mms-SADA / vocab.json
herwoww's picture
Upload tokenizer
bdf0390 verified
raw
history blame
517 Bytes
{
"ara": {
"[PAD]": 35,
"[UNK]": 34,
"|": 0,
"ء": 1,
"ؤ": 2,
"ئ": 3,
"ا": 4,
"ب": 5,
"ة": 6,
"ت": 7,
"ث": 8,
"ج": 9,
"ح": 10,
"خ": 11,
"د": 12,
"ذ": 13,
"ر": 14,
"ز": 15,
"س": 16,
"ش": 17,
"ص": 18,
"ض": 19,
"ط": 20,
"ظ": 21,
"ع": 22,
"غ": 23,
"ف": 24,
"ق": 25,
"ك": 26,
"ل": 27,
"م": 28,
"ن": 29,
"ه": 30,
"و": 31,
"ى": 32,
"ي": 33
}
}