Akashsingh commited on
Commit
da4593c
1 Parent(s): 965fc3f

Upload tokenizer

Browse files
added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "<unk>": 65
3
+ }
special_tokens_map.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "pad_token": "త",
3
+ "unk_token": "<unk>"
4
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_blank": true,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "త",
6
+ "lstrip": false,
7
+ "normalized": false,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "65": {
13
+ "content": "<unk>",
14
+ "lstrip": false,
15
+ "normalized": false,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ }
20
+ },
21
+ "clean_up_tokenization_spaces": true,
22
+ "is_uroman": false,
23
+ "language": "tel",
24
+ "model_max_length": 1000000000000000019884624838656,
25
+ "normalize": true,
26
+ "pad_token": "త",
27
+ "phonemize": false,
28
+ "tokenizer_class": "VitsTokenizer",
29
+ "unk_token": "<unk>",
30
+ "verbose": false
31
+ }
vocab.json ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ " ": 19,
3
+ "'": 8,
4
+ "-": 53,
5
+ "6": 60,
6
+ "_": 45,
7
+ "ం": 61,
8
+ "ః": 28,
9
+ "అ": 42,
10
+ "ఆ": 63,
11
+ "ఇ": 62,
12
+ "ఈ": 55,
13
+ "ఉ": 51,
14
+ "ఊ": 59,
15
+ "ఎ": 38,
16
+ "ఏ": 37,
17
+ "ఐ": 9,
18
+ "ఒ": 21,
19
+ "ఓ": 40,
20
+ "ఔ": 13,
21
+ "క": 20,
22
+ "ఖ": 30,
23
+ "గ": 10,
24
+ "ఘ": 7,
25
+ "చ": 46,
26
+ "ఛ": 57,
27
+ "జ": 12,
28
+ "ఞ": 1,
29
+ "ట": 56,
30
+ "ఠ": 29,
31
+ "డ": 16,
32
+ "ఢ": 52,
33
+ "ణ": 33,
34
+ "త": 0,
35
+ "థ": 3,
36
+ "ద": 49,
37
+ "ధ": 24,
38
+ "న": 5,
39
+ "ప": 27,
40
+ "ఫ": 34,
41
+ "బ": 54,
42
+ "భ": 2,
43
+ "మ": 17,
44
+ "య": 41,
45
+ "ర": 4,
46
+ "ల": 50,
47
+ "ళ": 58,
48
+ "వ": 23,
49
+ "శ": 25,
50
+ "ష": 31,
51
+ "స": 18,
52
+ "హ": 14,
53
+ "ా": 15,
54
+ "ి": 35,
55
+ "ీ": 43,
56
+ "ు": 47,
57
+ "ూ": 22,
58
+ "ృ": 39,
59
+ "ె": 44,
60
+ "ే": 11,
61
+ "ై": 64,
62
+ "ొ": 36,
63
+ "ో": 6,
64
+ "ౌ": 26,
65
+ "్": 48,
66
+ "–": 32
67
+ }