tokenizer_config.json

{
  "tokenizer_class": "SentencePieceTokenizerAdapter",
  "source_tokenizer_dir": "/home/nico/Code/PechaBridge/ext/BoSentencePiece",
  "model_max_length": 1000000000000000019884624838656,
  "unk_token": "<unk>",
  "pad_token": "<pad>",
  "bos_token": "<s>",
  "eos_token": "</s>",
  "cls_token": "[CLS]",
  "sep_token": "[SEP]",
  "mask_token": "[MASK]",
  "additional_special_tokens": [
    "<NL>",
    "<s_ocr>",
    "</s_ocr>",
    "<s_cls1>"
  ]
}