model.visual.patch_embed.proj.input_quantizer TensorQuantizer(disabled) model.visual.patch_embed.proj.output_quantizer TensorQuantizer(disabled) model.visual.patch_embed.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.0.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.0.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.0.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.0.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.0.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.0.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.0.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.0.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.0.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.0.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.0.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.1.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.1.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.1.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.1.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.1.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.1.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.1.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.1.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.1.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.1.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.1.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.2.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.2.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.2.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.2.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.2.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.2.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.2.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.2.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.2.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.2.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.2.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.3.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.3.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.3.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.3.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.3.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.3.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.3.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.3.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.3.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.3.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.3.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.4.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.4.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.4.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.4.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.4.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.4.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.4.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.4.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.4.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.4.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.4.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.5.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.5.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.5.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.5.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.5.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.5.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.5.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.5.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.5.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.5.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.5.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.6.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.6.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.6.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.6.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.6.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.6.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.6.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.6.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.6.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.6.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.6.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.7.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.7.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.7.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.7.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.7.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.7.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.7.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.7.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.7.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.7.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.7.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.8.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.8.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.8.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.8.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.8.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.8.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.8.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.8.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.8.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.8.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.8.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.9.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.9.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.9.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.9.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.9.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.9.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.9.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.9.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.9.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.9.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.9.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.10.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.10.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.10.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.10.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.10.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.10.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.10.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.10.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.10.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.10.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.10.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.11.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.11.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.11.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.11.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.11.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.11.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.11.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.11.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.11.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.11.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.11.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.12.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.12.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.12.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.12.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.12.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.12.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.12.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.12.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.12.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.12.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.12.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.13.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.13.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.13.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.13.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.13.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.13.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.13.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.13.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.13.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.13.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.13.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.14.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.14.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.14.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.14.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.14.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.14.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.14.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.14.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.14.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.14.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.14.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.15.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.15.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.15.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.15.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.15.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.15.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.15.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.15.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.15.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.15.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.15.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.16.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.16.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.16.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.16.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.16.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.16.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.16.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.16.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.16.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.16.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.16.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.17.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.17.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.17.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.17.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.17.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.17.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.17.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.17.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.17.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.17.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.17.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.18.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.18.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.18.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.18.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.18.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.18.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.18.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.18.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.18.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.18.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.18.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.19.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.19.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.19.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.19.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.19.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.19.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.19.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.19.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.19.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.19.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.19.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.20.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.20.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.20.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.20.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.20.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.20.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.20.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.20.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.20.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.20.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.20.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.21.