lynguyenminh commited on
Commit
05f5b55
·
1 Parent(s): 4baa897

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 +99 -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 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "%": 48,
3
+ "0": 32,
4
+ "1": 20,
5
+ "2": 10,
6
+ "3": 24,
7
+ "4": 37,
8
+ "5": 21,
9
+ "6": 1,
10
+ "7": 76,
11
+ "8": 6,
12
+ "9": 47,
13
+ "[PAD]": 96,
14
+ "[UNK]": 95,
15
+ "a": 86,
16
+ "b": 69,
17
+ "c": 87,
18
+ "d": 61,
19
+ "e": 16,
20
+ "f": 5,
21
+ "g": 74,
22
+ "h": 19,
23
+ "i": 44,
24
+ "k": 63,
25
+ "l": 89,
26
+ "m": 49,
27
+ "n": 75,
28
+ "o": 4,
29
+ "p": 29,
30
+ "q": 14,
31
+ "r": 15,
32
+ "s": 42,
33
+ "t": 52,
34
+ "u": 46,
35
+ "v": 2,
36
+ "w": 78,
37
+ "x": 71,
38
+ "y": 80,
39
+ "|": 3,
40
+ "à": 60,
41
+ "á": 79,
42
+ "â": 55,
43
+ "ã": 38,
44
+ "è": 35,
45
+ "é": 25,
46
+ "ê": 66,
47
+ "ì": 85,
48
+ "í": 51,
49
+ "ò": 94,
50
+ "ó": 93,
51
+ "ô": 39,
52
+ "ù": 34,
53
+ "ú": 17,
54
+ "ý": 45,
55
+ "ă": 62,
56
+ "đ": 82,
57
+ "ơ": 8,
58
+ "ư": 58,
59
+ "ạ": 59,
60
+ "ả": 92,
61
+ "ấ": 73,
62
+ "ầ": 31,
63
+ "ẩ": 70,
64
+ "ẫ": 40,
65
+ "ậ": 88,
66
+ "ắ": 0,
67
+ "ằ": 28,
68
+ "ẵ": 54,
69
+ "ặ": 68,
70
+ "ẹ": 27,
71
+ "ẻ": 7,
72
+ "ẽ": 83,
73
+ "ế": 65,
74
+ "ề": 22,
75
+ "ể": 56,
76
+ "ệ": 41,
77
+ "ỉ": 64,
78
+ "ị": 13,
79
+ "ọ": 50,
80
+ "ỏ": 12,
81
+ "ố": 11,
82
+ "ồ": 23,
83
+ "ổ": 91,
84
+ "ỗ": 81,
85
+ "ộ": 84,
86
+ "ớ": 67,
87
+ "ờ": 9,
88
+ "ở": 53,
89
+ "ỡ": 26,
90
+ "ợ": 36,
91
+ "ụ": 33,
92
+ "ủ": 18,
93
+ "ứ": 90,
94
+ "ừ": 30,
95
+ "ử": 43,
96
+ "ữ": 57,
97
+ "ự": 72,
98
+ "ỳ": 77
99
+ }