monideep2255 commited on
Commit
a54893a
1 Parent(s): 2b73d7a

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +6 -0
  2. tokenizer_config.json +10 -0
  3. vocab.json +34 -0
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "[PAD]",
5
+ "unk_token": "[UNK]"
6
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "do_lower_case": false,
4
+ "eos_token": "</s>",
5
+ "pad_token": "[PAD]",
6
+ "replace_word_delimiter_char": " ",
7
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
8
+ "unk_token": "[UNK]",
9
+ "word_delimiter_token": "|"
10
+ }
vocab.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "'": 29,
3
+ "1": 12,
4
+ "3": 18,
5
+ "[PAD]": 31,
6
+ "[UNK]": 30,
7
+ "a": 14,
8
+ "b": 27,
9
+ "c": 19,
10
+ "d": 24,
11
+ "e": 28,
12
+ "f": 7,
13
+ "g": 6,
14
+ "h": 3,
15
+ "i": 22,
16
+ "j": 0,
17
+ "k": 21,
18
+ "l": 16,
19
+ "m": 2,
20
+ "n": 5,
21
+ "o": 10,
22
+ "p": 17,
23
+ "q": 11,
24
+ "r": 8,
25
+ "s": 9,
26
+ "t": 20,
27
+ "u": 23,
28
+ "v": 4,
29
+ "w": 25,
30
+ "x": 1,
31
+ "y": 15,
32
+ "z": 13,
33
+ "|": 26
34
+ }