Hugging Face's logo

auliaadila
/
wav2vec2-large-xlsr-ind-v1

Model card Files Community
auliaadila commited on
Commit
c7ed267
•
1 Parent(s): 0a6e988

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +0 -16
  2. tokenizer_config.json +0 -1
  3. vocab.json +37 -37
special_tokens_map.json CHANGED
@@ -1,20 +1,4 @@
1
  {
2
- "additional_special_tokens": [
3
- {
4
- "content": "<s>",
5
- "lstrip": false,
6
- "normalized": true,
7
- "rstrip": false,
8
- "single_word": false
9
- },
10
- {
11
- "content": "</s>",
12
- "lstrip": false,
13
- "normalized": true,
14
- "rstrip": false,
15
- "single_word": false
16
- }
17
- ],
18
  "bos_token": "<s>",
19
  "eos_token": "</s>",
20
  "pad_token": "[PAD]",
 
1
  {
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  "bos_token": "<s>",
3
  "eos_token": "</s>",
4
  "pad_token": "[PAD]",
tokenizer_config.json CHANGED
@@ -5,7 +5,6 @@
5
  "eos_token": "</s>",
6
  "model_max_length": 1000000000000000019884624838656,
7
  "pad_token": "[PAD]",
8
- "processor_class": "Wav2Vec2Processor",
9
  "replace_word_delimiter_char": " ",
10
  "target_lang": null,
11
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
 
5
  "eos_token": "</s>",
6
  "model_max_length": 1000000000000000019884624838656,
7
  "pad_token": "[PAD]",
 
8
  "replace_word_delimiter_char": " ",
9
  "target_lang": null,
10
  "tokenizer_class": "Wav2Vec2CTCTokenizer",
vocab.json CHANGED
@@ -1,41 +1,41 @@
1
  {
2
- "0": 3,
3
- "1": 30,
4
- "2": 15,
5
- "3": 23,
6
- "4": 9,
7
- "5": 2,
8
- "6": 0,
9
- "7": 14,
10
- "8": 10,
11
- "9": 17,
12
  "[PAD]": 38,
13
  "[UNK]": 37,
14
- "a": 34,
15
- "b": 7,
16
- "c": 22,
17
- "d": 31,
18
- "e": 6,
19
- "f": 12,
20
- "g": 21,
21
- "h": 19,
22
- "i": 25,
23
- "j": 5,
24
- "k": 20,
25
- "l": 36,
26
- "m": 28,
27
- "n": 32,
28
- "o": 27,
29
- "p": 18,
30
- "q": 29,
31
- "r": 13,
32
- "s": 24,
33
- "t": 11,
34
- "u": 35,
35
- "v": 8,
36
- "w": 4,
37
- "x": 16,
38
- "y": 33,
39
- "z": 1,
40
- "|": 26
41
  }
 
1
  {
2
+ "0": 20,
3
+ "1": 26,
4
+ "2": 27,
5
+ "3": 14,
6
+ "4": 34,
7
+ "5": 16,
8
+ "6": 29,
9
+ "7": 2,
10
+ "8": 19,
11
+ "9": 23,
12
  "[PAD]": 38,
13
  "[UNK]": 37,
14
+ "a": 15,
15
+ "b": 28,
16
+ "c": 12,
17
+ "d": 4,
18
+ "e": 13,
19
+ "f": 0,
20
+ "g": 7,
21
+ "h": 9,
22
+ "i": 18,
23
+ "j": 30,
24
+ "k": 10,
25
+ "l": 17,
26
+ "m": 11,
27
+ "n": 3,
28
+ "o": 33,
29
+ "p": 6,
30
+ "q": 8,
31
+ "r": 22,
32
+ "s": 25,
33
+ "t": 31,
34
+ "u": 36,
35
+ "v": 21,
36
+ "w": 24,
37
+ "x": 1,
38
+ "y": 32,
39
+ "z": 35,
40
+ "|": 5
41
  }