add tokenizer
Browse files- vocab.json +110 -110
vocab.json
CHANGED
@@ -1,114 +1,114 @@
|
|
1 |
{
|
2 |
-
"1":
|
3 |
-
"2":
|
4 |
-
"3":
|
5 |
-
"4":
|
6 |
-
"5":
|
7 |
-
"7":
|
8 |
-
"8":
|
9 |
-
"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 |
"[PAD]": 111,
|
36 |
"[UNK]": 110,
|
37 |
-
"a":
|
38 |
-
"b":
|
39 |
-
"c":
|
40 |
-
"d":
|
41 |
-
"e":
|
42 |
-
"f":
|
43 |
-
"g":
|
44 |
-
"h":
|
45 |
-
"i":
|
46 |
-
"j":
|
47 |
-
"k":
|
48 |
-
"l":
|
49 |
-
"m":
|
50 |
-
"n":
|
51 |
-
"o":
|
52 |
-
"p":
|
53 |
-
"q":
|
54 |
-
"r":
|
55 |
-
"s":
|
56 |
-
"t":
|
57 |
-
"u":
|
58 |
-
"v":
|
59 |
-
"w":
|
60 |
-
"x":
|
61 |
-
"y":
|
62 |
-
"z":
|
63 |
-
"|":
|
64 |
-
"γ±":
|
65 |
-
"γ²":
|
66 |
-
"γ³":
|
67 |
-
"γ΄":
|
68 |
-
"γ΅":
|
69 |
-
"γΆ":
|
70 |
-
"γ·":
|
71 |
-
"γΈ":
|
72 |
-
"γΉ":
|
73 |
-
"γΊ":
|
74 |
-
"γ»":
|
75 |
-
"γΌ":
|
76 |
-
"γΎ":
|
77 |
-
"γΏ":
|
78 |
-
"γ
":
|
79 |
-
"γ
":
|
80 |
-
"γ
":
|
81 |
-
"γ
":
|
82 |
-
"γ
":
|
83 |
-
"γ
":
|
84 |
-
"γ
":
|
85 |
-
"γ
":
|
86 |
-
"γ
":
|
87 |
-
"γ
":
|
88 |
-
"γ
":
|
89 |
-
"γ
":
|
90 |
-
"γ
":
|
91 |
-
"γ
":
|
92 |
-
"γ
":
|
93 |
-
"γ
":
|
94 |
-
"γ
":
|
95 |
-
"γ
":
|
96 |
-
"γ
":
|
97 |
-
"γ
":
|
98 |
-
"γ
":
|
99 |
-
"γ
":
|
100 |
-
"γ
":
|
101 |
-
"γ
":
|
102 |
-
"γ
":
|
103 |
-
"γ
":
|
104 |
-
"γ
":
|
105 |
-
"γ
":
|
106 |
-
"γ
":
|
107 |
-
"γ
":
|
108 |
-
"γ
":
|
109 |
-
"γ
":
|
110 |
-
"γ
":
|
111 |
-
"γ
‘":
|
112 |
-
"γ
’":
|
113 |
-
"γ
£":
|
114 |
}
|
|
|
1 |
{
|
2 |
+
"1": 82,
|
3 |
+
"2": 94,
|
4 |
+
"3": 12,
|
5 |
+
"4": 97,
|
6 |
+
"5": 81,
|
7 |
+
"7": 48,
|
8 |
+
"8": 68,
|
9 |
+
"A": 77,
|
10 |
+
"B": 79,
|
11 |
+
"C": 98,
|
12 |
+
"D": 100,
|
13 |
+
"E": 62,
|
14 |
+
"F": 26,
|
15 |
+
"G": 16,
|
16 |
+
"H": 23,
|
17 |
+
"I": 58,
|
18 |
+
"J": 91,
|
19 |
+
"K": 20,
|
20 |
+
"L": 88,
|
21 |
+
"M": 102,
|
22 |
+
"N": 86,
|
23 |
+
"O": 37,
|
24 |
+
"P": 42,
|
25 |
+
"Q": 80,
|
26 |
+
"R": 15,
|
27 |
+
"S": 70,
|
28 |
+
"T": 17,
|
29 |
+
"U": 11,
|
30 |
+
"V": 108,
|
31 |
+
"W": 29,
|
32 |
+
"X": 107,
|
33 |
+
"Y": 14,
|
34 |
+
"Z": 9,
|
35 |
"[PAD]": 111,
|
36 |
"[UNK]": 110,
|
37 |
+
"a": 78,
|
38 |
+
"b": 2,
|
39 |
+
"c": 67,
|
40 |
+
"d": 59,
|
41 |
+
"e": 69,
|
42 |
+
"f": 21,
|
43 |
+
"g": 53,
|
44 |
+
"h": 33,
|
45 |
+
"i": 89,
|
46 |
+
"j": 47,
|
47 |
+
"k": 71,
|
48 |
+
"l": 57,
|
49 |
+
"m": 13,
|
50 |
+
"n": 56,
|
51 |
+
"o": 34,
|
52 |
+
"p": 61,
|
53 |
+
"q": 19,
|
54 |
+
"r": 36,
|
55 |
+
"s": 27,
|
56 |
+
"t": 64,
|
57 |
+
"u": 93,
|
58 |
+
"v": 39,
|
59 |
+
"w": 25,
|
60 |
+
"x": 32,
|
61 |
+
"y": 38,
|
62 |
+
"z": 40,
|
63 |
+
"|": 24,
|
64 |
+
"γ±": 43,
|
65 |
+
"γ²": 46,
|
66 |
+
"γ³": 22,
|
67 |
+
"γ΄": 103,
|
68 |
+
"γ΅": 109,
|
69 |
+
"γΆ": 74,
|
70 |
+
"γ·": 41,
|
71 |
+
"γΈ": 76,
|
72 |
+
"γΉ": 83,
|
73 |
+
"γΊ": 66,
|
74 |
+
"γ»": 51,
|
75 |
+
"γΌ": 4,
|
76 |
+
"γΎ": 105,
|
77 |
+
"γΏ": 31,
|
78 |
+
"γ
": 87,
|
79 |
+
"γ
": 52,
|
80 |
+
"γ
": 55,
|
81 |
+
"γ
": 35,
|
82 |
+
"γ
": 85,
|
83 |
+
"γ
": 92,
|
84 |
+
"γ
": 50,
|
85 |
+
"γ
": 7,
|
86 |
+
"γ
": 90,
|
87 |
+
"γ
": 84,
|
88 |
+
"γ
": 8,
|
89 |
+
"γ
": 72,
|
90 |
+
"γ
": 96,
|
91 |
+
"γ
": 60,
|
92 |
+
"γ
": 44,
|
93 |
+
"γ
": 95,
|
94 |
+
"γ
": 54,
|
95 |
+
"γ
": 0,
|
96 |
+
"γ
": 101,
|
97 |
+
"γ
": 49,
|
98 |
+
"γ
": 45,
|
99 |
+
"γ
": 75,
|
100 |
+
"γ
": 106,
|
101 |
+
"γ
": 63,
|
102 |
+
"γ
": 10,
|
103 |
+
"γ
": 104,
|
104 |
+
"γ
": 6,
|
105 |
+
"γ
": 3,
|
106 |
+
"γ
": 1,
|
107 |
+
"γ
": 99,
|
108 |
+
"γ
": 18,
|
109 |
+
"γ
": 30,
|
110 |
+
"γ
": 28,
|
111 |
+
"γ
‘": 5,
|
112 |
+
"γ
’": 65,
|
113 |
+
"γ
£": 73
|
114 |
}
|