tokenizer.json

{
  "type": "word",
  "unk_token": "[UNK]",
  "pad_token": "[PAD]",
  "vocab_size": 50000,
  "vocab_file": "vocab.txt"
}