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