Upload tokenizer
Browse files- vocab.json +40 -40
vocab.json
CHANGED
@@ -1,46 +1,46 @@
|
|
1 |
{
|
2 |
-
"'":
|
3 |
"[PAD]": 43,
|
4 |
"[UNK]": 42,
|
5 |
-
"a":
|
6 |
-
"b":
|
7 |
-
"c":
|
8 |
-
"d":
|
9 |
-
"e":
|
10 |
-
"f":
|
11 |
-
"g":
|
12 |
-
"h":
|
13 |
-
"i":
|
14 |
-
"j":
|
15 |
-
"k":
|
16 |
-
"l":
|
17 |
-
"m":
|
18 |
-
"n":
|
19 |
"o": 39,
|
20 |
-
"p":
|
21 |
-
"q":
|
22 |
-
"r":
|
23 |
-
"s":
|
24 |
-
"t":
|
25 |
-
"u":
|
26 |
-
"v":
|
27 |
-
"w":
|
28 |
-
"x":
|
29 |
-
"y":
|
30 |
-
"z":
|
31 |
-
"|":
|
32 |
-
"à":
|
33 |
-
"á":
|
34 |
-
"â":
|
35 |
-
"ä":
|
36 |
-
"è":
|
37 |
-
"é":
|
38 |
-
"ê":
|
39 |
-
"ë":
|
40 |
-
"ï":
|
41 |
-
"ó":
|
42 |
-
"ô":
|
43 |
-
"ú":
|
44 |
"û": 20,
|
45 |
-
"ü":
|
46 |
}
|
|
|
1 |
{
|
2 |
+
"'": 24,
|
3 |
"[PAD]": 43,
|
4 |
"[UNK]": 42,
|
5 |
+
"a": 17,
|
6 |
+
"b": 14,
|
7 |
+
"c": 7,
|
8 |
+
"d": 19,
|
9 |
+
"e": 13,
|
10 |
+
"f": 6,
|
11 |
+
"g": 16,
|
12 |
+
"h": 36,
|
13 |
+
"i": 0,
|
14 |
+
"j": 40,
|
15 |
+
"k": 1,
|
16 |
+
"l": 22,
|
17 |
+
"m": 41,
|
18 |
+
"n": 33,
|
19 |
"o": 39,
|
20 |
+
"p": 34,
|
21 |
+
"q": 23,
|
22 |
+
"r": 29,
|
23 |
+
"s": 31,
|
24 |
+
"t": 26,
|
25 |
+
"u": 18,
|
26 |
+
"v": 10,
|
27 |
+
"w": 25,
|
28 |
+
"x": 28,
|
29 |
+
"y": 38,
|
30 |
+
"z": 37,
|
31 |
+
"|": 5,
|
32 |
+
"à": 3,
|
33 |
+
"á": 9,
|
34 |
+
"â": 15,
|
35 |
+
"ä": 21,
|
36 |
+
"è": 4,
|
37 |
+
"é": 12,
|
38 |
+
"ê": 2,
|
39 |
+
"ë": 11,
|
40 |
+
"ï": 32,
|
41 |
+
"ó": 27,
|
42 |
+
"ô": 35,
|
43 |
+
"ú": 8,
|
44 |
"û": 20,
|
45 |
+
"ü": 30
|
46 |
}
|