unirec-0.1b/config.json

{
  "activation_dropout": 0.0,
  "activation_function": "relu",
  "architectures": [
    "VLMOCRForConditionalGenerationNew"
  ],
  "attention_dropout": 0.1,
  "bos_token_id": 0,
  "d_model": 768,
  "decoder_attention_heads": 6,
  "decoder_ffn_dim": 3072,
  "decoder_layerdrop": 0.0,
  "decoder_layers": 6,
  "decoder_start_token_id": 0,
  "depths": [
    2,
    2,
    9,
    2
  ],
  "dims": [
    128,
    256,
    512,
    768
  ],
  "drop_path_rate": 0.1,
  "dropout": 0.1,
  "encoder_layerdrop": 0.0,
  "eos_token_id": 2,
  "init_std": 0.02,
  "is_encoder_decoder": true,
  "kernel_size": [
    3,
    3
  ],
  "max_position_embeddings": 3072,
  "mixer": [
    [
      "Conv",
      "Conv"
    ],
    [
      "Conv",
      "Conv"
    ],
    [
      "Conv",
      "Conv",
      "Conv",
      "Conv",
      "Conv",
      "FGlobal",
      "Global",
      "Global",
      "Global"
    ],
    [
      "Global",
      "Global"
    ]
  ],
  "mlp_ratio": 4,
  "model_type": "m2m_100",
  "num_heads": [
    4,
    8,
    8,
    12
  ],
  "pad_token_id": 1,
  "scale_embedding": true,
  "sub_k": [
    [
      2,
      2
    ],
    [
      2,
      2
    ],
    [
      2,
      2
    ],
    [
      2,
      2
    ]
  ],
  "torch_dtype": "float32",
  "transformers_version": "4.49.0",
  "use_cache": true,
  "vocab_size": 56371,
  "max_length": 1024,
  "label_smoothing": 0.1
}