{
  "_name_or_path": "/aifs4su/zhuhan/multi_turn_alignment/GRPO-for-Llava-main/checkpoints/verl_safety/llava_next_grpo_1218_1631/global_step_150/actor/huggingface",
  "architectures": [
    "LlavaNextForConditionalGeneration"
  ],
  "bos_token_id": 1,
  "dtype": "float32",
  "eos_token_id": 2,
  "hidden_size": 4096,
  "ignore_index": -100,
  "image_grid_pinpoints": [
    [
      336,
      336
    ]
  ],
  "image_seq_length": 576,
  "image_token_index": 32000,
  "model_type": "llava_next",
  "multimodal_projector_bias": true,
  "pad_token_id": 0,
  "projector_hidden_act": "gelu",
  "text_config": {
    "_name_or_path": "lmsys/vicuna-7b-v1.5",
    "architectures": [
      "LlamaForCausalLM"
    ],
    "dtype": "float32",
    "max_position_embeddings": 4096,
    "model_type": "llama",
    "pad_token_id": 0,
    "rms_norm_eps": 1e-05,
    "torch_dtype": "float32",
    "vocab_size": 32064
  },
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.49.0",
  "use_cache": true,
  "use_image_newline_parameter": true,
  "vision_config": {
    "dtype": "float32",
    "hidden_size": 1024,
    "image_size": 336,
    "intermediate_size": 4096,
    "model_type": "clip_vision_model",
    "num_attention_heads": 16,
    "num_hidden_layers": 24,
    "patch_size": 14,
    "projection_dim": 768,
    "torch_dtype": "float32",
    "vocab_size": 32000
  },
  "vision_feature_layer": -2,
  "vision_feature_select_strategy": "default",
  "vocab_size": 32064
}