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.21.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.21.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.21.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.21.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.21.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.21.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.21.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.21.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.21.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.21.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.22.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.22.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.22.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.22.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.22.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.22.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.22.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.22.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.22.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.22.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.22.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.23.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.23.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.23.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.23.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.23.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.23.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.23.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.23.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.23.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.23.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.23.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.24.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.24.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.24.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.24.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.24.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.24.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.24.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.24.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.24.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.24.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.24.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.25.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.25.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.25.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.25.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.25.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.25.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.25.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.25.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.25.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.25.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.25.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.26.norm1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.26.norm1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.26.norm2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.26.norm2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.qkv.input_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.qkv.output_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.qkv.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.proj.input_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.proj.output_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.proj.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.q_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.k_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.v_bmm_quantizer TensorQuantizer(disabled) model.visual.blocks.26.attn.softmax_quantizer TensorQuantizer(disabled) model.visual.blocks.26.mlp.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.blocks.26.mlp.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.blocks.26.mlp.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.blocks.26.mlp.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.blocks.26.mlp.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.blocks.26.mlp.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.visual.merger.norm.input_quantizer TensorQuantizer(disabled) model.visual.merger.norm.output_quantizer TensorQuantizer(disabled) model.visual.merger.linear_fc1.input_quantizer TensorQuantizer(disabled) model.visual.merger.linear_fc1.output_quantizer TensorQuantizer(disabled) model.visual.merger.linear_fc1.weight_quantizer TensorQuantizer(disabled) model.visual.merger.linear_fc2.input_quantizer TensorQuantizer(disabled) model.visual.merger.linear_fc2.output_quantizer TensorQuantizer(disabled) model.visual.merger.linear_fc2.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.8750 calibrator=MaxCalibrator quant) model.language_model.layers.0.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4277 calibrator=MaxCalibrator quant) model.language_model.layers.0.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.0000 calibrator=MaxCalibrator quant) model.language_model.layers.0.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3848 calibrator=MaxCalibrator quant) model.language_model.layers.0.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.0000 calibrator=MaxCalibrator quant) model.language_model.layers.0.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2871 calibrator=MaxCalibrator quant) model.language_model.layers.0.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.0.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.6250 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0442 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0469 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0461 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0457 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=99.0000 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.6250 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.6250 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.8750 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.language_model.layers.0.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.0.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.0.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.4453 calibrator=MaxCalibrator quant) model.language_model.layers.1.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6094 calibrator=MaxCalibrator quant) model.language_model.layers.1.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=32.2500 calibrator=MaxCalibrator quant) model.language_model.layers.1.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2773 calibrator=MaxCalibrator quant) model.language_model.layers.1.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=32.2500 calibrator=MaxCalibrator quant) model.language_model.layers.1.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2930 calibrator=MaxCalibrator quant) model.language_model.layers.1.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.1.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.2500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0454 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=13.7500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.2500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.2500 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.0625 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4082 calibrator=MaxCalibrator quant) model.language_model.layers.1.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.1.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.1.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.9141 calibrator=MaxCalibrator quant) model.language_model.layers.2.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6680 calibrator=MaxCalibrator quant) model.language_model.layers.2.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.2500 calibrator=MaxCalibrator quant) model.language_model.layers.2.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2656 calibrator=MaxCalibrator quant) model.language_model.layers.2.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.2500 calibrator=MaxCalibrator quant) model.language_model.layers.2.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2910 calibrator=MaxCalibrator quant) model.language_model.layers.2.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.2.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=13.1875 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.8750 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=13.1875 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=13.1875 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=21.2500 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4492 calibrator=MaxCalibrator quant) model.language_model.layers.2.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.2.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.2.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.5000 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3418 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.5000 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.5000 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.3203 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.5859 calibrator=MaxCalibrator quant) model.language_model.layers.3.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.3.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.3.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.1250 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.5000 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5195 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.1250 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=16.1250 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=14.6875 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.language_model.layers.3.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.3.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.3.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.3359 calibrator=MaxCalibrator quant) model.language_model.layers.4.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7383 calibrator=MaxCalibrator quant) model.language_model.layers.4.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=23.6250 calibrator=MaxCalibrator quant) model.language_model.layers.4.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2734 calibrator=MaxCalibrator quant) model.language_model.layers.4.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=23.6250 calibrator=MaxCalibrator quant) model.language_model.layers.4.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2773 calibrator=MaxCalibrator quant) model.language_model.layers.4.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.4.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.2500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=19.7500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.2500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.2500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=19.2500 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4355 calibrator=MaxCalibrator quant) model.language_model.layers.4.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.4.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.4.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.2344 calibrator=MaxCalibrator quant) model.language_model.layers.5.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7461 calibrator=MaxCalibrator quant) model.language_model.layers.5.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=29.7500 calibrator=MaxCalibrator quant) model.language_model.layers.5.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3281 calibrator=MaxCalibrator quant) model.language_model.layers.5.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=29.7500 calibrator=MaxCalibrator quant) model.language_model.layers.5.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2871 calibrator=MaxCalibrator quant) model.language_model.layers.5.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.5.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=18.1250 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0459 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0476 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3906 calibrator=MaxCalibrator quant) model.language_model.layers.5.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.5.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.5.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=17.8750 calibrator=MaxCalibrator quant) model.language_model.layers.6.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6602 calibrator=MaxCalibrator quant) model.language_model.layers.6.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=29.1250 calibrator=MaxCalibrator quant) model.language_model.layers.6.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3887 calibrator=MaxCalibrator quant) model.language_model.layers.6.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=29.1250 calibrator=MaxCalibrator quant) model.language_model.layers.6.