tokenizer_config.json

{
  "clean_up_tokenization_spaces": true,
  "do_lower_case": false,
  "model_max_length": 512,
  "special_tokens": {
    "mask_token": "[MASK]",
    "pad_token": "[PAD]",
    "sep_token": "[SEP]"
  },
  "tokenize_chinese_chars": false,
  "tokenizer_class": "PreTrainedTokenizerFast",
  "trust_remote_code": true
}