Hugging Face's logo

Anjaan-Khadka
/
asr

Model card Files Community
Anjaan-Khadka commited on
Commit
60c49df
·
1 Parent(s): 7f08c40

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 +65 -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 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "[PAD]": 62,
3
+ "[UNK]": 61,
4
+ "|": 43,
5
+ "ँ": 46,
6
+ "ं": 54,
7
+ "ः": 37,
8
+ "अ": 40,
9
+ "आ": 51,
10
+ "इ": 30,
11
+ "ई": 0,
12
+ "उ": 42,
13
+ "ऊ": 24,
14
+ "ऋ": 55,
15
+ "ए": 14,
16
+ "ऐ": 45,
17
+ "ओ": 31,
18
+ "औ": 2,
19
+ "क": 12,
20
+ "ख": 4,
21
+ "ग": 58,
22
+ "घ": 39,
23
+ "ङ": 25,
24
+ "च": 60,
25
+ "छ": 19,
26
+ "ज": 48,
27
+ "झ": 1,
28
+ "ञ": 15,
29
+ "ट": 11,
30
+ "ठ": 22,
31
+ "ड": 50,
32
+ "ढ": 8,
33
+ "ण": 57,
34
+ "त": 33,
35
+ "थ": 13,
36
+ "द": 6,
37
+ "ध": 7,
38
+ "न": 27,
39
+ "प": 20,
40
+ "फ": 17,
41
+ "ब": 28,
42
+ "भ": 21,
43
+ "म": 59,
44
+ "य": 23,
45
+ "र": 16,
46
+ "ल": 3,
47
+ "व": 32,
48
+ "श": 38,
49
+ "ष": 34,
50
+ "स": 41,
51
+ "ह": 49,
52
+ "ा": 26,
53
+ "ि": 35,
54
+ "ी": 36,
55
+ "ु": 18,
56
+ "ू": 52,
57
+ "ृ": 56,
58
+ "े": 44,
59
+ "ै": 9,
60
+ "ो": 10,
61
+ "ौ": 29,
62
+ "्": 47,
63
+ "।": 53,
64
+ "‍": 5
65
+ }