{ "bits": 4, "group_size": 64, "lm_head": { "bits": 6, "group_size": 64, "mode": "affine" }, "mode": "affine", "model.embed_tokens": { "bits": 6, "group_size": 64, "mode": "affine" }, "model.layers.0.mlp.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.mlp.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.mlp.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.0.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.1.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.1.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.1.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.1.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.10.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.10.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.10.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.10.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.11.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.11.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.11.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.11.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.12.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.12.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.12.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.12.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.13.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.13.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.13.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.13.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.14.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.14.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.14.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.14.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.15.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.15.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.15.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.15.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.16.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.16.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.16.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.16.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.17.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.17.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.17.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.17.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.18.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.18.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.18.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.18.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.19.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.19.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.19.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.19.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.2.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.2.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.2.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.2.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.20.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.20.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.20.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.20.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.21.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.21.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.21.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.21.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.22.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.22.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.22.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.22.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.23.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.23.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.23.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.23.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.24.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.24.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.24.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.24.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.25.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.25.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.25.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.25.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.26.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.26.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.26.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.26.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.27.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.27.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.27.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.27.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.28.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.28.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.28.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.28.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.29.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.29.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.29.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.29.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.3.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.3.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.3.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.3.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.30.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.30.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.30.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.30.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.31.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.31.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.31.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.31.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.32.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.32.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.32.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.32.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.33.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.33.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.33.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.33.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.34.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.34.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.34.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.34.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.35.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.35.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.35.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.35.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.36.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.36.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.36.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.36.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.37.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.37.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.37.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.37.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.38.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.38.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.38.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.38.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.39.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.39.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.39.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.39.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.4.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.4.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.4.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.4.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.40.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.40.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.40.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.40.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.41.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.41.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.41.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.41.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.42.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.42.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.42.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.42.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.43.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.43.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.43.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.43.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.44.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.44.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.44.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.44.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.45.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.45.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.45.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.45.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.46.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.46.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.46.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.46.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.47.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.47.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.47.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.47.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.5.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.5.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.5.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.5.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.6.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.6.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.6.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.6.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.7.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.7.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.7.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.7.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.8.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.8.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.8.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.8.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.mlp.share_expert.down_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.mlp.share_expert.gate_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.mlp.share_expert.up_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.mlp.switch_mlp.down_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.9.mlp.switch_mlp.gate_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.9.mlp.switch_mlp.up_proj": { "bits": 3, "group_size": 128, "mode": "affine" }, "model.layers.9.self_attn.g_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.self_attn.k_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.self_attn.o_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.self_attn.q_proj": { "bits": 4, "group_size": 64, "mode": "affine" }, "model.layers.9.self_attn.v_proj": { "bits": 4, "group_size": 64, "mode": "affine" } }