Hugging Face's logo

spsither
/
wav2vec2_run1

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Model card Files
xet
 Community
1
spsither commited on
Commit
548bef2
·
1 Parent(s): fa92729

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +6 -0
  2. tokenizer_config.json +11 -0
  3. vocab.json +110 -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 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "do_lower_case": false,
4
+ "eos_token": "</s>",
5
+ "pad_token": "[PAD]",
6
+ "processor_class": "Wav2Vec2Processor",
7
+ "replace_word_delimiter_char": " ",
8
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
9
+ "unk_token": "[UNK]",
10
+ "word_delimiter_token": "|"
11
+ }
vocab.json ADDED
@@ -0,0 +1,110 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "'": 76,
3
+ "[PAD]": 107,
4
+ "[UNK]": 106,
5
+ "|": 71,
6
+ "་": 77,
7
+ "༌": 51,
8
+ "༔": 81,
9
+ "༡": 61,
10
+ "༢": 32,
11
+ "༤": 62,
12
+ "༥": 95,
13
+ "༦": 86,
14
+ "༧": 41,
15
+ "༨": 14,
16
+ "༩": 85,
17
+ "༷": 83,
18
+ "༸": 80,
19
+ "ཀ": 4,
20
+ "ཁ": 34,
21
+ "ག": 11,
22
+ "གྷ": 20,
23
+ "ང": 38,
24
+ "ཅ": 22,
25
+ "ཆ": 8,
26
+ "ཇ": 23,
27
+ "ཉ": 78,
28
+ "ཊ": 63,
29
+ "ཋ": 44,
30
+ "ཌ": 69,
31
+ "ཌྷ": 3,
32
+ "ཎ": 1,
33
+ "ཏ": 91,
34
+ "ཐ": 16,
35
+ "ད": 52,
36
+ "དྷ": 94,
37
+ "ན": 70,
38
+ "པ": 102,
39
+ "ཕ": 89,
40
+ "བ": 21,
41
+ "བྷ": 101,
42
+ "མ": 64,
43
+ "ཙ": 88,
44
+ "ཚ": 47,
45
+ "ཛ": 55,
46
+ "ཝ": 66,
47
+ "ཞ": 92,
48
+ "ཟ": 6,
49
+ "འ": 40,
50
+ "ཡ": 39,
51
+ "ར": 24,
52
+ "ལ": 25,
53
+ "ཤ": 96,
54
+ "ཥ": 5,
55
+ "ས": 103,
56
+ "ཧ": 53,
57
+ "ཨ": 7,
58
+ "ཀྵ": 12,
59
+ "ཪ": 9,
60
+ "ཱ": 31,
61
+ "ི": 43,
62
+ "ཱི": 82,
63
+ "ུ": 27,
64
+ "ཱུ": 19,
65
+ "ེ": 59,
66
+ "ཻ": 58,
67
+ "ོ": 73,
68
+ "ཽ": 50,
69
+ "ཾ": 60,
70
+ "ཿ": 68,
71
+ "ྀ": 93,
72
+ "ྂ": 97,
73
+ "ྃ": 2,
74
+ "྄": 36,
75
+ "྅": 105,
76
+ "ྐ": 30,
77
+ "ྑ": 46,
78
+ "ྒ": 26,
79
+ "ྔ": 17,
80
+ "ྕ": 37,
81
+ "ྖ": 104,
82
+ "ྗ": 28,
83
+ "ྙ": 45,
84
+ "ྚ": 67,
85
+ "ྛ": 13,
86
+ "ྜ": 65,
87
+ "ྞ": 74,
88
+ "ྟ": 84,
89
+ "ྠ": 56,
90
+ "ྡ": 29,
91
+ "ྡྷ": 98,
92
+ "ྣ": 79,
93
+ "ྤ": 0,
94
+ "ྥ": 49,
95
+ "ྦ": 90,
96
+ "ྦྷ": 18,
97
+ "ྨ": 35,
98
+ "ྩ": 72,
99
+ "ྪ": 87,
100
+ "ྫ": 15,
101
+ "ྭ": 99,
102
+ "ྰ": 48,
103
+ "ྱ": 57,
104
+ "ྲ": 54,
105
+ "ླ": 42,
106
+ "ྴ": 75,
107
+ "ྵ": 100,
108
+ "ྶ": 10,
109
+ "ྷ": 33
110
+ }