malayalam_combined_ / vocab.json
akrishnan's picture
Upload tokenizer
ca49120 verified
raw
history blame
1.17 kB
{
"*": 1,
"0": 2,
"1": 3,
"2": 4,
"3": 5,
"4": 6,
"5": 7,
"6": 8,
"7": 9,
"8": 10,
"9": 11,
"=": 12,
"@": 13,
"[PAD]": 91,
"[UNK]": 90,
"|": 0,
"ം": 14,
"ഃ": 15,
"അ": 16,
"ആ": 17,
"ഇ": 18,
"ഈ": 19,
"ഉ": 20,
"ഊ": 21,
"ഋ": 22,
"എ": 23,
"ഏ": 24,
"ഐ": 25,
"ഒ": 26,
"ഓ": 27,
"ഔ": 28,
"ക": 29,
"ഖ": 30,
"ഗ": 31,
"ഘ": 32,
"ങ": 33,
"ച": 34,
"ഛ": 35,
"ജ": 36,
"ഝ": 37,
"ഞ": 38,
"ട": 39,
"ഠ": 40,
"ഡ": 41,
"ഢ": 42,
"ണ": 43,
"ത": 44,
"ഥ": 45,
"ദ": 46,
"ധ": 47,
"ന": 48,
"പ": 49,
"ഫ": 50,
"ബ": 51,
"ഭ": 52,
"മ": 53,
"യ": 54,
"ര": 55,
"റ": 56,
"ല": 57,
"ള": 58,
"ഴ": 59,
"വ": 60,
"ശ": 61,
"ഷ": 62,
"സ": 63,
"ഹ": 64,
"ാ": 65,
"ി": 66,
"ീ": 67,
"ു": 68,
"ൂ": 69,
"ൃ": 70,
"െ": 71,
"േ": 72,
"ൈ": 73,
"ൊ": 74,
"ോ": 75,
"ൌ": 76,
"്": 77,
"ൗ": 78,
"൦": 79,
"൨": 80,
"൪": 81,
"ൺ": 82,
"ൻ": 83,
"ർ": 84,
"ൽ": 85,
"ൾ": 86,
"‌": 87,
"‍": 88,
"’": 89
}