config.json

{
  "architectures": [
    "ImageCaptioningModel"
  ],
  "max_seq_len": 20,
  "torch_dtype": "float32",
  "transformers_version": "4.52.3",
  "vocab_size": 3478
}