Commit
·
d5ab5c7
1
Parent(s):
f20ecec
Upload 5 files
Browse files- added_tokens.json +4 -0
- preprocessor_config.json +9 -0
- special_tokens_map.json +22 -0
- vocab.json +93 -0
added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</s>": 92,
|
3 |
+
"<s>": 91
|
4 |
+
}
|
preprocessor_config.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"do_normalize": true,
|
3 |
+
"feature_extractor_type": "Wav2Vec2FeatureExtractor",
|
4 |
+
"feature_size": 1,
|
5 |
+
"padding_side": "right",
|
6 |
+
"padding_value": 0,
|
7 |
+
"return_attention_mask": true,
|
8 |
+
"sampling_rate": 16000
|
9 |
+
}
|
special_tokens_map.json
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
{
|
4 |
+
"content": "<s>",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": true,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false
|
9 |
+
},
|
10 |
+
{
|
11 |
+
"content": "</s>",
|
12 |
+
"lstrip": false,
|
13 |
+
"normalized": true,
|
14 |
+
"rstrip": false,
|
15 |
+
"single_word": false
|
16 |
+
}
|
17 |
+
],
|
18 |
+
"bos_token": "<s>",
|
19 |
+
"eos_token": "</s>",
|
20 |
+
"pad_token": "[PAD]",
|
21 |
+
"unk_token": "[UNK]"
|
22 |
+
}
|
vocab.json
ADDED
@@ -0,0 +1,93 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[PAD]": 90,
|
3 |
+
"[UNK]": 89,
|
4 |
+
"a": 86,
|
5 |
+
"b": 81,
|
6 |
+
"c": 82,
|
7 |
+
"d": 19,
|
8 |
+
"e": 71,
|
9 |
+
"g": 4,
|
10 |
+
"h": 21,
|
11 |
+
"i": 36,
|
12 |
+
"k": 49,
|
13 |
+
"l": 58,
|
14 |
+
"m": 54,
|
15 |
+
"n": 46,
|
16 |
+
"o": 80,
|
17 |
+
"p": 76,
|
18 |
+
"q": 12,
|
19 |
+
"r": 57,
|
20 |
+
"s": 48,
|
21 |
+
"t": 45,
|
22 |
+
"u": 85,
|
23 |
+
"v": 73,
|
24 |
+
"x": 22,
|
25 |
+
"y": 43,
|
26 |
+
"|": 69,
|
27 |
+
"à": 3,
|
28 |
+
"á": 42,
|
29 |
+
"â": 50,
|
30 |
+
"ã": 31,
|
31 |
+
"è": 52,
|
32 |
+
"é": 30,
|
33 |
+
"ê": 37,
|
34 |
+
"ì": 32,
|
35 |
+
"í": 61,
|
36 |
+
"ò": 13,
|
37 |
+
"ó": 40,
|
38 |
+
"ô": 67,
|
39 |
+
"õ": 20,
|
40 |
+
"ù": 35,
|
41 |
+
"ú": 65,
|
42 |
+
"ý": 47,
|
43 |
+
"ă": 75,
|
44 |
+
"đ": 39,
|
45 |
+
"ĩ": 79,
|
46 |
+
"ũ": 1,
|
47 |
+
"ơ": 29,
|
48 |
+
"ư": 87,
|
49 |
+
"ạ": 15,
|
50 |
+
"ả": 5,
|
51 |
+
"ấ": 53,
|
52 |
+
"ầ": 16,
|
53 |
+
"ẩ": 64,
|
54 |
+
"ẫ": 18,
|
55 |
+
"ậ": 33,
|
56 |
+
"ắ": 55,
|
57 |
+
"ằ": 68,
|
58 |
+
"ẳ": 28,
|
59 |
+
"ẵ": 14,
|
60 |
+
"ặ": 77,
|
61 |
+
"ẹ": 72,
|
62 |
+
"ẻ": 78,
|
63 |
+
"ẽ": 17,
|
64 |
+
"ế": 23,
|
65 |
+
"ề": 41,
|
66 |
+
"ể": 24,
|
67 |
+
"ễ": 10,
|
68 |
+
"ệ": 59,
|
69 |
+
"ỉ": 2,
|
70 |
+
"ị": 38,
|
71 |
+
"ọ": 83,
|
72 |
+
"ỏ": 60,
|
73 |
+
"ố": 44,
|
74 |
+
"ồ": 70,
|
75 |
+
"ổ": 25,
|
76 |
+
"ỗ": 66,
|
77 |
+
"ộ": 84,
|
78 |
+
"ớ": 8,
|
79 |
+
"ờ": 62,
|
80 |
+
"ở": 11,
|
81 |
+
"ỡ": 6,
|
82 |
+
"ợ": 9,
|
83 |
+
"ụ": 88,
|
84 |
+
"ủ": 7,
|
85 |
+
"ứ": 0,
|
86 |
+
"ừ": 56,
|
87 |
+
"ử": 63,
|
88 |
+
"ữ": 26,
|
89 |
+
"ự": 27,
|
90 |
+
"ỳ": 34,
|
91 |
+
"ỷ": 51,
|
92 |
+
"ỹ": 74
|
93 |
+
}
|