{
  "architectures": [
    "Qwen3ForCausalLM"
  ],
  "attention_bias": false,
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "eos_token_id": 151645,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 12288,
  "max_position_embeddings": 40960,
  "max_window_layers": 36,
  "model_type": "qwen3",
  "num_attention_heads": 32,
  "num_hidden_layers": 36,
  "num_key_value_heads": 8,
  "rms_norm_eps": 1e-06,
  "rope_scaling": null,
  "rope_theta": 1000000,
  "sliding_window": null,
  "tie_word_embeddings": true,
  "torch_dtype": "bfloat16",
  "transformers_version": "4.51.2",
  "use_cache": true,
  "use_sliding_window": false,
  "vocab_size": 151665,
  "quantization": {
    "group_size": 64,
    "bits": 4,
    "mode": "affine",
    "model.layers.0.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.6.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.6.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.7.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.7.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.8.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.8.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.9.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.10.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.10.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.11.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.11.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.12.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.12.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.13.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.13.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.14.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.14.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.15.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.15.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.33.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.34.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    }
  },
  "quantization_config": {
    "group_size": 64,
    "bits": 4,
    "mode": "affine",
    "model.layers.0.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.0.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.1.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.2.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.3.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.4.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.mlp.down_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.q_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.5.self_attn.v_proj": {
      "bits": 6,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.6.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.6.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.7.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.7.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.8.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.8.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.9.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.10.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.10.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.11.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.11.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.12.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.12.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.13.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.13.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.14.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.14.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.15.mlp.down_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.15.self_attn.o_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.33.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    },
    "model.layers.34.self_attn.k_proj": {
      "bits": 5,
      "group_size": 64,
      "mode": "affine"
    }
  }
}