tokenizer_config.json

{
  "clean_up_tokenization_spaces": true,
  "cls_token": "[CLS]",
  "do_lower_case": false,
  "encoding": "utf8",
  "mask_token": "[MASK]",
  "model_max_length": 514,
  "pad_token": "[PAD]",
  "sep_token": "[SEP]",
  "sp_model_kwargs": {},
  "tokenizer_class": "ErnieMTokenizer",
  "unk_token": "[UNK]"
}