Upload tokenizer vocab
Browse files- vocab.json +103 -0
vocab.json
ADDED
|
@@ -0,0 +1,103 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[PAD]": 0,
|
| 3 |
+
"<s>": 1,
|
| 4 |
+
"</s>": 2,
|
| 5 |
+
"[INST]": 3,
|
| 6 |
+
"[/INST]": 4,
|
| 7 |
+
"\t": 5,
|
| 8 |
+
"\n": 6,
|
| 9 |
+
" ": 7,
|
| 10 |
+
"!": 8,
|
| 11 |
+
"\"": 9,
|
| 12 |
+
"#": 10,
|
| 13 |
+
"$": 11,
|
| 14 |
+
"%": 12,
|
| 15 |
+
"&": 13,
|
| 16 |
+
"'": 14,
|
| 17 |
+
"(": 15,
|
| 18 |
+
")": 16,
|
| 19 |
+
"*": 17,
|
| 20 |
+
"+": 18,
|
| 21 |
+
",": 19,
|
| 22 |
+
"-": 20,
|
| 23 |
+
".": 21,
|
| 24 |
+
"/": 22,
|
| 25 |
+
"0": 23,
|
| 26 |
+
"1": 24,
|
| 27 |
+
"2": 25,
|
| 28 |
+
"3": 26,
|
| 29 |
+
"4": 27,
|
| 30 |
+
"5": 28,
|
| 31 |
+
"6": 29,
|
| 32 |
+
"7": 30,
|
| 33 |
+
"8": 31,
|
| 34 |
+
"9": 32,
|
| 35 |
+
":": 33,
|
| 36 |
+
";": 34,
|
| 37 |
+
"<": 35,
|
| 38 |
+
"=": 36,
|
| 39 |
+
">": 37,
|
| 40 |
+
"?": 38,
|
| 41 |
+
"@": 39,
|
| 42 |
+
"A": 40,
|
| 43 |
+
"B": 41,
|
| 44 |
+
"C": 42,
|
| 45 |
+
"D": 43,
|
| 46 |
+
"E": 44,
|
| 47 |
+
"F": 45,
|
| 48 |
+
"G": 46,
|
| 49 |
+
"H": 47,
|
| 50 |
+
"I": 48,
|
| 51 |
+
"J": 49,
|
| 52 |
+
"K": 50,
|
| 53 |
+
"L": 51,
|
| 54 |
+
"M": 52,
|
| 55 |
+
"N": 53,
|
| 56 |
+
"O": 54,
|
| 57 |
+
"P": 55,
|
| 58 |
+
"Q": 56,
|
| 59 |
+
"R": 57,
|
| 60 |
+
"S": 58,
|
| 61 |
+
"T": 59,
|
| 62 |
+
"U": 60,
|
| 63 |
+
"V": 61,
|
| 64 |
+
"W": 62,
|
| 65 |
+
"X": 63,
|
| 66 |
+
"Y": 64,
|
| 67 |
+
"Z": 65,
|
| 68 |
+
"[": 66,
|
| 69 |
+
"]": 67,
|
| 70 |
+
"^": 68,
|
| 71 |
+
"_": 69,
|
| 72 |
+
"`": 70,
|
| 73 |
+
"a": 71,
|
| 74 |
+
"b": 72,
|
| 75 |
+
"c": 73,
|
| 76 |
+
"d": 74,
|
| 77 |
+
"e": 75,
|
| 78 |
+
"f": 76,
|
| 79 |
+
"g": 77,
|
| 80 |
+
"h": 78,
|
| 81 |
+
"i": 79,
|
| 82 |
+
"j": 80,
|
| 83 |
+
"k": 81,
|
| 84 |
+
"l": 82,
|
| 85 |
+
"m": 83,
|
| 86 |
+
"n": 84,
|
| 87 |
+
"o": 85,
|
| 88 |
+
"p": 86,
|
| 89 |
+
"q": 87,
|
| 90 |
+
"r": 88,
|
| 91 |
+
"s": 89,
|
| 92 |
+
"t": 90,
|
| 93 |
+
"u": 91,
|
| 94 |
+
"v": 92,
|
| 95 |
+
"w": 93,
|
| 96 |
+
"x": 94,
|
| 97 |
+
"y": 95,
|
| 98 |
+
"z": 96,
|
| 99 |
+
"{": 97,
|
| 100 |
+
"|": 98,
|
| 101 |
+
"}": 99,
|
| 102 |
+
"~": 100
|
| 103 |
+
}
|