Upload tokenizer
13d0731
verified
|
{ |
|
"\t": 0, |
|
"\n": 1, |
|
"\r": 2, |
|
"#": 4, |
|
"&": 5, |
|
"(": 6, |
|
")": 7, |
|
"+": 8, |
|
"/": 9, |
|
"0": 10, |
|
"1": 11, |
|
"2": 12, |
|
"3": 13, |
|
"4": 14, |
|
"5": 15, |
|
"6": 16, |
|
"7": 17, |
|
"8": 18, |
|
"9": 19, |
|
"[": 20, |
|
"[PAD]": 122, |
|
"[UNK]": 121, |
|
"\\": 21, |
|
"]": 22, |
|
"_": 23, |
|
"{": 24, |
|
"|": 3, |
|
"}": 25, |
|
"~": 26, |
|
" ": 27, |
|
"¥": 28, |
|
"©": 29, |
|
"¬": 30, |
|
"": 31, |
|
"·": 32, |
|
"ë": 33, |
|
"ŋ": 34, |
|
"ɕ": 35, |
|
"ə": 36, |
|
"ɵ": 37, |
|
"ʨ": 38, |
|
"̀": 39, |
|
"́": 40, |
|
"̃": 41, |
|
"̆": 42, |
|
"̇": 43, |
|
"̈": 44, |
|
"̜": 45, |
|
"̟": 46, |
|
"̥": 47, |
|
"̩": 48, |
|
"̰": 49, |
|
"̶": 50, |
|
"̹": 51, |
|
"υ": 52, |
|
"а": 53, |
|
"б": 54, |
|
"в": 55, |
|
"г": 56, |
|
"д": 57, |
|
"е": 58, |
|
"ж": 59, |
|
"з": 60, |
|
"и": 61, |
|
"й": 62, |
|
"к": 63, |
|
"л": 64, |
|
"м": 65, |
|
"н": 66, |
|
"о": 67, |
|
"п": 68, |
|
"р": 69, |
|
"с": 70, |
|
"т": 71, |
|
"у": 72, |
|
"ф": 73, |
|
"х": 74, |
|
"ц": 75, |
|
"ч": 76, |
|
"ш": 77, |
|
"щ": 78, |
|
"ъ": 79, |
|
"ы": 80, |
|
"ь": 81, |
|
"э": 82, |
|
"ю": 83, |
|
"я": 84, |
|
"ё": 85, |
|
"і": 86, |
|
"ї": 87, |
|
"ў": 88, |
|
"ѳ": 89, |
|
"ғ": 90, |
|
"қ": 91, |
|
"ң": 92, |
|
"ү": 93, |
|
"ұ": 94, |
|
"һ": 95, |
|
"ӊ": 96, |
|
"ӏ": 97, |
|
"ә": 98, |
|
"ө": 99, |
|
"": 100, |
|
"": 101, |
|
"‐": 102, |
|
"‒": 103, |
|
"–": 104, |
|
"—": 105, |
|
"―": 106, |
|
"’": 107, |
|
"‚": 108, |
|
"„": 109, |
|
"…": 110, |
|
"№": 111, |
|
"−": 112, |
|
"─": 113, |
|
"●": 114, |
|
"☺": 115, |
|
"♥": 116, |
|
"⠀": 117, |
|
"": 118, |
|
"": 119, |
|
"-": 120 |
|
} |
|
|