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2734 calibrator=MaxCalibrator quant) model.language_model.layers.6.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.6.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=31.8750 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=18.0000 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0466 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=31.8750 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=31.8750 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=24.1250 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.language_model.layers.6.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.6.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.6.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2949 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3105 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.9609 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6211 calibrator=MaxCalibrator quant) model.language_model.layers.7.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.7.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.7.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.5000 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=18.3750 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.5000 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.5000 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.6250 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4395 calibrator=MaxCalibrator quant) model.language_model.layers.7.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.7.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.7.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.7578 calibrator=MaxCalibrator quant) model.language_model.layers.8.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6406 calibrator=MaxCalibrator quant) model.language_model.layers.8.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.7500 calibrator=MaxCalibrator quant) model.language_model.layers.8.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3047 calibrator=MaxCalibrator quant) model.language_model.layers.8.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.7500 calibrator=MaxCalibrator quant) model.language_model.layers.8.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3516 calibrator=MaxCalibrator quant) model.language_model.layers.8.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.8.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.2500 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=21.1250 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0474 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.2500 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.2500 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=19.8750 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4414 calibrator=MaxCalibrator quant) model.language_model.layers.8.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.8.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.8.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.3438 calibrator=MaxCalibrator quant) model.language_model.layers.9.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7188 calibrator=MaxCalibrator quant) model.language_model.layers.9.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=38.7500 calibrator=MaxCalibrator quant) model.language_model.layers.9.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2852 calibrator=MaxCalibrator quant) model.language_model.layers.9.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=38.7500 calibrator=MaxCalibrator quant) model.language_model.layers.9.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.9.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.9.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=34.2500 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.2500 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=34.2500 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=34.2500 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=11.4375 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.language_model.layers.9.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.9.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.9.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=61.7500 calibrator=MaxCalibrator quant) model.language_model.layers.10.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8281 calibrator=MaxCalibrator quant) model.language_model.layers.10.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.5000 calibrator=MaxCalibrator quant) model.language_model.layers.10.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2520 calibrator=MaxCalibrator quant) model.language_model.layers.10.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.5000 calibrator=MaxCalibrator quant) model.language_model.layers.10.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3047 calibrator=MaxCalibrator quant) model.language_model.layers.10.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.10.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=45.7500 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=19.6250 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=45.7500 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=45.7500 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.0000 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4570 calibrator=MaxCalibrator quant) model.language_model.layers.10.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.10.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.10.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=39.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2559 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=39.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=39.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=3.1719 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6367 calibrator=MaxCalibrator quant) model.language_model.layers.11.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.11.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.11.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=18.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.2500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.7500 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.language_model.layers.11.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.11.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.11.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.4453 calibrator=MaxCalibrator quant) model.language_model.layers.12.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7383 calibrator=MaxCalibrator quant) model.language_model.layers.12.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2812 calibrator=MaxCalibrator quant) model.language_model.layers.12.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2402 calibrator=MaxCalibrator quant) model.language_model.layers.12.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.12.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=45.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=19.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0464 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=45.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=45.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.2500 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5234 calibrator=MaxCalibrator quant) model.language_model.layers.12.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.12.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.12.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=6.9688 calibrator=MaxCalibrator quant) model.language_model.layers.13.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7812 calibrator=MaxCalibrator quant) model.language_model.layers.13.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.13.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4297 calibrator=MaxCalibrator quant) model.language_model.layers.13.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.13.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3398 calibrator=MaxCalibrator quant) model.language_model.layers.13.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.13.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.7500 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=22.3750 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0481 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.7500 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.7500 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=21.5000 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5352 calibrator=MaxCalibrator quant) model.language_model.layers.13.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.13.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.13.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.7656 calibrator=MaxCalibrator quant) model.language_model.layers.14.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7070 calibrator=MaxCalibrator quant) model.language_model.layers.14.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=49.2500 calibrator=MaxCalibrator quant) model.language_model.layers.14.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3828 calibrator=MaxCalibrator quant) model.language_model.layers.14.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=49.2500 calibrator=MaxCalibrator quant) model.language_model.layers.14.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3203 calibrator=MaxCalibrator quant) model.language_model.layers.14.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.14.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.0000 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=21.5000 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.0000 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.0000 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=32.7500 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.language_model.layers.14.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.14.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.14.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2178 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.4844 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7070 calibrator=MaxCalibrator quant) model.language_model.layers.15.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.15.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.15.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=21.5000 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4199 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=24.6250 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.language_model.layers.15.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.15.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.15.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.2812 calibrator=MaxCalibrator quant) model.language_model.layers.16.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7422 calibrator=MaxCalibrator quant) model.language_model.layers.16.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.7500 calibrator=MaxCalibrator quant) model.language_model.layers.16.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4043 calibrator=MaxCalibrator quant) model.language_model.layers.16.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.7500 calibrator=MaxCalibrator quant) model.language_model.layers.16.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.16.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.16.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=49.0000 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=17.8750 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=49.0000 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=49.0000 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=24.3750 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5273 calibrator=MaxCalibrator quant) model.language_model.layers.16.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.16.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.16.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.1406 calibrator=MaxCalibrator quant) model.language_model.layers.17.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7500 calibrator=MaxCalibrator quant) model.language_model.layers.17.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=42.2500 calibrator=MaxCalibrator quant) model.language_model.layers.17.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3223 calibrator=MaxCalibrator quant) model.language_model.layers.17.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=42.2500 calibrator=MaxCalibrator quant) model.language_model.layers.17.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3066 calibrator=MaxCalibrator quant) model.language_model.layers.17.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.17.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.8750 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4316 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3711 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.1250 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.language_model.layers.17.