jhauret's picture
Upload tokenizer
9c8d017 verified
raw
history blame
410 Bytes
{
"<pad>": 35,
"<unk>": 34,
"a": 1,
"b": 2,
"d": 3,
"e": 4,
"f": 5,
"i": 6,
"j": 7,
"k": 8,
"l": 9,
"m": 10,
"n": 11,
"o": 12,
"p": 13,
"s": 14,
"t": 15,
"u": 16,
"v": 17,
"w": 18,
"y": 19,
"z": 20,
"|": 0,
"ø": 21,
"ŋ": 22,
"œ": 23,
"ɑ": 24,
"ɔ": 25,
"ə": 26,
"ɛ": 27,
"ɡ": 28,
"ɲ": 29,
"ʁ": 30,
"ʃ": 31,
"ʒ": 32,
"̃": 33
}