Upload tokenizer
Browse files- vocab.json +68 -68
vocab.json
CHANGED
@@ -1,73 +1,73 @@
|
|
1 |
{
|
2 |
-
"\n":
|
3 |
-
"&":
|
4 |
-
"'":
|
5 |
-
"(":
|
6 |
-
")":
|
7 |
-
"/":
|
8 |
-
"0":
|
9 |
-
"1":
|
10 |
-
"2":
|
11 |
-
"3":
|
12 |
-
"4":
|
13 |
-
"5":
|
14 |
-
"6":
|
15 |
-
"7":
|
16 |
-
"8":
|
17 |
-
"9":
|
18 |
-
"[":
|
19 |
"[PAD]": 70,
|
20 |
"[UNK]": 69,
|
21 |
-
"]":
|
22 |
-
"a":
|
23 |
-
"b":
|
24 |
-
"c":
|
25 |
-
"d":
|
26 |
-
"e":
|
27 |
-
"f":
|
28 |
-
"g":
|
29 |
-
"h":
|
30 |
-
"i":
|
31 |
-
"j":
|
32 |
-
"k":
|
33 |
-
"l":
|
34 |
-
"m":
|
35 |
-
"n":
|
36 |
-
"o":
|
37 |
-
"p":
|
38 |
-
"q":
|
39 |
-
"r":
|
40 |
-
"s":
|
41 |
-
"t":
|
42 |
-
"u":
|
43 |
-
"v":
|
44 |
-
"w":
|
45 |
-
"x":
|
46 |
-
"y":
|
47 |
-
"z":
|
48 |
-
"|":
|
49 |
-
" ":
|
50 |
-
"´":
|
51 |
-
"ß":
|
52 |
-
"à":
|
53 |
-
"á":
|
54 |
-
"â":
|
55 |
-
"ä":
|
56 |
-
"ç":
|
57 |
-
"è":
|
58 |
-
"é":
|
59 |
-
"ê":
|
60 |
-
"ë":
|
61 |
-
"î":
|
62 |
-
"ï":
|
63 |
-
"ô":
|
64 |
"ö": 35,
|
65 |
-
"û":
|
66 |
-
"ü":
|
67 |
-
"ě":
|
68 |
-
"ğ":
|
69 |
-
"‒":
|
70 |
-
"–":
|
71 |
-
"’":
|
72 |
-
"„":
|
73 |
}
|
|
|
1 |
{
|
2 |
+
"\n": 42,
|
3 |
+
"&": 66,
|
4 |
+
"'": 61,
|
5 |
+
"(": 32,
|
6 |
+
")": 16,
|
7 |
+
"/": 11,
|
8 |
+
"0": 48,
|
9 |
+
"1": 1,
|
10 |
+
"2": 27,
|
11 |
+
"3": 20,
|
12 |
+
"4": 45,
|
13 |
+
"5": 43,
|
14 |
+
"6": 34,
|
15 |
+
"7": 21,
|
16 |
+
"8": 39,
|
17 |
+
"9": 54,
|
18 |
+
"[": 37,
|
19 |
"[PAD]": 70,
|
20 |
"[UNK]": 69,
|
21 |
+
"]": 40,
|
22 |
+
"a": 33,
|
23 |
+
"b": 58,
|
24 |
+
"c": 46,
|
25 |
+
"d": 7,
|
26 |
+
"e": 63,
|
27 |
+
"f": 57,
|
28 |
+
"g": 18,
|
29 |
+
"h": 15,
|
30 |
+
"i": 25,
|
31 |
+
"j": 31,
|
32 |
+
"k": 44,
|
33 |
+
"l": 59,
|
34 |
+
"m": 17,
|
35 |
+
"n": 10,
|
36 |
+
"o": 6,
|
37 |
+
"p": 13,
|
38 |
+
"q": 47,
|
39 |
+
"r": 22,
|
40 |
+
"s": 38,
|
41 |
+
"t": 67,
|
42 |
+
"u": 51,
|
43 |
+
"v": 30,
|
44 |
+
"w": 68,
|
45 |
+
"x": 14,
|
46 |
+
"y": 26,
|
47 |
+
"z": 5,
|
48 |
+
"|": 60,
|
49 |
+
" ": 36,
|
50 |
+
"´": 0,
|
51 |
+
"ß": 9,
|
52 |
+
"à": 52,
|
53 |
+
"á": 49,
|
54 |
+
"â": 55,
|
55 |
+
"ä": 23,
|
56 |
+
"ç": 56,
|
57 |
+
"è": 8,
|
58 |
+
"é": 4,
|
59 |
+
"ê": 29,
|
60 |
+
"ë": 28,
|
61 |
+
"î": 2,
|
62 |
+
"ï": 3,
|
63 |
+
"ô": 12,
|
64 |
"ö": 35,
|
65 |
+
"û": 53,
|
66 |
+
"ü": 64,
|
67 |
+
"ě": 50,
|
68 |
+
"ğ": 24,
|
69 |
+
"‒": 19,
|
70 |
+
"–": 62,
|
71 |
+
"’": 65,
|
72 |
+
"„": 41
|
73 |
}
|