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.17.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.17.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.1406 calibrator=MaxCalibrator quant) model.language_model.layers.18.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6602 calibrator=MaxCalibrator quant) model.language_model.layers.18.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.5000 calibrator=MaxCalibrator quant) model.language_model.layers.18.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3301 calibrator=MaxCalibrator quant) model.language_model.layers.18.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.5000 calibrator=MaxCalibrator quant) model.language_model.layers.18.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3125 calibrator=MaxCalibrator quant) model.language_model.layers.18.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.18.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=21.1250 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3379 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.language_model.layers.18.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.18.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.18.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3965 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2236 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.5156 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6719 calibrator=MaxCalibrator quant) model.language_model.layers.19.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.19.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.19.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.2500 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=27.1250 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.2500 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.2500 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.language_model.layers.19.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.19.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.19.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.5391 calibrator=MaxCalibrator quant) model.language_model.layers.20.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7852 calibrator=MaxCalibrator quant) model.language_model.layers.20.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=58.2500 calibrator=MaxCalibrator quant) model.language_model.layers.20.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4102 calibrator=MaxCalibrator quant) model.language_model.layers.20.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=58.2500 calibrator=MaxCalibrator quant) model.language_model.layers.20.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3262 calibrator=MaxCalibrator quant) model.language_model.layers.20.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.20.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.7500 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=31.3750 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0488 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.7500 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.7500 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=18.8750 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.language_model.layers.20.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.20.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.20.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=3.0312 calibrator=MaxCalibrator quant) model.language_model.layers.21.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8633 calibrator=MaxCalibrator quant) model.language_model.layers.21.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.7500 calibrator=MaxCalibrator quant) model.language_model.layers.21.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.5156 calibrator=MaxCalibrator quant) model.language_model.layers.21.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.7500 calibrator=MaxCalibrator quant) model.language_model.layers.21.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3555 calibrator=MaxCalibrator quant) model.language_model.layers.21.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.21.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.5000 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=24.5000 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0486 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.5000 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.5000 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=15.3125 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4961 calibrator=MaxCalibrator quant) model.language_model.layers.21.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.21.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.21.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.4375 calibrator=MaxCalibrator quant) model.language_model.layers.22.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.9297 calibrator=MaxCalibrator quant) model.language_model.layers.22.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=39.0000 calibrator=MaxCalibrator quant) model.language_model.layers.22.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3301 calibrator=MaxCalibrator quant) model.language_model.layers.22.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=39.0000 calibrator=MaxCalibrator quant) model.language_model.layers.22.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3652 calibrator=MaxCalibrator quant) model.language_model.layers.22.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.22.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=53.5000 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=19.6250 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0505 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3457 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=53.5000 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=53.5000 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=31.7500 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3555 calibrator=MaxCalibrator quant) model.language_model.layers.22.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.22.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.22.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=50.2500 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4102 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=50.2500 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=50.2500 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.6406 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6953 calibrator=MaxCalibrator quant) model.language_model.layers.23.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.23.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.23.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=56.7500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=20.2500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4219 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0493 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=56.7500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=56.7500 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=13.3125 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5312 calibrator=MaxCalibrator quant) model.language_model.layers.23.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.23.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.23.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=1.9141 calibrator=MaxCalibrator quant) model.language_model.layers.24.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8047 calibrator=MaxCalibrator quant) model.language_model.layers.24.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.24.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3555 calibrator=MaxCalibrator quant) model.language_model.layers.24.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.24.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3125 calibrator=MaxCalibrator quant) model.language_model.layers.24.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.24.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.2500 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=25.1250 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0479 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0498 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0491 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.2500 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.2500 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=13.5000 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.8438 calibrator=MaxCalibrator quant) model.language_model.layers.24.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.24.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.24.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=7.9375 calibrator=MaxCalibrator quant) model.language_model.layers.25.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8945 calibrator=MaxCalibrator quant) model.language_model.layers.25.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.5000 calibrator=MaxCalibrator quant) model.language_model.layers.25.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3105 calibrator=MaxCalibrator quant) model.language_model.layers.25.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.5000 calibrator=MaxCalibrator quant) model.language_model.layers.25.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3105 calibrator=MaxCalibrator quant) model.language_model.layers.25.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.25.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.7500 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=25.0000 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0471 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.7500 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.7500 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=30.1250 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4805 calibrator=MaxCalibrator quant) model.language_model.layers.25.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.25.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.25.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.3906 calibrator=MaxCalibrator quant) model.language_model.layers.26.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6797 calibrator=MaxCalibrator quant) model.language_model.layers.26.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=47.7500 calibrator=MaxCalibrator quant) model.language_model.layers.26.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4766 calibrator=MaxCalibrator quant) model.language_model.layers.26.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=47.7500 calibrator=MaxCalibrator quant) model.language_model.layers.26.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3555 calibrator=MaxCalibrator quant) model.language_model.layers.26.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.26.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.2500 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0503 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=25.1250 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0513 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0510 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.2500 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.2500 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=47.5000 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.language_model.layers.26.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.26.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.26.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3340 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=47.2500 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2559 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=4.5625 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6602 calibrator=MaxCalibrator quant) model.language_model.layers.27.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.27.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.27.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=27.2500 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0483 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0496 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0508 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=49.5000 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4863 calibrator=MaxCalibrator quant) model.language_model.layers.27.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.27.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.27.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.0938 calibrator=MaxCalibrator quant) model.language_model.layers.28.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7695 calibrator=MaxCalibrator quant) model.language_model.layers.28.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=48.7500 calibrator=MaxCalibrator quant) model.language_model.layers.28.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.5664 calibrator=MaxCalibrator quant) model.language_model.layers.28.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=48.7500 calibrator=MaxCalibrator quant) model.language_model.layers.28.