bart-caser / tokenizer_config.json
XvKuoMing's picture
End of training
eba73e3 verified
{
"add_prefix_space": true,
"added_tokens_decoder": {
"0": {
"content": "<s>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1": {
"content": "<pad>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"2": {
"content": "</s>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"3": {
"content": "<unk>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24208": {
"content": "ar_AR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24209": {
"content": "cs_CZ",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24210": {
"content": "de_DE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24211": {
"content": "en_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24212": {
"content": "es_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24213": {
"content": "et_EE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24214": {
"content": "fi_FI",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24215": {
"content": "fr_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24216": {
"content": "gu_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24217": {
"content": "hi_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24218": {
"content": "it_IT",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24219": {
"content": "ja_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24220": {
"content": "kk_KZ",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24221": {
"content": "ko_KR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24222": {
"content": "lt_LT",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24223": {
"content": "lv_LV",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24224": {
"content": "my_MM",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24225": {
"content": "ne_NP",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24226": {
"content": "nl_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24227": {
"content": "ro_RO",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24228": {
"content": "ru_RU",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24229": {
"content": "si_LK",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24230": {
"content": "tr_TR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24231": {
"content": "vi_VN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24232": {
"content": "zh_CN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24233": {
"content": "af_ZA",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24234": {
"content": "az_AZ",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24235": {
"content": "bn_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24236": {
"content": "fa_IR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24237": {
"content": "he_IL",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24238": {
"content": "hr_HR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24239": {
"content": "id_ID",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24240": {
"content": "ka_GE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24241": {
"content": "km_KH",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24242": {
"content": "mk_MK",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24243": {
"content": "ml_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24244": {
"content": "mn_MN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24245": {
"content": "mr_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24246": {
"content": "pl_PL",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24247": {
"content": "ps_AF",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24248": {
"content": "pt_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24249": {
"content": "sv_SE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24250": {
"content": "sw_KE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24251": {
"content": "ta_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24252": {
"content": "te_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24253": {
"content": "th_TH",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24254": {
"content": "tl_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24255": {
"content": "uk_UA",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24256": {
"content": "ur_PK",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24257": {
"content": "xh_ZA",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24258": {
"content": "gl_ES",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24259": {
"content": "sl_SI",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"24260": {
"content": "<mask>",
"lstrip": true,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": true
}
},
"additional_special_tokens": [
"ar_AR",
"cs_CZ",
"de_DE",
"en_XX",
"es_XX",
"et_EE",
"fi_FI",
"fr_XX",
"gu_IN",
"hi_IN",
"it_IT",
"ja_XX",
"kk_KZ",
"ko_KR",
"lt_LT",
"lv_LV",
"my_MM",
"ne_NP",
"nl_XX",
"ro_RO",
"ru_RU",
"si_LK",
"tr_TR",
"vi_VN",
"zh_CN",
"af_ZA",
"az_AZ",
"bn_IN",
"fa_IR",
"he_IL",
"hr_HR",
"id_ID",
"ka_GE",
"km_KH",
"mk_MK",
"ml_IN",
"mn_MN",
"mr_IN",
"pl_PL",
"ps_AF",
"pt_XX",
"sv_SE",
"sw_KE",
"ta_IN",
"te_IN",
"th_TH",
"tl_XX",
"uk_UA",
"ur_PK",
"xh_ZA",
"gl_ES",
"sl_SI"
],
"bos_token": "<s>",
"clean_up_tokenization_spaces": true,
"cls_token": "<s>",
"device_map": "auto",
"eos_token": "</s>",
"mask_token": "<mask>",
"model_max_length": 1024,
"pad_token": "<pad>",
"sep_token": "</s>",
"sp_model_kwargs": {},
"src_lang": "en_XX",
"tgt_lang": null,
"tokenizer_class": "MBart50Tokenizer",
"unk_token": "<unk>"
}