YassineBenlaria commited on
Commit
57b7737
1 Parent(s): c797e48

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +6 -0
  2. tokenizer_config.json +13 -0
  3. vocab.json +41 -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 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "[PAD]": 38,
3
+ "[UNK]": 37,
4
+ "|": 4,
5
+ "ء": 23,
6
+ "ؤ": 7,
7
+ "ئ": 19,
8
+ "ا": 17,
9
+ "ب": 20,
10
+ "ة": 27,
11
+ "ت": 14,
12
+ "ث": 3,
13
+ "ج": 18,
14
+ "ح": 15,
15
+ "خ": 2,
16
+ "د": 1,
17
+ "ذ": 13,
18
+ "ر": 5,
19
+ "ز": 35,
20
+ "س": 31,
21
+ "ش": 24,
22
+ "ص": 34,
23
+ "ض": 33,
24
+ "ط": 30,
25
+ "ظ": 9,
26
+ "ع": 26,
27
+ "غ": 25,
28
+ "ف": 8,
29
+ "ق": 12,
30
+ "ك": 22,
31
+ "ل": 21,
32
+ "م": 10,
33
+ "ن": 36,
34
+ "ه": 29,
35
+ "و": 0,
36
+ "ى": 28,
37
+ "ي": 11,
38
+ "ڤ": 6,
39
+ "ک": 32,
40
+ "ی": 16
41
+ }