Shubham09 commited on
Commit
30d5ff1
1 Parent(s): 8939208

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 +31 -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 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "[PAD]": 28,
3
+ "[UNK]": 27,
4
+ "a": 8,
5
+ "b": 6,
6
+ "c": 23,
7
+ "d": 10,
8
+ "e": 26,
9
+ "f": 21,
10
+ "g": 24,
11
+ "h": 11,
12
+ "i": 5,
13
+ "j": 1,
14
+ "k": 19,
15
+ "l": 3,
16
+ "m": 7,
17
+ "n": 15,
18
+ "o": 13,
19
+ "p": 9,
20
+ "q": 4,
21
+ "r": 25,
22
+ "s": 12,
23
+ "t": 18,
24
+ "u": 22,
25
+ "v": 0,
26
+ "w": 17,
27
+ "x": 2,
28
+ "y": 14,
29
+ "z": 16,
30
+ "|": 20
31
+ }