{
  "name_or_path": "new_corpus_bert_reference",
  "special_tokens_map_file": "new_corpus_bert_reference/special_tokens_map.json",
  "tokenizer_class": "PreTrainedTokenizerFast"
}