Matthijs commited on
Commit
06fcb9d
1 Parent(s): 1664b85

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +4 -0
  2. tokenizer_config.json +9 -0
  3. vocab.json +45 -0
special_tokens_map.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "pad_token": "<pad>",
3
+ "unk_token": "<unk>"
4
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_blank": true,
3
+ "clean_up_tokenization_spaces": true,
4
+ "language": null,
5
+ "model_max_length": 1000000000000000019884624838656,
6
+ "pad_token": "<pad>",
7
+ "tokenizer_class": "VitsMmsTokenizer",
8
+ "unk_token": "<unk>"
9
+ }
vocab.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ " ": 22,
3
+ "'": 37,
4
+ "-": 2,
5
+ "<pad>": 41,
6
+ "<unk>": 42,
7
+ "_": 6,
8
+ "a": 32,
9
+ "b": 18,
10
+ "c": 29,
11
+ "d": 35,
12
+ "e": 40,
13
+ "f": 33,
14
+ "g": 0,
15
+ "h": 11,
16
+ "i": 5,
17
+ "j": 10,
18
+ "k": 23,
19
+ "l": 24,
20
+ "m": 15,
21
+ "n": 20,
22
+ "o": 27,
23
+ "p": 28,
24
+ "q": 7,
25
+ "r": 31,
26
+ "s": 1,
27
+ "t": 38,
28
+ "u": 21,
29
+ "v": 8,
30
+ "w": 9,
31
+ "x": 39,
32
+ "y": 13,
33
+ "z": 16,
34
+ "á": 30,
35
+ "ä": 34,
36
+ "è": 14,
37
+ "é": 19,
38
+ "ë": 25,
39
+ "í": 26,
40
+ "ï": 17,
41
+ "ó": 4,
42
+ "ú": 36,
43
+ "ü": 12,
44
+ "–": 3
45
+ }