tokenizer_config.json

{
  "add_prefix_space": false,
  "clean_up_tokenization_spaces": true,
  "cls_token": "<|beginoftext|>",
  "do_basic_tokenize": true,
  "do_lower_case": true,
  "mask_token": "[MASK]",
  "model_max_length": 1024,
  "never_split": null,
  "pad_token": "[PAD]",
  "sep_token": "<|endoftext|>",
  "strip_accents": null,
  "tokenize_chinese_chars": true,
  "tokenizer_class": "BertTokenizer",
  "unk_token": "[UNK]"
}