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3848 calibrator=MaxCalibrator quant) model.language_model.layers.28.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.28.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=56.2500 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=24.7500 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=56.2500 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=56.2500 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=39.0000 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.language_model.layers.28.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.28.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.28.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.2969 calibrator=MaxCalibrator quant) model.language_model.layers.29.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8398 calibrator=MaxCalibrator quant) model.language_model.layers.29.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.5000 calibrator=MaxCalibrator quant) model.language_model.layers.29.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4336 calibrator=MaxCalibrator quant) model.language_model.layers.29.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.5000 calibrator=MaxCalibrator quant) model.language_model.layers.29.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3301 calibrator=MaxCalibrator quant) model.language_model.layers.29.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.29.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0518 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=27.3750 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0500 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=24.3750 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4727 calibrator=MaxCalibrator quant) model.language_model.layers.29.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.29.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.29.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=4.1562 calibrator=MaxCalibrator quant) model.language_model.layers.30.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8828 calibrator=MaxCalibrator quant) model.language_model.layers.30.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.0000 calibrator=MaxCalibrator quant) model.language_model.layers.30.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3418 calibrator=MaxCalibrator quant) model.language_model.layers.30.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=43.0000 calibrator=MaxCalibrator quant) model.language_model.layers.30.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4355 calibrator=MaxCalibrator quant) model.language_model.layers.30.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.30.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0520 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0527 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0522 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0552 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=32.7500 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0515 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=55.7500 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=75.5000 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.language_model.layers.30.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.30.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.30.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=38.7500 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3848 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=38.7500 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=38.7500 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2676 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=17.0000 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8320 calibrator=MaxCalibrator quant) model.language_model.layers.31.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.31.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.31.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=53.2500 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=40.0000 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0537 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0535 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=53.2500 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=53.2500 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=39.5000 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4648 calibrator=MaxCalibrator quant) model.language_model.layers.31.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.31.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.31.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=7.2812 calibrator=MaxCalibrator quant) model.language_model.layers.32.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8633 calibrator=MaxCalibrator quant) model.language_model.layers.32.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.2500 calibrator=MaxCalibrator quant) model.language_model.layers.32.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.5898 calibrator=MaxCalibrator quant) model.language_model.layers.32.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.2500 calibrator=MaxCalibrator quant) model.language_model.layers.32.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3379 calibrator=MaxCalibrator quant) model.language_model.layers.32.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.32.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.5000 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=38.2500 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.5000 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=50.5000 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=40.0000 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4688 calibrator=MaxCalibrator quant) model.language_model.layers.32.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.32.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.32.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=12.5625 calibrator=MaxCalibrator quant) model.language_model.layers.33.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.9570 calibrator=MaxCalibrator quant) model.language_model.layers.33.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=33.7500 calibrator=MaxCalibrator quant) model.language_model.layers.33.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4844 calibrator=MaxCalibrator quant) model.language_model.layers.33.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=33.7500 calibrator=MaxCalibrator quant) model.language_model.layers.33.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3574 calibrator=MaxCalibrator quant) model.language_model.layers.33.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.33.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=52.2500 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0530 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0532 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2432 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=51.7500 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0544 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3770 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3828 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0542 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=52.2500 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=52.2500 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=85.0000 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.language_model.layers.33.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.33.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.33.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=4.7500 calibrator=MaxCalibrator quant) model.language_model.layers.34.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.8594 calibrator=MaxCalibrator quant) model.language_model.layers.34.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=34.0000 calibrator=MaxCalibrator quant) model.language_model.layers.34.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3457 calibrator=MaxCalibrator quant) model.language_model.layers.34.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=34.0000 calibrator=MaxCalibrator quant) model.language_model.layers.34.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3242 calibrator=MaxCalibrator quant) model.language_model.layers.34.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.34.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3145 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3281 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3691 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2041 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=46.0000 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2070 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=44.2500 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=167.0000 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3730 calibrator=MaxCalibrator quant) model.language_model.layers.34.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.34.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.34.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4336 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2988 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=35.5000 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3730 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=7.8750 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6172 calibrator=MaxCalibrator quant) model.language_model.layers.35.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.35.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.35.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.2500 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4004 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4023 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3574 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=79.5000 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0540 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0525 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.2500 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.2500 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=132.0000 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4180 calibrator=MaxCalibrator quant) model.language_model.layers.35.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.35.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.35.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=46.0000 calibrator=MaxCalibrator quant) model.language_model.layers.36.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=2.1406 calibrator=MaxCalibrator quant) model.language_model.layers.36.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.7500 calibrator=MaxCalibrator quant) model.language_model.layers.36.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4512 calibrator=MaxCalibrator quant) model.language_model.layers.36.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.7500 calibrator=MaxCalibrator quant) model.language_model.layers.36.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3926 calibrator=MaxCalibrator quant) model.language_model.layers.36.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.36.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3809 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4258 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3418 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2178 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3105 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3398 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=76.5000 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0586 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0569 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0559 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0554 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=48.0000 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2969 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=80.5000 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4629 calibrator=MaxCalibrator quant) model.language_model.layers.36.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.36.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.36.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=28.3750 calibrator=MaxCalibrator quant) model.language_model.layers.37.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6328 calibrator=MaxCalibrator quant) model.language_model.layers.37.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.2500 calibrator=MaxCalibrator quant) model.language_model.layers.37.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.6211 calibrator=MaxCalibrator quant) model.language_model.layers.37.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.2500 calibrator=MaxCalibrator quant) model.language_model.layers.37.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3965 calibrator=MaxCalibrator quant) model.language_model.layers.37.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.37.