khaoulaoub commited on
Commit
8cf6454
1 Parent(s): 7f82b48

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 +44 -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 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "*": 18,
3
+ "[PAD]": 41,
4
+ "[UNK]": 40,
5
+ "d": 32,
6
+ "|": 29,
7
+ "ء": 23,
8
+ "آ": 14,
9
+ "أ": 35,
10
+ "ؤ": 34,
11
+ "إ": 9,
12
+ "ئ": 2,
13
+ "ا": 8,
14
+ "ب": 7,
15
+ "ة": 13,
16
+ "ت": 33,
17
+ "ث": 37,
18
+ "ج": 20,
19
+ "ح": 3,
20
+ "خ": 11,
21
+ "د": 30,
22
+ "ذ": 1,
23
+ "ر": 27,
24
+ "ز": 10,
25
+ "س": 17,
26
+ "ش": 36,
27
+ "ص": 4,
28
+ "ض": 26,
29
+ "ط": 6,
30
+ "ظ": 21,
31
+ "ع": 28,
32
+ "غ": 38,
33
+ "ف": 25,
34
+ "ق": 5,
35
+ "ك": 22,
36
+ "ل": 24,
37
+ "م": 15,
38
+ "ن": 19,
39
+ "ه": 39,
40
+ "و": 0,
41
+ "ى": 31,
42
+ "ي": 16,
43
+ "ڭ": 12
44
+ }