bert-base-japanese-whole-word-masking/tokenizer_config.json

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