{"cls_token": "<s>", "unk_token": "<unk>", "pad_token": "<pad>", "mask_token": "<mask>", "sep_token": "</s>", "do_lowercase_and_remove_accent": false, "bos_token": "<s>", "additional_special_tokens": [], "model_max_length": 512, "do_lower_case": false, "lang2id": null, "id2lang": null, "tokenizer_file": null, "tokenizer_class": "HerbertTokenizer"}