tokenizer_config.json

{
  "bos_token": "[CLS]",
  "clean_up_tokenization_spaces": true,
  "cls_token": "[CLS]",
  "do_lower_case": true,
  "eos_token": "[SEP]",
  "keep_accents": false,
  "mask_token": {
    "__type": "AddedToken",
    "content": "[MASK]",
    "lstrip": true,
    "normalized": false,
    "rstrip": false,
    "single_word": false
  },
  "model_max_length": 512,
  "pad_token": "<pad>",
  "remove_space": true,
  "sep_token": "[SEP]",
  "sp_model_kwargs": {},
  "tokenizer_class": "AlbertTokenizer",
  "unk_token": "<unk>"
}