config.json

{
  "architectures": [
    "Glm4MoeForCausalLM"
  ],
  "attention_bias": true,
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoModelForCausalLM": "modeling_glm4_moe.Glm4MoeForCausalLM"
  },
  "dtype": "bfloat16",
  "eos_token_id": [
    151329,
    151336,
    151338
  ],
  "first_k_dense_replace": 1,
  "head_dim": 128,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 10944,
  "max_position_embeddings": 131072,
  "model_type": "glm4_moe",
  "moe_intermediate_size": 1408,
  "n_group": 1,
  "n_routed_experts": 128,
  "n_shared_experts": 1,
  "norm_topk_prob": true,
  "num_attention_heads": 96,
  "num_experts_per_tok": 8,
  "num_hidden_layers": 46,
  "num_key_value_heads": 8,
  "num_nextn_predict_layers": 1,
  "pad_token_id": 151329,
  "partial_rotary_factor": 0.5,
  "quantization_config": {
    "bits": 4,
    "data_type": "int",
    "extra_config": {
      "model.layers.1.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.10.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.11.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.12.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.13.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.14.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.15.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.16.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.17.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.18.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.19.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.2.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.20.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.21.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.22.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.23.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.24.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.25.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.26.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.27.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.28.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.29.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.3.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.30.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.31.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.32.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.33.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.34.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.35.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.36.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.37.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.38.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.39.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.4.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.40.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.41.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.42.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.43.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.44.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.45.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.5.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.6.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.7.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.8.mlp.gate.gate_linear": {
        "bits": 8
      },
      "model.layers.9.mlp.gate.gate_linear": {
        "bits": 8
      }
    },
    "group_size": 128,
    "packing_format": "auto_round:auto_gptq",
    "quant_method": "auto-round",
    "sym": true
  },
  "rms_norm_eps": 1e-05,
  "rope_scaling": null,
  "rope_theta": 1000000,
  "routed_scaling_factor": 1.0,
  "tie_word_embeddings": false,
  "topk_group": 1,
  "transformers_version": "4.57.1",
  "use_cache": true,
  "use_qk_norm": false,
  "vocab_size": 151552
}