{
  "architectures": [
    "EmovaQwen2ForCausalLM"
  ],
  "attention_dropout": 0.0,
  "bos_token_id": 151643,
  "eos_token_id": 151645,
  "hidden_act": "silu",
  "hidden_size": 2048,
  "image_aspect_ratio": "native_anyres",
  "initializer_range": 0.02,
  "intermediate_size": 11008,
  "max_position_embeddings": 32768,
  "max_window_layers": 70,
  "mm_hidden_size": 3584,
  "mm_patch_merge_type": "flat",
  "mm_projector": {
    "mlp_depth": 2,
    "trainable": true,
    "type": "MLPProjector"
  },
  "mm_projector_lr": 2e-05,
  "mm_use_im_patch_token": false,
  "mm_use_im_start_end": false,
  "mm_vision_tower": {
    "max_pixels": 3211264,
    "pretrained_model_name_or_path": "Emova-ollm/qwen2vit600m",
    "trainable": true,
    "type": "Qwen2VisionTower",
    "unfreeze_mm_vision_tower": true
  },
  "model_type": "emova_qwen2",
  "num_attention_heads": 16,
  "num_hidden_layers": 36,
  "num_key_value_heads": 2,
  "rms_norm_eps": 1e-06,
  "rope_theta": 1000000.0,
  "sliding_window": null,
  "tie_word_embeddings": true,
  "tokenizer_model_max_length": 8192,
  "tokenizer_padding_side": "left",
  "torch_dtype": "bfloat16",
  "transformers_version": "4.44.2",
  "use_cache": true,
  "use_mm_proj": true,
  "use_sliding_window": false,
  "vocab_size": 155776
}