w2v-bert-odia_v1 / vocab.json
cdactvm's picture
Upload tokenizer
cd3f61d verified
raw
history blame
1.14 kB
{
"0": 1,
"1": 2,
"2": 3,
"3": 4,
"5": 5,
"7": 6,
"8": 7,
"9": 8,
"[PAD]": 88,
"[UNK]": 87,
"|": 0,
"।": 9,
"ଁ": 10,
"ଂ": 11,
"ଃ": 12,
"ଅ": 13,
"ଆ": 14,
"ଇ": 15,
"ଈ": 16,
"ଉ": 17,
"ଊ": 18,
"ଋ": 19,
"ଏ": 20,
"ଐ": 21,
"ଓ": 22,
"ଔ": 23,
"କ": 24,
"ଖ": 25,
"ଗ": 26,
"ଘ": 27,
"ଙ": 28,
"ଚ": 29,
"ଛ": 30,
"ଜ": 31,
"ଝ": 32,
"ଞ": 33,
"ଟ": 34,
"ଠ": 35,
"ଡ": 36,
"ଢ": 37,
"ଣ": 38,
"ତ": 39,
"ଥ": 40,
"ଦ": 41,
"ଧ": 42,
"ନ": 43,
"ପ": 44,
"ଫ": 45,
"ବ": 46,
"ଭ": 47,
"ମ": 48,
"ଯ": 49,
"ର": 50,
"ଲ": 51,
"ଳ": 52,
"ଵ": 53,
"ଶ": 54,
"ଷ": 55,
"ସ": 56,
"ହ": 57,
"଼": 58,
"ା": 59,
"ି": 60,
"ୀ": 61,
"ୁ": 62,
"ୂ": 63,
"ୃ": 64,
"େ": 65,
"ୈ": 66,
"ୋ": 67,
"ୌ": 68,
"୍": 69,
"ଡ଼": 70,
"ଢ଼": 71,
"ୟ": 72,
"୦": 73,
"୧": 74,
"୨": 75,
"୩": 76,
"୪": 77,
"୫": 78,
"୬": 79,
"୭": 80,
"୮": 81,
"୯": 82,
"ୱ": 83,
"‌": 84,
"–": 85,
"’": 86
}