Hugging Face's logo

ogbi
/
wav2vec2-mms-300m-ikk-1

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
wav2vec2
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
ogbi commited on
Commit
4f3f5a1
·
verified ·
1 Parent(s): f622bbd

Upload tokenizer

Browse files
Files changed (3) hide show
  1. added_tokens.json +2 -2
  2. tokenizer_config.json +4 -4
  3. vocab.json +33 -32
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 34,
3
- "<s>": 33
4
  }
 
1
  {
2
+ "</s>": 35,
3
+ "<s>": 34
4
  }
tokenizer_config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "31": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
@@ -8,7 +8,7 @@
8
  "single_word": false,
9
  "special": false
10
  },
11
- "32": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,7 +16,7 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "33": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
@@ -24,7 +24,7 @@
24
  "single_word": false,
25
  "special": true
26
  },
27
- "34": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "32": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
11
+ "33": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
 
16
  "single_word": false,
17
  "special": false
18
  },
19
+ "34": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
 
24
  "single_word": false,
25
  "special": true
26
  },
27
+ "35": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
vocab.json CHANGED
@@ -1,35 +1,36 @@
1
  {
2
- "[PAD]": 32,
3
- "[UNK]": 31,
4
- "a": 1,
5
- "b": 2,
6
- "c": 3,
7
- "d": 4,
8
- "e": 5,
9
- "f": 6,
10
- "g": 7,
11
- "h": 8,
12
- "i": 9,
13
- "j": 10,
14
- "k": 11,
15
- "l": 12,
16
- "m": 13,
17
- "n": 14,
18
- "o": 15,
19
- "p": 16,
20
- "r": 17,
21
- "s": 18,
22
- "t": 19,
23
- "u": 20,
24
- "v": 21,
25
- "w": 22,
26
- "y": 23,
27
- "z": 24,
 
28
  "|": 0,
29
- "ṅ": 25,
30
- "ẹ": 26,
31
- "ị": 27,
32
- "ọ": 28,
33
- "ụ": 29,
34
- "’": 30
35
  }
 
1
  {
2
+ "'": 1,
3
+ "[PAD]": 33,
4
+ "[UNK]": 32,
5
+ "a": 2,
6
+ "b": 3,
7
+ "c": 4,
8
+ "d": 5,
9
+ "e": 6,
10
+ "f": 7,
11
+ "g": 8,
12
+ "h": 9,
13
+ "i": 10,
14
+ "j": 11,
15
+ "k": 12,
16
+ "l": 13,
17
+ "m": 14,
18
+ "n": 15,
19
+ "o": 16,
20
+ "p": 17,
21
+ "r": 18,
22
+ "s": 19,
23
+ "t": 20,
24
+ "u": 21,
25
+ "v": 22,
26
+ "w": 23,
27
+ "y": 24,
28
+ "z": 25,
29
  "|": 0,
30
+ "ṅ": 26,
31
+ "ẹ": 27,
32
+ "ị": 28,
33
+ "ọ": 29,
34
+ "ụ": 30,
35
+ "’": 31
36
  }