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=43.7500 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3496 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2227 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3789 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3242 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0571 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2471 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0615 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2461 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2812 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=77.5000 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1670 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0618 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0952 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0840 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0608 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0547 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0596 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0698 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0576 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0574 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0623 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0601 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0581 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0557 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=43.7500 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=43.7500 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=84.5000 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3926 calibrator=MaxCalibrator quant) model.language_model.layers.37.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.37.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.37.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.conv1d.input_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.conv1d.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.conv1d.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.out_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=40.2500 calibrator=MaxCalibrator quant) model.language_model.layers.38.linear_attn.out_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.out_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.5898 calibrator=MaxCalibrator quant) model.language_model.layers.38.linear_attn.in_proj_qkv.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.5000 calibrator=MaxCalibrator quant) model.language_model.layers.38.linear_attn.in_proj_qkv.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_qkv.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.7305 calibrator=MaxCalibrator quant) model.language_model.layers.38.linear_attn.in_proj_z.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=36.5000 calibrator=MaxCalibrator quant) model.language_model.layers.38.linear_attn.in_proj_z.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_z.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4102 calibrator=MaxCalibrator quant) model.language_model.layers.38.linear_attn.in_proj_b.input_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_b.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_b.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_a.input_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_a.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.linear_attn.in_proj_a.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.38.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=25.8750 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1914 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0598 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0579 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2715 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2871 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1650 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0605 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0562 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2314 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2236 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2441 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2402 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0884 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0613 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3301 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1504 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2383 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0591 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3594 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2637 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3945 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1963 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2422 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1089 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2246 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=152.0000 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0796 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0593 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1865 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1807 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0564 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0583 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0693 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0947 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1094 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2559 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1074 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0669 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0640 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0938 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0806 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1211 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0708 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1025 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0566 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0674 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1245 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0620 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0645 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1787 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0732 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0894 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1030 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0635 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1855 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1216 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0718 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1816 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0728 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0825 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0603 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1543 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0879 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1040 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=25.8750 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=25.8750 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=176.0000 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3965 calibrator=MaxCalibrator quant) model.language_model.layers.38.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.38.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.38.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.q_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=37.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.q_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.q_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.4219 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.k_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=37.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.k_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.k_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.v_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=37.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.v_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.v_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.3867 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.o_proj.input_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=15.6250 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.o_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.o_proj.weight_quantizer TensorQuantizer((4, 3) bit fake per-tensor amax=0.5430 calibrator=MaxCalibrator quant) model.language_model.layers.39.self_attn.q_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.k_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.v_bmm_quantizer TensorQuantizer(disabled) model.language_model.layers.39.self_attn.softmax_quantizer TensorQuantizer(disabled) model.language_model.layers.39.mlp.experts.gate_up_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1128 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0771 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0918 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1592 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2090 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0986 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1240 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0850 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0811 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1602 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3125 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2129 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2305 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1768 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2158 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0962 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0762 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0869 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3066 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1680 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0742 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1299 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3203 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3184 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0625 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0981 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2363 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1416 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0845 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0610 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1084 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2334 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2656 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2285 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1846 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1050 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2373 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1738 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2988 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0898 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2676 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1836 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0786 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0933 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2080 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1152 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1011 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1108 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0864 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1187 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0815 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2061 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0996 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0928 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1191 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2393 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2754 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0791 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2051 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1221 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3750 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2168 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2891 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0713 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1426 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1777 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2734 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1133 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1206 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2451 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0820 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1045 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2109 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1162 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2002 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1177 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1079 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0977 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.gate_up_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=125.