{ "[PAD]": 114, "[UNK]": 113, "_": 1, "`": 2, "a": 3, "b": 4, "c": 5, "d": 6, "e": 7, "f": 8, "g": 9, "h": 10, "i": 11, "j": 12, "k": 13, "l": 14, "m": 15, "n": 16, "o": 17, "p": 18, "q": 19, "r": 20, "s": 21, "t": 22, "u": 23, "v": 24, "w": 25, "x": 26, "y": 27, "z": 28, "|": 0, "¥": 29, "°": 30, "½": 31, "¾": 32, "é": 33, "í": 34, "،": 35, "؛": 36, "؟": 37, "ء": 38, "آ": 39, "أ": 40, "ؤ": 41, "ئ": 42, "ا": 43, "ب": 44, "ت": 45, "ث": 46, "ج": 47, "ح": 48, "خ": 49, "د": 50, "ذ": 51, "ر": 52, "ز": 53, "س": 54, "ش": 55, "ص": 56, "ض": 57, "ط": 58, "ظ": 59, "ع": 60, "غ": 61, "ـ": 62, "ف": 63, "ق": 64, "ك": 65, "ل": 66, "م": 67, "ن": 68, "ه": 69, "و": 70, "ى": 71, "ي": 72, "ً": 73, "ٌ": 74, "َ": 75, "ُ": 76, "ّ": 77, "٪": 78, "ټ": 79, "پ": 80, "ځ": 81, "څ": 82, "چ": 83, "ډ": 84, "ړ": 85, "ږ": 86, "ژ": 87, "ښ": 88, "ک": 89, "ګ": 90, "گ": 91, "ڼ": 92, "ھ": 93, "ی": 94, "ۍ": 95, "ې": 96, "ے": 97, "۔": 98, "۰": 99, "۱": 100, "۲": 101, "۳": 102, "۴": 103, "۵": 104, "۶": 105, "۷": 106, "۸": 107, "۹": 108, "‌": 109, "‍": 110, "–": 111, "—": 112 }