Upload tokenizer
6cd0398
verified
|
{ |
|
"!": 1, |
|
"\"": 2, |
|
"&": 3, |
|
"'": 4, |
|
",": 5, |
|
"-": 6, |
|
".": 7, |
|
":": 8, |
|
"?": 9, |
|
"[PAD]": 111, |
|
"[UNK]": 110, |
|
"a": 10, |
|
"b": 11, |
|
"c": 12, |
|
"d": 13, |
|
"e": 14, |
|
"f": 15, |
|
"g": 16, |
|
"h": 17, |
|
"i": 18, |
|
"j": 19, |
|
"k": 20, |
|
"l": 21, |
|
"m": 22, |
|
"n": 23, |
|
"o": 24, |
|
"p": 25, |
|
"q": 26, |
|
"r": 27, |
|
"s": 28, |
|
"t": 29, |
|
"u": 30, |
|
"v": 31, |
|
"w": 32, |
|
"x": 33, |
|
"y": 34, |
|
"z": 35, |
|
"|": 0, |
|
"ँ": 37, |
|
"ं": 38, |
|
"ः": 39, |
|
"अ": 40, |
|
"आ": 41, |
|
"इ": 42, |
|
"ई": 43, |
|
"उ": 44, |
|
"ऊ": 45, |
|
"ऋ": 46, |
|
"ए": 47, |
|
"ऐ": 48, |
|
"ऑ": 49, |
|
"ओ": 50, |
|
"औ": 51, |
|
"क": 52, |
|
"ख": 53, |
|
"ग": 54, |
|
"घ": 55, |
|
"च": 56, |
|
"छ": 57, |
|
"ज": 58, |
|
"झ": 59, |
|
"ञ": 60, |
|
"ट": 61, |
|
"ठ": 62, |
|
"ड": 63, |
|
"ढ": 64, |
|
"ण": 65, |
|
"त": 66, |
|
"थ": 67, |
|
"द": 68, |
|
"ध": 69, |
|
"न": 70, |
|
"प": 71, |
|
"फ": 72, |
|
"ब": 73, |
|
"भ": 74, |
|
"म": 75, |
|
"य": 76, |
|
"र": 77, |
|
"ल": 78, |
|
"व": 79, |
|
"श": 80, |
|
"ष": 81, |
|
"स": 82, |
|
"ह": 83, |
|
"़": 84, |
|
"ा": 85, |
|
"ि": 86, |
|
"ी": 87, |
|
"ु": 88, |
|
"ू": 89, |
|
"ृ": 90, |
|
"ॅ": 91, |
|
"े": 92, |
|
"ै": 93, |
|
"ॉ": 94, |
|
"ो": 95, |
|
"ौ": 96, |
|
"्": 97, |
|
"क़": 98, |
|
"ख़": 99, |
|
"ग़": 100, |
|
"ज़": 101, |
|
"ड़": 102, |
|
"ढ़": 103, |
|
"फ़": 104, |
|
"।": 105, |
|
"–": 106, |
|
"‘": 107, |
|
"’": 108, |
|
"“": 109, |
|
"”": 110 |
|
} |
|
|