w2v-bert-yoruba-CV16 / vocab.json
olanigan's picture
Upload tokenizer
65da849 verified
raw
history blame
522 Bytes
{
"[PAD]": 44,
"[UNK]": 43,
"a": 1,
"b": 2,
"d": 3,
"e": 4,
"f": 5,
"g": 6,
"h": 7,
"i": 8,
"j": 9,
"k": 10,
"l": 11,
"m": 12,
"n": 13,
"o": 14,
"p": 15,
"r": 16,
"s": 17,
"t": 18,
"u": 19,
"w": 20,
"y": 21,
"|": 0,
"à": 22,
"á": 23,
"è": 24,
"é": 25,
"ì": 26,
"í": 27,
"ò": 28,
"ó": 29,
"ù": 30,
"ú": 31,
"ń": 32,
"ǹ": 33,
"̀": 34,
"́": 35,
"̄": 36,
"̣": 37,
"ṣ": 38,
"ẹ": 39,
"ọ": 40,
"ụ": 41,
"’": 42
}