DylanonWic
commited on
Commit
•
a04e3ce
1
Parent(s):
cbfd4eb
Upload 4 files
Browse files- added_tokens.json +4 -0
- special_tokens_map.json +22 -0
- tokenizer_config.json +14 -0
- vocab.json +72 -0
added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</s>": 71,
|
3 |
+
"<s>": 70
|
4 |
+
}
|
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 |
+
}
|
tokenizer_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"do_lower_case": false,
|
4 |
+
"eos_token": "</s>",
|
5 |
+
"model_max_length": 1000000000000000019884624838656,
|
6 |
+
"name_or_path": "airesearch/wav2vec2-large-xlsr-53-th",
|
7 |
+
"num_proc": 4,
|
8 |
+
"pad_token": "[PAD]",
|
9 |
+
"replace_word_delimiter_char": " ",
|
10 |
+
"special_tokens_map_file": "/root/.cache/huggingface/hub/models--airesearch--wav2vec2-large-xlsr-53-th/snapshots/3155938c549b23eee16b1d4b55dcb161b7fe4bcf/special_tokens_map.json",
|
11 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
12 |
+
"unk_token": "[UNK]",
|
13 |
+
"word_delimiter_token": "|"
|
14 |
+
}
|
vocab.json
ADDED
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[PAD]": 69,
|
3 |
+
"[UNK]": 68,
|
4 |
+
"|": 42,
|
5 |
+
"ก": 2,
|
6 |
+
"ข": 58,
|
7 |
+
"ฃ": 66,
|
8 |
+
"ค": 46,
|
9 |
+
"ฆ": 44,
|
10 |
+
"ง": 3,
|
11 |
+
"จ": 49,
|
12 |
+
"ฉ": 53,
|
13 |
+
"ช": 63,
|
14 |
+
"ซ": 9,
|
15 |
+
"ฌ": 8,
|
16 |
+
"ญ": 14,
|
17 |
+
"ฎ": 52,
|
18 |
+
"ฏ": 48,
|
19 |
+
"ฐ": 34,
|
20 |
+
"ฑ": 0,
|
21 |
+
"ฒ": 4,
|
22 |
+
"ณ": 39,
|
23 |
+
"ด": 10,
|
24 |
+
"ต": 35,
|
25 |
+
"ถ": 36,
|
26 |
+
"ท": 21,
|
27 |
+
"ธ": 65,
|
28 |
+
"น": 62,
|
29 |
+
"บ": 33,
|
30 |
+
"ป": 31,
|
31 |
+
"ผ": 57,
|
32 |
+
"ฝ": 30,
|
33 |
+
"พ": 43,
|
34 |
+
"ฟ": 38,
|
35 |
+
"ภ": 23,
|
36 |
+
"ม": 51,
|
37 |
+
"ย": 24,
|
38 |
+
"ร": 41,
|
39 |
+
"ฤ": 17,
|
40 |
+
"ล": 26,
|
41 |
+
"ว": 47,
|
42 |
+
"ศ": 37,
|
43 |
+
"ษ": 55,
|
44 |
+
"ส": 67,
|
45 |
+
"ห": 40,
|
46 |
+
"ฬ": 20,
|
47 |
+
"อ": 19,
|
48 |
+
"ฮ": 29,
|
49 |
+
"ฯ": 11,
|
50 |
+
"ะ": 5,
|
51 |
+
"ั": 45,
|
52 |
+
"า": 16,
|
53 |
+
"ำ": 56,
|
54 |
+
"ิ": 64,
|
55 |
+
"ี": 32,
|
56 |
+
"ึ": 13,
|
57 |
+
"ื": 61,
|
58 |
+
"ุ": 27,
|
59 |
+
"ู": 60,
|
60 |
+
"เ": 28,
|
61 |
+
"แ": 50,
|
62 |
+
"โ": 22,
|
63 |
+
"ใ": 12,
|
64 |
+
"ไ": 59,
|
65 |
+
"ๅ": 1,
|
66 |
+
"็": 25,
|
67 |
+
"่": 15,
|
68 |
+
"้": 7,
|
69 |
+
"๊": 6,
|
70 |
+
"๋": 18,
|
71 |
+
"์": 54
|
72 |
+
}
|