mt5-small-ft-Chinese-Braille / tokenizer_config.json
Violet-yo's picture
Upload tokenizer
445e6dd verified
{
"add_prefix_space": true,
"added_tokens_decoder": {
"0": {
"content": "<pad>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"1": {
"content": "</s>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"2": {
"content": "<unk>",
"lstrip": false,
"normalized": false,
"rstrip": false,
"single_word": false,
"special": true
},
"244576": {
"content": "⠁",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"246357": {
"content": "⠙",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"246634": {
"content": "⠫",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"247570": {
"content": "⠃",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"249368": {
"content": "⠈",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250100": {
"content": "⠢",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250101": {
"content": "⠖",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250102": {
"content": "⠶",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250103": {
"content": "⠦",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250104": {
"content": "⠔",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250105": {
"content": "⠴",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250106": {
"content": "⠉",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250107": {
"content": "⠑",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250108": {
"content": "⠋",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250109": {
"content": "⠛",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250110": {
"content": "⠓",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250111": {
"content": "⠊",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250112": {
"content": "⠚",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250113": {
"content": "⠅",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250114": {
"content": "⠇",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250115": {
"content": "⠍",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250116": {
"content": "⠝",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250117": {
"content": "⠕",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250118": {
"content": "⠏",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250119": {
"content": "⠟",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250120": {
"content": "⠗",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250121": {
"content": "⠎",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250122": {
"content": "⠞",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250123": {
"content": "⠥",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250124": {
"content": "⠧",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250125": {
"content": "⠺",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250126": {
"content": "⠭",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250127": {
"content": "⠽",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250128": {
"content": "⠵",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250129": {
"content": "⠮",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250130": {
"content": "⠐",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250131": {
"content": "⠼",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250132": {
"content": "⠯",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250133": {
"content": "⠄",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250134": {
"content": "⠷",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250135": {
"content": "⠾",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250136": {
"content": "⠡",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250137": {
"content": "⠬",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250138": {
"content": "⠠",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250139": {
"content": "⠤",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250140": {
"content": "⠨",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250141": {
"content": "⠌",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250142": {
"content": "⠆",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250143": {
"content": "⠰",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250144": {
"content": "⠣",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250145": {
"content": "⠿",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250146": {
"content": "⠜",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250147": {
"content": "⠹",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250148": {
"content": "⠪",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250149": {
"content": "⠳",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250150": {
"content": "⠻",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250151": {
"content": "⠘",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250152": {
"content": "⠸",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250153": {
"content": "⠲",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250154": {
"content": "⠒",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
},
"250155": {
"content": "⠂",
"lstrip": false,
"normalized": true,
"rstrip": false,
"single_word": false,
"special": false
}
},
"additional_special_tokens": [],
"clean_up_tokenization_spaces": true,
"eos_token": "</s>",
"extra_ids": 0,
"legacy": true,
"model_max_length": 1000000000000000019884624838656,
"pad_token": "<pad>",
"sp_model_kwargs": {},
"tokenizer_class": "T5Tokenizer",
"unk_token": "<unk>"
}