config.json

{
  "architectures": [
    "Qwen3VLMoeForConditionalGeneration"
  ],
  "dtype": "float32",
  "image_token_id": 151655,
  "model_type": "qwen3_vl_moe",
  "quantization_config": {
    "algo_config": null,
    "exclude": [
      "model.language_model.layers.0.mlp.gate",
      "model.language_model.layers.1.mlp.gate",
      "model.language_model.layers.2.mlp.gate",
      "model.language_model.layers.3.mlp.gate",
      "model.language_model.layers.4.mlp.gate",
      "model.language_model.layers.5.mlp.gate",
      "model.language_model.layers.6.mlp.gate",
      "model.language_model.layers.7.mlp.gate",
      "model.language_model.layers.8.mlp.gate",
      "model.language_model.layers.9.mlp.gate",
      "model.language_model.layers.10.mlp.gate",
      "model.language_model.layers.11.mlp.gate",
      "model.language_model.layers.12.mlp.gate",
      "model.language_model.layers.13.mlp.gate",
      "model.language_model.layers.14.mlp.gate",
      "model.language_model.layers.15.mlp.gate",
      "model.language_model.layers.16.mlp.gate",
      "model.language_model.layers.17.mlp.gate",
      "model.language_model.layers.18.mlp.gate",
      "model.language_model.layers.19.mlp.gate",
      "model.language_model.layers.20.mlp.gate",
      "model.language_model.layers.21.mlp.gate",
      "model.language_model.layers.22.mlp.gate",
      "model.language_model.layers.23.mlp.gate",
      "model.language_model.layers.24.mlp.gate",
      "model.language_model.layers.25.mlp.gate",
      "model.language_model.layers.26.mlp.gate",
      "model.language_model.layers.27.mlp.gate",
      "model.language_model.layers.28.mlp.gate",
      "model.language_model.layers.29.mlp.gate",
      "model.language_model.layers.30.mlp.gate",
      "model.language_model.layers.31.mlp.gate",
      "model.language_model.layers.32.mlp.gate",
      "model.language_model.layers.33.mlp.gate",
      "model.language_model.layers.34.mlp.gate",
      "model.language_model.layers.35.mlp.gate",
      "model.language_model.layers.36.mlp.gate",
      "model.language_model.layers.37.mlp.gate",
      "model.language_model.layers.38.mlp.gate",
      "model.language_model.layers.39.mlp.gate",
      "model.language_model.layers.40.mlp.gate",
      "model.language_model.layers.41.mlp.gate",
      "model.language_model.layers.42.mlp.gate",
      "model.language_model.layers.43.mlp.gate",
      "model.language_model.layers.44.mlp.gate",
      "model.language_model.layers.45.mlp.gate",
      "model.language_model.layers.46.mlp.gate",
      "model.language_model.layers.47.mlp.gate",
      "model.language_model.layers.48.mlp.gate",
      "model.language_model.layers.49.mlp.gate",
      "model.language_model.layers.50.mlp.gate",
      "model.language_model.layers.51.mlp.gate",
      "model.language_model.layers.52.mlp.gate",
      "model.language_model.layers.53.mlp.gate",
      "model.language_model.layers.54.mlp.gate",
      "model.language_model.layers.55.mlp.gate",
      "model.language_model.layers.56.mlp.gate",
      "model.language_model.layers.57.mlp.gate",
      "model.language_model.layers.58.mlp.gate",
      "model.language_model.layers.59.mlp.gate",
      "model.language_model.layers.60.mlp.gate",
      "model.language_model.layers.61.mlp.gate",
      "model.language_model.layers.62.mlp.gate",
      "model.language_model.layers.63.mlp.gate",
      "model.language_model.layers.64.mlp.gate",
      "model.language_model.layers.65.mlp.gate",
      "model.language_model.layers.66.mlp.gate",
      "model.language_model.layers.67.mlp.gate",
      "model.language_model.layers.68.mlp.gate",
      "model.language_model.layers.69.mlp.gate",
      "model.language_model.layers.70.mlp.gate",
      "model.language_model.layers.71.mlp.gate",
      "model.language_model.layers.72.mlp.gate",
      "model.language_model.layers.73.mlp.gate",
      "model.language_model.layers.74.mlp.gate",
