{ "auto_map": { "AutoTokenizer": [ "tokenization_luke_bert_japanese.LukeBertJapaneseTokenizer", null ] }, "clean_up_tokenization_spaces": true, "cls_token": "[CLS]", "do_lower_case": false, "do_subword_tokenize": true, "do_word_tokenize": true, "entity_mask2_token": "[MASK2]", "entity_mask_token": "[MASK]", "entity_pad_token": "[PAD]", "entity_token_1": "", "entity_token_2": "", "entity_unk_token": "[UNK]", "jumanpp_kwargs": null, "mask_token": "[MASK]", "max_entity_length": 32, "max_mention_length": 30, "mecab_kwargs": { "mecab_dic": "unidic_lite" }, "model_max_length": 512, "never_split": null, "pad_token": "[PAD]", "sep_token": "[SEP]", "spm_file": null, "subword_tokenizer_type": "wordpiece", "sudachi_kwargs": null, "task": null, "tokenizer_class": "LukeBertJapaneseTokenizer", "unk_token": "[UNK]", "word_tokenizer_type": "mecab" }