YassineBenlaria commited on
Commit
8447aac
1 Parent(s): 5aecab8

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