kkahadze's picture
Upload tokenizer
b9c40eb verified
raw
history blame contribute delete
527 Bytes
{
"[PAD]": 45,
"[UNK]": 44,
"a": 1,
"b": 2,
"c": 3,
"d": 4,
"e": 5,
"g": 6,
"h": 7,
"i": 8,
"j": 9,
"k": 10,
"l": 11,
"m": 12,
"n": 13,
"o": 14,
"p": 15,
"q": 16,
"r": 17,
"s": 18,
"t": 19,
"u": 20,
"v": 21,
"w": 22,
"x": 23,
"z": 24,
"|": 0,
"ă": 25,
"č": 26,
"ĕ": 27,
"ħ": 28,
"ŏ": 29,
"š": 30,
"ž": 31,
"ǧ": 32,
"ǰ": 33,
"ɣ": 34,
"ɬ": 35,
"ʒ": 36,
"ʔ": 37,
"ʕ": 38,
"ʡ": 39,
"ʲ": 40,
"ʷ": 41,
"ː": 42,
"ⁿ": 43
}