{
  "architectures": [
    "VisionEncoderDecoderModel"
  ],
  "decoder": {
    "_attn_implementation_autoset": true,
    "_name_or_path": "vinai/phobert-base",
    "add_cross_attention": true,
    "architectures": [
      "RobertaForMaskedLM"
    ],
    "attention_probs_dropout_prob": 0.1,
    "classifier_dropout": null,
    "gradient_checkpointing": false,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.1,
    "hidden_size": 768,
    "initializer_range": 0.02,
    "intermediate_size": 3072,
    "is_decoder": true,
    "layer_norm_eps": 1e-05,
    "max_position_embeddings": 258,
    "model_type": "roberta",
    "num_attention_heads": 12,
    "num_hidden_layers": 12,
    "position_embedding_type": "absolute",
    "tokenizer_class": "PhobertTokenizer",
    "torch_dtype": "float32",
    "type_vocab_size": 1,
    "use_cache": true,
    "vocab_size": 64001
  },
  "decoder_start_token_id": 0,
  "early_stopping": null,
  "encoder": {
    "attention_probs_dropout_prob": 0.0,
    "encoder_stride": 16,
    "hidden_act": "gelu",
    "hidden_dropout_prob": 0.0,
    "hidden_size": 1024,
    "image_size": 384,
    "initializer_range": 0.02,
    "intermediate_size": 4096,
    "layer_norm_eps": 1e-12,
    "model_type": "vit",
    "num_attention_heads": 16,
    "num_channels": 3,
    "num_hidden_layers": 24,
    "patch_size": 16,
    "pooler_act": "tanh",
    "pooler_output_size": 1024,
    "qkv_bias": false,
    "torch_dtype": "float32"
  },
  "eos_token_id": 2,
  "is_encoder_decoder": true,
  "length_penalty": null,
  "max_length": null,
  "model_type": "vision-encoder-decoder",
  "no_repeat_ngram_size": null,
  "num_beams": null,
  "pad_token_id": 1,
  "tie_word_embeddings": false,
  "torch_dtype": "float32",
  "transformers_version": "4.51.3",
  "vocab_size": 64001
}