sanchit-gandhi HF staff commited on
Commit
3e44724
1 Parent(s): 2521fa9

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +4 -0
  2. tokenizer_config.json +12 -0
  3. vocab.json +40 -0
special_tokens_map.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "pad_token": "|",
3
+ "unk_token": "<unk>"
4
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_blank": true,
3
+ "clean_up_tokenization_spaces": true,
4
+ "is_uroman": false,
5
+ "language": "peg",
6
+ "model_max_length": 1000000000000000019884624838656,
7
+ "normalize": true,
8
+ "pad_token": "|",
9
+ "phonemize": false,
10
+ "tokenizer_class": "VitsTokenizer",
11
+ "unk_token": "<unk>"
12
+ }
vocab.json ADDED
@@ -0,0 +1,40 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ " ": 37,
3
+ "|": 0,
4
+ "ଂ": 15,
5
+ "ଃ": 36,
6
+ "ଅ": 34,
7
+ "ଆ": 17,
8
+ "ଇ": 16,
9
+ "ଉ": 33,
10
+ "ଏ": 30,
11
+ "ଓ": 35,
12
+ "କ": 8,
13
+ "ଗ": 13,
14
+ "ଙ": 26,
15
+ "ଚ": 23,
16
+ "ଜ": 10,
17
+ "ଞ": 32,
18
+ "ଟ": 29,
19
+ "ଡ": 25,
20
+ "ଣ": 28,
21
+ "ତ": 7,
22
+ "ଦ": 19,
23
+ "ନ": 6,
24
+ "ପ": 20,
25
+ "ବ": 21,
26
+ "ମ": 14,
27
+ "ର": 5,
28
+ "ଲ": 24,
29
+ "ସ": 18,
30
+ "ହ": 9,
31
+ "଼": 22,
32
+ "ା": 1,
33
+ "ି": 3,
34
+ "ୁ": 12,
35
+ "େ": 4,
36
+ "୍": 2,
37
+ "ୟ": 27,
38
+ "ୱ": 11,
39
+ "‍": 31
40
+ }