config.json

{
  "architectures": [
    "DonutSwinModel"
  ],
  "attention_probs_dropout_prob": 0.1,
  "decoder_start_token_id": 0,
  "early_stopping": true,
  "encoder_layer_norm_eps": 1e-05,
  "hidden_act": "gelu",
  "hidden_dropout_prob": 0.1,
  "hidden_size": 768,
  "id2label": {
    "0": "header",
    "1": "question",
    "2": "answer",
    "3": "footer"
  },
  "image_size": [1280, 960],
  "initializer_range": 0.02,
  "is_encoder_decoder": true,
  "label2id": {
    "answer": 2,
    "footer": 3,
    "header": 0,
    "question": 1
  },
  "max_length": 768,
  "model_type": "donut",
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "pad_token_id": 1,
  "torch_dtype": "float32",
  "transformers_version": "4.36.0",
  "window_size": 7
}