manjugeorge commited on
Commit
11187bc
1 Parent(s): b127f30

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +6 -0
  2. tokenizer_config.json +10 -0
  3. vocab.json +31 -0
special_tokens_map.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "eos_token": "</s>",
4
+ "pad_token": "[PAD]",
5
+ "unk_token": "[UNK]"
6
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "<s>",
3
+ "do_lower_case": false,
4
+ "eos_token": "</s>",
5
+ "pad_token": "[PAD]",
6
+ "replace_word_delimiter_char": " ",
7
+ "tokenizer_class": "Wav2Vec2CTCTokenizer",
8
+ "unk_token": "[UNK]",
9
+ "word_delimiter_token": "|"
10
+ }
vocab.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "[PAD]": 28,
3
+ "[UNK]": 27,
4
+ "|": 23,
5
+ "ആ": 4,
6
+ "ഏ": 0,
7
+ "ഒ": 20,
8
+ "ക": 15,
9
+ "ജ": 9,
10
+ "ണ": 18,
11
+ "ത": 7,
12
+ "ദ": 16,
13
+ "ന": 17,
14
+ "പ": 5,
15
+ "മ": 3,
16
+ "യ": 12,
17
+ "ര": 2,
18
+ "ല": 8,
19
+ "ള": 26,
20
+ "ഴ": 11,
21
+ "വ": 19,
22
+ "സ": 25,
23
+ "ാ": 6,
24
+ "ി": 14,
25
+ "ു": 10,
26
+ "െ": 21,
27
+ "േ": 22,
28
+ "്": 13,
29
+ "ർ": 24,
30
+ "ൽ": 1
31
+ }