bert-base-japanese-v3-jsts / tokenizer_config.json
ryo0634's picture
Upload tokenizer
d1e7f3d
raw
history blame contribute delete
606 Bytes
{
"cls_token": "[CLS]",
"do_lower_case": false,
"do_subword_tokenize": true,
"do_word_tokenize": true,
"jumanpp_kwargs": null,
"mask_token": "[MASK]",
"mecab_kwargs": {
"mecab_dic": "unidic_lite"
},
"model_max_length": 512,
"name_or_path": "results_jsts/bert-base-japanese-v3/lr_2e-05_epochs_3_seed_42",
"never_split": null,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"special_tokens_map_file": null,
"subword_tokenizer_type": "wordpiece",
"sudachi_kwargs": null,
"tokenizer_class": "BertJapaneseTokenizer",
"unk_token": "[UNK]",
"word_tokenizer_type": "mecab"
}