config.json

{
  "architectures": [
    "PerceptionLMForConditionalGeneration"
  ],
  "dtype": "float16",
  "image_token_id": 128002,
  "model_type": "perception_lm",
  "projector_pooling_ratio": 2,
  "text_config": {
    "attention_bias": false,
    "attention_dropout": 0.0,
    "bos_token_id": 128000,
    "dtype": "float16",
    "eos_token_id": [
      128001,
      128009
    ],
    "head_dim": 128,
    "hidden_act": "silu",
    "hidden_size": 4096,
    "initializer_range": 0.02,
    "intermediate_size": 14336,
    "max_position_embeddings": 11520,
    "mlp_bias": false,
    "model_type": "llama",
    "num_attention_heads": 32,
    "num_hidden_layers": 32,
    "num_key_value_heads": 8,
    "pretraining_tp": 1,
    "rms_norm_eps": 1e-05,
    "rope_scaling": null,
    "rope_theta": 500000.0,
    "use_cache": true,
    "vocab_size": 128256
  },
  "transformers_version": "4.57.1",
  "video_token_id": 128003,
  "vision_config": {
    "architecture": "vit_pe_core_gigantic_patch14_448",
    "do_pooling": true,
    "dtype": "float16",
    "global_pool": "map",
    "initializer_range": 0.02,
    "label_names": [
      "LABEL_0",
      "LABEL_1"
    ],
    "model_args": {
      "depth": 47,
      "embed_dim": 1536,
      "global_pool": "",
      "img_size": [
        448,
        448
      ],
      "init_values": 0.1,
      "ref_feat_shape": [
        32,
        32
      ],
      "use_post_transformer_norm": false
    },
    "model_type": "timm_wrapper",
    "num_classes": 2,
    "num_features": 1536,
    "pretrained_cfg": {
      "classifier": "head",
      "crop_mode": "center",
      "crop_pct": 1.0,
      "custom_load": false,
      "first_conv": "patch_embed.proj",
      "fixed_input_size": true,
      "input_size": [
        3,
        448,
        448
      ],
      "interpolation": "bicubic",
      "license": "custom",
      "mean": [
        0.5,
        0.5,
        0.5
      ],
      "pool_size": null,
      "std": [
        0.5,
        0.5,
        0.5
      ],
      "tag": "fb"
    }
  },
  "vision_use_cls_token": false,
  "int4_exclude_patterns": [],
  "auto_map": {
    "AutoModelForImageTextToText": "modeling_perception_lm_int4.PerceptionLMForConditionalGenerationInt4"
  }
}