mms-MGB5 / vocab.json
herwoww's picture
Upload tokenizer
e00eef5 verified
raw
history blame
652 Bytes
{
"ara": {
"#": 1,
"*": 2,
".": 3,
"@": 4,
"[PAD]": 45,
"[UNK]": 44,
"_": 5,
"|": 0,
"؟": 6,
"ء": 7,
"آ": 8,
"أ": 9,
"ؤ": 10,
"إ": 11,
"ئ": 12,
"ا": 13,
"ب": 14,
"ة": 15,
"ت": 16,
"ث": 17,
"ج": 18,
"ح": 19,
"خ": 20,
"د": 21,
"ذ": 22,
"ر": 23,
"ز": 24,
"س": 25,
"ش": 26,
"ص": 27,
"ض": 28,
"ط": 29,
"ظ": 30,
"ع": 31,
"غ": 32,
"ف": 33,
"ق": 34,
"ك": 35,
"ل": 36,
"م": 37,
"ن": 38,
"ه": 39,
"و": 40,
"ى": 41,
"ي": 42,
"ّ": 43
}
}