gpt-oss-120b-oQ4 / config.json
cjnielson44's picture
Add files using upload-large-folder tool
9960f0e verified
Raw
History Blame Contribute Delete
27.1 kB
{
"architectures": [
"GptOssForCausalLM"
],
"attention_bias": true,
"attention_dropout": 0.0,
"eos_token_id": 200002,
"experts_per_token": 4,
"head_dim": 64,
"hidden_act": "silu",
"hidden_size": 2880,
"initial_context_length": 4096,
"initializer_range": 0.02,
"intermediate_size": 2880,
"layer_types": [
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention",
"sliding_attention",
"full_attention"
],
"max_position_embeddings": 131072,
"model_type": "gpt_oss",
"num_attention_heads": 64,
"num_experts_per_tok": 4,
"num_hidden_layers": 36,
"num_key_value_heads": 8,
"num_local_experts": 128,
"output_router_logits": false,
"pad_token_id": 199999,
"quantization_config": {
"group_size": 64,
"bits": 4,
"mode": "affine",
"model.layers.0.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.0.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.0.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.1.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.1.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.1.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.2.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.2.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.2.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.3.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.3.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.3.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.4.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.4.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.4.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.5.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.5.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.5.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.6.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.6.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.6.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.7.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.7.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.7.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.8.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.8.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.8.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.9.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.9.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.9.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.10.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.10.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.10.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.11.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.11.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.11.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.12.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.12.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.12.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.13.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.13.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.13.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.14.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.14.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.14.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.15.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.15.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.15.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.16.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.16.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.16.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.17.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.17.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.17.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.18.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.18.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.18.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.19.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.19.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.19.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.20.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.20.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.20.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.21.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.21.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.21.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.22.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.22.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.22.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.23.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.23.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.23.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.24.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.24.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.24.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.25.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.25.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.25.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.26.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.26.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.26.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.27.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.27.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.27.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.28.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.28.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.28.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.29.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.29.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.29.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.30.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.30.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.30.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.31.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.31.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.31.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.32.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.32.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.32.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.33.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.33.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.33.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.34.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.34.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.34.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.35.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.35.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.35.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
}
},
"rms_norm_eps": 1e-05,
"rope_scaling": {
"beta_fast": 32.0,
"beta_slow": 1.0,
"factor": 32.0,
"original_max_position_embeddings": 4096,
"rope_type": "yarn",
"truncate": false
},
"rope_theta": 150000,
"router_aux_loss_coef": 0.9,
"sliding_window": 128,
"swiglu_limit": 7.0,
"tie_word_embeddings": false,
"transformers_version": "4.55.0.dev0",
"use_cache": true,
"vocab_size": 201088,
"quantization": {
"group_size": 64,
"bits": 4,
"mode": "affine",
"model.layers.0.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.0.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.0.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.1.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.1.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.1.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.2.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.2.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.2.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.3.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.3.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.3.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.4.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.4.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.4.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.5.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.5.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.5.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.6.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.6.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.6.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.7.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.7.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.7.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.8.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.8.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.8.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.9.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.9.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.9.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.10.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.10.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.10.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.11.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.11.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.11.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.12.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.12.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.12.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.13.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.13.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.13.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.14.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.14.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.14.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.15.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.15.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.15.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.16.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.16.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.16.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.17.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.17.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.17.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.18.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.18.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.18.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.19.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.19.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.19.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.20.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.20.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.20.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.21.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.21.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.21.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.22.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.22.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.22.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.23.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.23.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.23.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.24.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.24.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.24.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.25.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.25.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.25.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.26.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.26.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.26.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.27.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.27.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.27.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.28.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.28.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.28.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.29.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.29.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.29.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.30.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.30.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.30.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.31.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.31.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.31.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.32.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.32.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.32.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.33.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.33.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.33.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.34.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.34.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.34.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.35.mlp.experts.gate_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.35.mlp.experts.up_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
},
"model.layers.35.mlp.experts.down_proj": {
"group_size": 32,
"bits": 4,
"mode": "mxfp4"
}
}
}