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.0 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.1 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1973 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.2 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.3 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2793 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.4 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.5 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3672 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.6 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1377 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.7 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.8 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.9 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0874 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.10 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0776 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.11 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.12 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2617 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.13 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.14 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0679 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.15 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1064 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.16 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.17 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.18 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1621 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.19 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.20 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1270 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.21 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.22 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1895 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.23 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.24 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.25 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2266 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.26 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1235 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.27 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.28 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1318 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.29 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.30 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.31 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2910 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.32 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.33 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.34 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.35 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.36 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.37 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.38 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1006 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.39 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.40 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2695 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.41 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1582 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.42 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2295 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.43 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1182 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.44 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0781 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.45 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.46 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.47 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.48 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.49 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.50 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5781 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.51 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.52 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0913 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.53 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.54 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1758 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.55 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2139 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.56 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.57 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3848 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.58 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2275 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.59 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0752 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.60 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.61 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0549 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.62 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.63 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.64 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.65 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0967 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.66 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1147 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.67 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.68 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1475 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.69 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.70 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1943 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.71 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2344 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.72 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.73 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1455 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.74 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2773 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.75 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3086 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.76 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1953 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.77 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.78 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0737 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.79 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.80 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.81 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1250 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.82 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1279 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.83 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4707 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.84 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0649 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.85 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2188 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.86 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.87 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.88 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0835 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.89 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.90 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1885 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.91 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.92 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2930 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.93 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2031 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.94 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.95 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.96 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4590 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.97 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1060 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.98 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2324 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.99 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.100 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1387 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.101 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4434 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.102 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.103 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.104 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2119 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.105 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.106 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.107 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.108 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2520 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.109 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.110 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.111 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.112 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.113 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2412 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.114 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1230 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.115 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0688 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.116 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.117 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.118 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1641 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.119 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.120 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.121 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1196 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.122 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1035 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.123 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.124 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.125 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0889 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.126 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.127 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3613 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.128 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.129 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.130 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0991 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.131 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0923 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.132 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5156 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.133 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.134 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1118 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.135 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1709 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.136 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1167 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.137 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0630 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.138 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1484 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.139 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1289 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.140 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.141 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3340 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.142 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.143 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1875 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.144 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.145 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1553 