Hugging Face's logo

dmatekenya
/
wav2vec2-large-mms-1b-chich-colab

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Community
dmatekenya commited on
Commit
7b75834
•
1 Parent(s): 9757b92

Upload tokenizer

Browse files
Files changed (3) hide show
  1. added_tokens.json +2 -2
  2. tokenizer_config.json +4 -4
  3. vocab.json +41 -45
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 48,
3
- "<s>": 47
4
  }
 
1
  {
2
+ "</s>": 44,
3
+ "<s>": 43
4
  }
tokenizer_config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "45": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
@@ -8,7 +8,7 @@
8
  "single_word": false,
9
  "special": false
10
  },
11
- "46": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,7 +16,7 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "47": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": true,
@@ -24,7 +24,7 @@
24
  "single_word": false,
25
  "special": true
26
  },
27
- "48": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": true,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "41": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
11
+ "42": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
 
16
  "single_word": false,
17
  "special": false
18
  },
19
+ "43": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": true,
 
24
  "single_word": false,
25
  "special": true
26
  },
27
+ "44": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": true,
vocab.json CHANGED
@@ -1,51 +1,47 @@
1
  {
2
  "nya": {
3
  "\t": 0,
4
- "$": 2,
5
- "(": 3,
6
- ")": 4,
7
- "/": 5,
8
- "0": 6,
9
- "1": 7,
10
- "2": 8,
11
- "3": 9,
12
- "4": 10,
13
- "5": 11,
14
- "6": 12,
15
- "7": 13,
16
- "8": 14,
17
- "9": 15,
18
- "[PAD]": 46,
19
- "[UNK]": 45,
20
- "a": 16,
21
- "b": 17,
22
- "c": 18,
23
- "d": 19,
24
- "e": 20,
25
- "f": 21,
26
- "g": 22,
27
- "h": 23,
28
- "i": 24,
29
- "j": 25,
30
- "k": 26,
31
- "l": 27,
32
- "m": 28,
33
- "n": 29,
34
- "o": 30,
35
- "p": 31,
36
- "q": 32,
37
- "r": 33,
38
- "s": 34,
39
- "t": 35,
40
- "u": 36,
41
- "v": 37,
42
- "w": 38,
43
- "x": 39,
44
- "y": 40,
45
- "z": 41,
46
  "|": 1,
47
- "’": 42,
48
- "…": 43,
49
- "": 44
50
  }
51
  }
 
1
  {
2
  "nya": {
3
  "\t": 0,
4
+ "/": 2,
5
+ "0": 3,
6
+ "1": 4,
7
+ "2": 5,
8
+ "3": 6,
9
+ "4": 7,
10
+ "5": 8,
11
+ "6": 9,
12
+ "7": 10,
13
+ "8": 11,
14
+ "9": 12,
15
+ "[PAD]": 42,
16
+ "[UNK]": 41,
17
+ "a": 13,
18
+ "b": 14,
19
+ "c": 15,
20
+ "d": 16,
21
+ "e": 17,
22
+ "f": 18,
23
+ "g": 19,
24
+ "h": 20,
25
+ "i": 21,
26
+ "j": 22,
27
+ "k": 23,
28
+ "l": 24,
29
+ "m": 25,
30
+ "n": 26,
31
+ "o": 27,
32
+ "p": 28,
33
+ "q": 29,
34
+ "r": 30,
35
+ "s": 31,
36
+ "t": 32,
37
+ "u": 33,
38
+ "v": 34,
39
+ "w": 35,
40
+ "x": 36,
41
+ "y": 37,
42
+ "z": 38,
 
 
 
43
  "|": 1,
44
+ "’": 39,
45
+ "": 40
 
46
  }
47
  }