{
  "_name_or_path": "google/vit-base-patch16-224",
  "architectures": [
    "ViTForImageClassification"
  ],
  "attention_probs_dropout_prob": 0.0,
  "encoder_stride": 16,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.0,
  "hidden_size": 768,
  "id2label": {
    "0": "Beige",
    "1": "Black",
    "2": "Black and White",
    "3": "Blue",
    "4": "Brown",
    "5": "Deep Beige",
    "6": "Deep Brown",
    "7": "Deep Gray",
    "8": "Gray",
    "9": "Green",
    "10": "Ivory",
    "11": "Light Beige",
    "12": "Light Brown",
    "13": "Light Gray",
    "14": "Orange",
    "15": "Pink",
    "16": "Purple",
    "17": "Red",
    "18": "White",
    "19": "Yellow"
  },
  "image_size": 224,
  "initializer_range": 0.02,
  "intermediate_size": 3072,
  "label2id": {
    "Beige": 0,
    "Black": 1,
    "Black and White": 2,
    "Blue": 3,
    "Brown": 4,
    "Deep Beige": 5,
    "Deep Brown": 6,
    "Deep Gray": 7,
    "Gray": 8,
    "Green": 9,
    "Ivory": 10,
    "Light Beige": 11,
    "Light Brown": 12,
    "Light Gray": 13,
    "Orange": 14,
    "Pink": 15,
    "Purple": 16,
    "Red": 17,
    "White": 18,
    "Yellow": 19
  },
  "layer_norm_eps": 1e-12,
  "model_type": "vit",
  "num_attention_heads": 12,
  "num_channels": 3,
  "num_hidden_layers": 12,
  "patch_size": 16,
  "problem_type": "single_label_classification",
  "qkv_bias": true,
  "torch_dtype": "float32",
  "transformers_version": "4.37.0"
}