config.json

{
  "_name_or_path": "model_weights/pvit",
  "architectures": [
    "LlavaLlamaForCausalLM"
  ],
  "bos_token_id": 0,
  "eos_token_id": 1,
  "freeze_mm_mlp_adapter": false,
  "hidden_act": "silu",
  "hidden_size": 4096,
  "initializer_range": 0.02,
  "intermediate_size": 11008,
  "max_position_embeddings": 2048,
  "max_sequence_length": 2048,
  "mm_hidden_size": 1024,
  "mm_use_bbox_fc": true,
  "mm_use_im_start_end": true,
  "mm_use_prompt_encoder": false,
  "mm_use_prompt_token": false,
  "mm_use_region_clip": "model_weights/regionclip_pretrained-cc_rn50x4.pth",
  "mm_vision_select_layer": -2,
  "mm_vision_tower": "openai/clip-vit-large-patch14",
  "model_type": "llava",
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "pad_token_id": -1,
  "region_end_token": 32004,
  "region_start_token": 32003,
  "rms_norm_eps": 1e-06,
  "tie_word_embeddings": false,
  "torch_dtype": "float16",
  "transformers_version": "4.28.0.dev0",
  "tune_mm_mlp_adapter": false,
  "tune_mm_prompt_only": false,
  "use_cache": false,
  "use_mm_proj": true,
  "vocab_size": 32005
}