tokenizer_config.json

{
  "add_prefix_space": true,
  "backend": "tokenizers",
  "bos_token": "<s>",
  "cls_token": "[CLS]",
  "eos_token": "</s>",
  "mask_token": "[MASK]",
  "model_max_length": 512,
  "pad_token": "[PAD]",
  "sep_token": "[SEP]",
  "sp_model_kwargs": {
    "enable_sampling": false
  },
  "tokenizer_class": "XLMRobertaTokenizer",
  "unk_token": "[UNK]",
  "language": "bal",
  "keep_accents": true,
  "strip_accents": false,
  "do_lower_case": false
}