      "model.language_model.layers.75.mlp.gate",
      "model.language_model.layers.76.mlp.gate",
      "model.language_model.layers.77.mlp.gate",
      "model.language_model.layers.78.mlp.gate",
      "model.language_model.layers.79.mlp.gate",
      "model.language_model.layers.80.mlp.gate",
      "model.language_model.layers.81.mlp.gate",
      "model.language_model.layers.82.mlp.gate",
      "model.language_model.layers.83.mlp.gate",
      "model.language_model.layers.84.mlp.gate",
      "model.language_model.layers.85.mlp.gate",
      "model.language_model.layers.86.mlp.gate",
      "model.language_model.layers.87.mlp.gate",
      "model.language_model.layers.88.mlp.gate",
      "model.language_model.layers.89.mlp.gate",
      "model.language_model.layers.90.mlp.gate",
      "model.language_model.layers.91.mlp.gate",
      "model.language_model.layers.92.mlp.gate",
      "model.language_model.layers.93.mlp.gate",
      "lm_head"
    ],
    "export": {
      "kv_cache_group": [],
      "min_kv_scale": 0.0,
      "pack_method": "reorder",
      "weight_format": "real_quantized",
      "weight_merge_groups": null
    },
    "global_quant_config": {
      "bias": null,
      "input_tensors": {
        "ch_axis": 1,
        "dtype": "fp8_e4m3",
        "group_size": null,
        "is_dynamic": true,
        "is_scale_quant": false,
        "mx_element_dtype": null,
        "observer_cls": "PerChannelMinMaxObserver",
        "qscheme": "per_channel",
        "round_method": null,
        "scale_calculation_mode": null,
        "scale_format": null,
        "scale_type": null,
        "symmetric": null
      },
      "output_tensors": null,
      "target_device": null,
      "weight": {
        "ch_axis": 0,
        "dtype": "fp8_e4m3",
        "group_size": null,
        "is_dynamic": false,
        "is_scale_quant": false,
        "mx_element_dtype": null,
        "observer_cls": "PerChannelMinMaxObserver",
        "qscheme": "per_channel",
        "round_method": null,
        "scale_calculation_mode": null,
        "scale_format": null,
        "scale_type": null,
        "symmetric": null
      }
    },
    "kv_cache_quant_config": {},
    "layer_quant_config": {},
    "layer_type_quant_config": {},
    "quant_method": "quark",
    "quant_mode": "eager_mode",
    "softmax_quant_spec": null,
    "version": "0.10"
  },
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "bos_token_id": 151643,
    "decoder_sparse_step": 1,
    "dtype": "float32",
    "eos_token_id": 151645,
    "head_dim": 128,
    "hidden_act": "silu",
    "hidden_size": 4096,
    "initializer_range": 0.02,
    "intermediate_size": 12288,
    "max_position_embeddings": 262144,
    "mlp_only_layers": [],
    "model_type": "qwen3_vl_moe_text",
    "moe_intermediate_size": 1536,
    "norm_topk_prob": true,
    "num_attention_heads": 64,
    "num_experts": 128,
    "num_experts_per_tok": 8,
    "num_hidden_layers": 94,
    "num_key_value_heads": 4,
    "rms_norm_eps": 1e-06,
    "rope_scaling": {
      "mrope_interleaved": true,
      "mrope_section": [
        24,
        20,
        20
      ],
      "rope_type": "default"
    },
    "rope_theta": 5000000,
    "router_aux_loss_coef": 0.001,
    "use_cache": true,
    "vocab_size": 151936
  },
  "tie_word_embeddings": false,
  "transformers_version": "4.57.1",
  "video_token_id": 151656,
  "vision_config": {
    "deepstack_visual_indexes": [
      8,
      16,
      24
    ],
    "depth": 27,
    "dtype": "float32",
    "hidden_act": "gelu_pytorch_tanh",
    "hidden_size": 1152,
    "in_channels": 3,
    "initializer_range": 0.02,
    "intermediate_size": 4304,
    "model_type": "qwen3_vl_moe",
    "num_heads": 16,
    "num_position_embeddings": 2304,
    "out_hidden_size": 4096,
    "patch_size": 16,
    "spatial_merge_size": 2,
    "temporal_patch_size": 2
  },
  "vision_end_token_id": 151653,
  "vision_start_token_id": 151652
}