natmin322 commited on
Commit
3e6cc3d
·
1 Parent(s): 9d3d4f7

Upload 3 files

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +6 -0
  2. tokenizer_config.json +11 -0
  3. vocab.json +96 -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 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "'": 62,
3
+ "[PAD]": 93,
4
+ "[UNK]": 92,
5
+ "a": 49,
6
+ "b": 0,
7
+ "c": 71,
8
+ "d": 34,
9
+ "e": 8,
10
+ "f": 87,
11
+ "g": 40,
12
+ "h": 29,
13
+ "i": 80,
14
+ "k": 3,
15
+ "l": 28,
16
+ "m": 63,
17
+ "n": 66,
18
+ "o": 85,
19
+ "p": 16,
20
+ "q": 54,
21
+ "r": 19,
22
+ "s": 10,
23
+ "t": 39,
24
+ "u": 23,
25
+ "v": 50,
26
+ "x": 53,
27
+ "y": 55,
28
+ "|": 17,
29
+ "à": 32,
30
+ "á": 59,
31
+ "â": 35,
32
+ "ã": 18,
33
+ "è": 70,
34
+ "é": 86,
35
+ "ê": 64,
36
+ "ì": 44,
37
+ "í": 69,
38
+ "ò": 79,
39
+ "ó": 82,
40
+ "ô": 67,
41
+ "õ": 57,
42
+ "ù": 22,
43
+ "ú": 48,
44
+ "ý": 91,
45
+ "ă": 89,
46
+ "đ": 41,
47
+ "ĩ": 6,
48
+ "ũ": 81,
49
+ "ơ": 1,
50
+ "ư": 43,
51
+ "ạ": 37,
52
+ "ả": 36,
53
+ "ấ": 2,
54
+ "ầ": 27,
55
+ "ẩ": 13,
56
+ "ẫ": 4,
57
+ "ậ": 60,
58
+ "ắ": 65,
59
+ "ằ": 25,
60
+ "ẳ": 20,
61
+ "ẵ": 84,
62
+ "ặ": 88,
63
+ "ẹ": 75,
64
+ "ẻ": 45,
65
+ "ẽ": 21,
66
+ "ế": 68,
67
+ "ề": 61,
68
+ "ể": 47,
69
+ "ễ": 26,
70
+ "ệ": 56,
71
+ "ỉ": 74,
72
+ "ị": 72,
73
+ "ọ": 58,
74
+ "ỏ": 9,
75
+ "ố": 7,
76
+ "ồ": 38,
77
+ "ổ": 14,
78
+ "ỗ": 46,
79
+ "ộ": 90,
80
+ "ớ": 5,
81
+ "ờ": 77,
82
+ "ở": 15,
83
+ "ỡ": 42,
84
+ "ợ": 76,
85
+ "ụ": 73,
86
+ "ủ": 30,
87
+ "ứ": 24,
88
+ "ừ": 52,
89
+ "ử": 51,
90
+ "ữ": 83,
91
+ "ự": 11,
92
+ "ỳ": 12,
93
+ "ỵ": 33,
94
+ "ỷ": 78,
95
+ "ỹ": 31
96
+ }