config.json

{
  "_name_or_path": "kawn_idefics3_siglib14_384",
  "architectures": [
    "KawnIdefics3ForConditionalGeneration"
  ],
  "ignore_index": -100,
  "image_grid_pinpoints": [
    [
      384,
      384
    ],
    [
      384,
      768
    ],
    [
      384,
      1152
    ],
    [
      384,
      1536
    ],
    [
      1536,
      384
    ],
    [
      1152,
      384
    ],
    [
      768,
      384
    ],
    [
      768,
      768
    ],
    [
      768,
      1152
    ],
    [
      768,
      1536
    ],
    [
      1536,
      768
    ],
    [
      1152,
      768
    ],
    [
      1152,
      1536
    ],
    [
      1536,
      1152
    ],
    [
      1152,
      1152
    ]
  ],
  "image_seq_length": 182,
  "image_token_id": 255030,
  "model_type": "kawn_idefics3",
  "scale_factor": 2,
  "text_config": {
    "_attn_implementation_autoset": true,
    "_name_or_path": "CohereForAI/aya-expanse-8b",
    "architectures": [
      "CohereForCausalLM"
    ],
    "hidden_size": 4096,
    "intermediate_size": 14336,
    "logit_scale": 0.125,
    "model_type": "cohere",
    "num_attention_heads": 32,
    "num_hidden_layers": 32,
    "num_key_value_heads": 8,
    "torch_dtype": "bfloat16"
  },
  "torch_dtype": "bfloat16",
  "transformers_version": "4.47.1",
  "use_cache": true,
  "vision_config": {
    "_attn_implementation_autoset": true,
    "hidden_size": 1152,
    "image_size": 384,
    "intermediate_size": 4304,
    "model_type": "siglip_vision_model",
    "num_attention_heads": 16,
    "num_hidden_layers": 27,
    "patch_size": 14
  },
  "vision_feature_layer": null
}