tokenizer_config.json

{
  "tokenizer_class": "PreTrainedTokenizerFast",
  "model_type": "BPE",
  "vocab_size": 50000,
  "language": "kn",
  "special_tokens": {
    "pad_token": "[PAD]",
    "unk_token": "[UNK]",
    "cls_token": "[CLS]",
    "sep_token": "[SEP]",
    "mask_token": "[MASK]"
  }
}