{
  "do_lower_case": false,
  "subword_tokenizer_type": "wordpiece",
  "word_tokenizer_type": "mecab"
}