tokenizer_config.json

{
  "add_prefix_space": false,
  "backend": "tokenizers",
  "bos_token": "<s>",
  "cls_token": "<s>",
  "do_lower_case": false,
  "eos_token": "</s>",
  "errors": "replace",
  "is_local": false,
  "mask_token": "<mask>",
  "model_max_length": 512,
  "pad_token": "<pad>",
  "sep_token": "</s>",
  "strip_accents": null,
  "tokenize_chinese_chars": true,
  "tokenizer_class": "BertTokenizer",
  "trim_offsets": true,
  "unk_token": "<unk>"
}