config.json

{
  "architectures": [
    "DeepQwenVLForCausalLM"
  ],
  "auto_map": {
    "AutoConfig": "model.DeepQwenVLConfig",
    "AutoModel": "model.DeepQwenVLForCausalLM",
    "AutoModelForCausalLM": "model.DeepQwenVLForCausalLM"
  },
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "deepseek_vision_hidden_size": 2048,
  "dtype": "float32",
  "eos_token_id": 151645,
  "hidden_act": "silu",
  "hidden_size": 1536,
  "image_newline_dim": 1536,
  "image_token_id": 151655,
  "initializer_range": 0.02,
  "intermediate_size": 8960,
  "max_position_embeddings": 32768,
  "model_type": "deepqwen_vl",
  "num_attention_heads": 12,
  "num_hidden_layers": 28,
  "num_key_value_heads": 2,
  "pad_token_id": 151643,
  "projector_hidden_dim": 1280,
  "projector_input_dim": 2048,
  "projector_output_dim": 1536,
  "projector_type": "mlp",
  "rms_norm_eps": 1e-06,
  "rope_scaling": {
    "mrope_section": [
      16,
      24,
      24
    ],
    "rope_type": "default",
    "type": "default"
  },
  "rope_theta": 1000000.0,
  "transformers_version": "4.57.3",
  "use_cache": true,
  "video_token_id": 151656,
  "view_separator_dim": 1536,
  "vision_end_token_id": 151653,
  "vision_start_token_id": 151652,
  "vision_token_id": 151654,
  "vocab_size": 151936
}