Commit
·
e5d934b
1
Parent(s):
ef22448
Upload tokenizer
Browse files- added_tokens.json +4 -0
- special_tokens_map.json +22 -0
- tokenizer_config.json +13 -0
- vocab.json +96 -0
added_tokens.json
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"</s>": 95,
|
3 |
+
"<s>": 94
|
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 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "<s>",
|
3 |
+
"clean_up_tokenization_spaces": true,
|
4 |
+
"do_lower_case": false,
|
5 |
+
"eos_token": "</s>",
|
6 |
+
"model_max_length": 1000000000000000019884624838656,
|
7 |
+
"pad_token": "[PAD]",
|
8 |
+
"replace_word_delimiter_char": " ",
|
9 |
+
"target_lang": null,
|
10 |
+
"tokenizer_class": "Wav2Vec2CTCTokenizer",
|
11 |
+
"unk_token": "[UNK]",
|
12 |
+
"word_delimiter_token": "|"
|
13 |
+
}
|
vocab.json
ADDED
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"0": 1,
|
3 |
+
"1": 2,
|
4 |
+
"2": 3,
|
5 |
+
"3": 4,
|
6 |
+
"4": 5,
|
7 |
+
"5": 6,
|
8 |
+
"6": 7,
|
9 |
+
"7": 8,
|
10 |
+
"8": 9,
|
11 |
+
"9": 10,
|
12 |
+
"C": 11,
|
13 |
+
"D": 12,
|
14 |
+
"I": 13,
|
15 |
+
"N": 14,
|
16 |
+
"O": 15,
|
17 |
+
"V": 16,
|
18 |
+
"[PAD]": 93,
|
19 |
+
"[UNK]": 92,
|
20 |
+
"n": 17,
|
21 |
+
"s": 18,
|
22 |
+
"|": 0,
|
23 |
+
"ಂ": 19,
|
24 |
+
"ಃ": 20,
|
25 |
+
"ಅ": 21,
|
26 |
+
"ಆ": 22,
|
27 |
+
"ಇ": 23,
|
28 |
+
"ಈ": 24,
|
29 |
+
"ಉ": 25,
|
30 |
+
"ಊ": 26,
|
31 |
+
"ಎ": 27,
|
32 |
+
"ಏ": 28,
|
33 |
+
"ಐ": 29,
|
34 |
+
"ಒ": 30,
|
35 |
+
"ಓ": 31,
|
36 |
+
"ಔ": 32,
|
37 |
+
"ಕ": 33,
|
38 |
+
"ಖ": 34,
|
39 |
+
"ಗ": 35,
|
40 |
+
"ಘ": 36,
|
41 |
+
"ಚ": 37,
|
42 |
+
"ಛ": 38,
|
43 |
+
"ಜ": 39,
|
44 |
+
"ಝ": 40,
|
45 |
+
"ಞ": 41,
|
46 |
+
"ಟ": 42,
|
47 |
+
"ಠ": 43,
|
48 |
+
"ಡ": 44,
|
49 |
+
"ಢ": 45,
|
50 |
+
"ಣ": 46,
|
51 |
+
"ತ": 47,
|
52 |
+
"ಥ": 48,
|
53 |
+
"ದ": 49,
|
54 |
+
"ಧ": 50,
|
55 |
+
"ನ": 51,
|
56 |
+
"ಪ": 52,
|
57 |
+
"ಫ": 53,
|
58 |
+
"ಬ": 54,
|
59 |
+
"ಭ": 55,
|
60 |
+
"ಮ": 56,
|
61 |
+
"ಯ": 57,
|
62 |
+
"ರ": 58,
|
63 |
+
"ಲ": 59,
|
64 |
+
"ಳ": 60,
|
65 |
+
"ವ": 61,
|
66 |
+
"ಶ": 62,
|
67 |
+
"ಷ": 63,
|
68 |
+
"ಸ": 64,
|
69 |
+
"ಹ": 65,
|
70 |
+
"ಾ": 66,
|
71 |
+
"ಿ": 67,
|
72 |
+
"ೀ": 68,
|
73 |
+
"ು": 69,
|
74 |
+
"ೂ": 70,
|
75 |
+
"ೃ": 71,
|
76 |
+
"ೆ": 72,
|
77 |
+
"ೇ": 73,
|
78 |
+
"ೈ": 74,
|
79 |
+
"ೊ": 75,
|
80 |
+
"ೋ": 76,
|
81 |
+
"ೌ": 77,
|
82 |
+
"್": 78,
|
83 |
+
"ೕ": 79,
|
84 |
+
"೦": 80,
|
85 |
+
"೧": 81,
|
86 |
+
"೨": 82,
|
87 |
+
"೩": 83,
|
88 |
+
"೫": 84,
|
89 |
+
"೭": 85,
|
90 |
+
"೮": 86,
|
91 |
+
"೯": 87,
|
92 |
+
"": 88,
|
93 |
+
"": 89,
|
94 |
+
"₹": 90,
|
95 |
+
"": 91
|
96 |
+
}
|