mBART_cc25_finetune_en-vi_translation / tokenizer_config.json
Tohrumi's picture
Training in progress, step 500
1ba8f4a verified
{
"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
},
"250001": {
"content": "ar_AR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250002": {
"content": "cs_CZ",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250003": {
"content": "de_DE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250004": {
"content": "en_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250005": {
"content": "es_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250006": {
"content": "et_EE",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250007": {
"content": "fi_FI",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250008": {
"content": "fr_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250009": {
"content": "gu_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250010": {
"content": "hi_IN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250011": {
"content": "it_IT",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250012": {
"content": "ja_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250013": {
"content": "kk_KZ",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250014": {
"content": "ko_KR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250015": {
"content": "lt_LT",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250016": {
"content": "lv_LV",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250017": {
"content": "my_MM",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250018": {
"content": "ne_NP",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250019": {
"content": "nl_XX",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250020": {
"content": "ro_RO",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250021": {
"content": "ru_RU",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250022": {
"content": "si_LK",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250023": {
"content": "tr_TR",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250024": {
"content": "vi_VN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250025": {
"content": "zh_CN",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"250026": {
"content": "<mask>",
"lstrip": true,
"normalized": false,
"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"
],
"bos_token": "<s>",
"clean_up_tokenization_spaces": true,
"cls_token": "<s>",
"eos_token": "</s>",
"mask_token": "<mask>",
"model_max_length": 1000000000000000019884624838656,
"pad_token": "<pad>",
"sep_token": "</s>",
"src_lang": "en_XX",
"tgt_lang": "vi_VN",
"tokenizer_class": "MBartTokenizer",
"unk_token": "<unk>"
}