Hugging Face's logo

rashmi035
/
wav2vec2-large-mms-1b-hindi-colab

Automatic Speech Recognition
Transformers
PyTorch
TensorBoard
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
rashmi035 commited on
Commit
9351182
•
1 Parent(s): 991cd1b

Upload tokenizer

Browse files
Files changed (2) hide show
  1. added_tokens.json +2 -2
  2. vocab.json +82 -81
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 83,
3
- "<s>": 82
4
  }
 
1
  {
2
+ "</s>": 84,
3
+ "<s>": 83
4
  }
vocab.json CHANGED
@@ -1,86 +1,87 @@
1
  {
2
  "hin": {
3
- "!": 1,
4
- "\"": 2,
5
- "'": 3,
6
- ",": 4,
7
- "-": 5,
8
- ".": 6,
9
- ":": 7,
10
- "?": 8,
11
- "F": 9,
12
- "M": 10,
13
- "W": 11,
14
- "[PAD]": 81,
15
- "[UNK]": 80,
16
- "a": 12,
17
- "e": 13,
18
- "i": 14,
19
- "l": 15,
20
- "m": 16,
21
- "p": 17,
22
- "r": 18,
23
- "u": 19,
24
  "|": 0,
25
- "ँ": 20,
26
- "ं": 21,
27
- "ः": 22,
28
- "अ": 23,
29
- "आ": 24,
30
- "इ": 25,
31
- "ई": 26,
32
- "उ": 27,
33
- "ऊ": 28,
34
- "ए": 29,
35
- "ऐ": 30,
36
- "ऑ": 31,
37
- "ओ": 32,
38
- "औ": 33,
39
- "क": 34,
40
- "ख": 35,
41
- "ग": 36,
42
- "घ": 37,
43
- "च": 38,
44
- "छ": 39,
45
- "ज": 40,
46
- "झ": 41,
47
- "ट": 42,
48
- "ठ": 43,
49
- "ड": 44,
50
- "ढ": 45,
51
- "ण": 46,
52
- "त": 47,
53
- "थ": 48,
54
- "द": 49,
55
- "ध": 50,
56
- "न": 51,
57
- "प": 52,
58
- "फ": 53,
59
- "ब": 54,
60
- "भ": 55,
61
- "म": 56,
62
- "य": 57,
63
- "र": 58,
64
- "ल": 59,
65
- "व": 60,
66
- "श": 61,
67
- "ष": 62,
68
- "स": 63,
69
- "ह": 64,
70
- "़": 65,
71
- "ा": 66,
72
- "ि": 67,
73
- "ी": 68,
74
- "ु": 69,
75
- "ू": 70,
76
- "ृ": 71,
77
- "े": 72,
78
- "ै": 73,
79
- "ॉ": 74,
80
- "ो": 75,
81
- "ौ": 76,
82
- "्": 77,
83
- "ड़": 78,
84
- "।": 79
 
 
 
 
 
 
 
 
 
 
 
85
  }
86
  }
 
1
  {
2
  "hin": {
3
+ "#": 1,
4
+ "(": 2,
5
+ ")": 3,
6
+ "0": 4,
7
+ "1": 5,
8
+ "2": 6,
9
+ "3": 7,
10
+ "4": 8,
11
+ "[PAD]": 82,
12
+ "[UNK]": 81,
13
+ "i": 9,
 
 
 
 
 
 
 
 
 
 
14
  "|": 0,
15
+ "ँ": 10,
16
+ "ं": 11,
17
+ "ः": 12,
18
+ "अ": 13,
19
+ "आ": 14,
20
+ "इ": 15,
21
+ "ई": 16,
22
+ "उ": 17,
23
+ "ऊ": 18,
24
+ "ए": 19,
25
+ "ऐ": 20,
26
+ "ऑ": 21,
27
+ "ओ": 22,
28
+ "औ": 23,
29
+ "क": 24,
30
+ "ख": 25,
31
+ "ग": 26,
32
+ "घ": 27,
33
+ "च": 28,
34
+ "छ": 29,
35
+ "ज": 30,
36
+ "झ": 31,
37
+ "ञ": 32,
38
+ "ट": 33,
39
+ "ठ": 34,
40
+ "ड": 35,
41
+ "ढ": 36,
42
+ "ण": 37,
43
+ "त": 38,
44
+ "थ": 39,
45
+ "द": 40,
46
+ "ध": 41,
47
+ "न": 42,
48
+ "प": 43,
49
+ "फ": 44,
50
+ "ब": 45,
51
+ "भ": 46,
52
+ "म": 47,
53
+ "य": 48,
54
+ "र": 49,
55
+ "ल": 50,
56
+ "व": 51,
57
+ "श": 52,
58
+ "ष": 53,
59
+ "स": 54,
60
+ "ह": 55,
61
+ "़": 56,
62
+ "ा": 57,
63
+ "ि": 58,
64
+ "ी": 59,
65
+ "ु": 60,
66
+ "ू": 61,
67
+ "ृ": 62,
68
+ "ॅ": 63,
69
+ "े": 64,
70
+ "ै": 65,
71
+ "ॉ": 66,
72
+ "ो": 67,
73
+ "ौ": 68,
74
+ "्": 69,
75
+ "क़": 70,
76
+ "ख़": 71,
77
+ "ग़": 72,
78
+ "ज़": 73,
79
+ "ड़": 74,
80
+ "ढ़": 75,
81
+ "फ़": 76,
82
+ "।": 77,
83
+ "०": 78,
84
+ "–": 79,
85
+ "’": 80
86
  }
87
  }