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.146 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.147 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0654 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.148 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.149 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5664 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.150 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2539 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.151 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1719 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.152 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.153 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1982 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.154 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1533 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.155 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.4531 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.156 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.157 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.158 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.159 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3027 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.160 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1367 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.161 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0723 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.162 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2832 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.163 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1562 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.164 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1099 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.165 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2207 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.166 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0801 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.167 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1055 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.168 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0659 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.169 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.170 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2500 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.171 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0903 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.172 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0957 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.173 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0703 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.174 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.175 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.176 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1572 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.177 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1016 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.178 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.179 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1001 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.180 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1309 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.181 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0767 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.182 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1113 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.183 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1748 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.184 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3633 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.185 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1445 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.186 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2256 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.187 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1729 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.188 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0588 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.189 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2100 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.190 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2354 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.191 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1494 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.192 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0854 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.193 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1924 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.194 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1069 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.195 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2021 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.196 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0684 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.197 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1157 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.198 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1138 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.199 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1123 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.200 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5977 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.201 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1143 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.202 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1226 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.203 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.204 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1465 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.205 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1172 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.206 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0747 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.207 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1436 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.208 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1104 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.209 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1660 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.210 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1260 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.211 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1797 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.212 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1201 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.213 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2148 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.214 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1631 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.215 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1406 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.216 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0664 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.217 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2217 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.218 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2490 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.219 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.220 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0859 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.221 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2197 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.222 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0757 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.223 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.224 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0972 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.225 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1396 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.226 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.227 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.228 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1826 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.229 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.230 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2852 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.231 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3223 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.232 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1992 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.233 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.234 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1904 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.235 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2578 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.236 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1689 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.237 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1021 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.238 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.6523 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.239 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1523 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.240 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1514 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.241 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.242 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0830 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.243 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.244 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1328 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.245 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1348 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.246 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1699 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.247 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2480 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.248 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1611 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.249 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1338 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.250 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2012 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.251 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3477 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.252 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0908 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.253 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1357 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.254 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.0942 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.experts.down_proj_weight_quantizers.255 TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.5117 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert.gate_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert.gate_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.mlp.shared_expert.gate_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.1934 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert.up_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=33.0000 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert.up_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.mlp.shared_expert.up_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.2949 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert.down_proj.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=54.5000 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert.down_proj.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.mlp.shared_expert.down_proj.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3047 calibrator=MaxCalibrator quant) model.language_model.layers.39.mlp.shared_expert_gate.input_quantizer TensorQuantizer(disabled) model.language_model.layers.39.mlp.shared_expert_gate.output_quantizer TensorQuantizer(disabled) model.language_model.layers.39.mlp.shared_expert_gate.weight_quantizer TensorQuantizer(disabled) lm_head.input_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=74.5000 calibrator=MaxCalibrator quant) lm_head.output_quantizer TensorQuantizer(disabled) lm_head.weight_quantizer TensorQuantizer((2, 1) bit fake block_sizes={-1: 16, 'type': 'dynamic', 'scale_bits': (4, 3)}, amax=0.3008 calibrator=MaxCalibrator quant) 22294 TensorQuantizers found in model