hippopotam commited on
Commit
7521451
1 Parent(s): 4c5b155

Upload tokenizer

Browse files
Files changed (2) hide show
  1. added_tokens.json +2 -2
  2. vocab.json +36 -42
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 43,
3
- "<s>": 42
4
  }
 
1
  {
2
+ "</s>": 37,
3
+ "<s>": 36
4
  }
vocab.json CHANGED
@@ -1,44 +1,38 @@
1
  {
2
- " ": 0,
3
- "(": 1,
4
- ")": 2,
5
- "[PAD]": 41,
6
- "[UNK]": 40,
7
- "a": 3,
8
- "b": 4,
9
- "c": 5,
10
- "d": 6,
11
- "e": 7,
12
- "f": 8,
13
- "g": 9,
14
- "h": 10,
15
- "i": 11,
16
- "j": 12,
17
- "k": 13,
18
- "l": 14,
19
- "m": 15,
20
- "n": 16,
21
- "o": 17,
22
- "p": 18,
23
- "q": 19,
24
- "r": 20,
25
- "s": 21,
26
- "t": 22,
27
- "u": 23,
28
- "v": 24,
29
- "w": 25,
30
- "x": 26,
31
- "y": 27,
32
- "z": 28,
33
- "ç": 29,
34
- "é": 30,
35
- "ë": 31,
36
- "ö": 32,
37
- "ü": 33,
38
- "ğ": 34,
39
- "ı": 35,
40
- "ş": 36,
41
- "̇": 37,
42
- "’": 38,
43
- "…": 39
44
  }
 
1
  {
2
+ "[PAD]": 35,
3
+ "[UNK]": 34,
4
+ "a": 1,
5
+ "b": 2,
6
+ "c": 3,
7
+ "d": 4,
8
+ "e": 5,
9
+ "f": 6,
10
+ "g": 7,
11
+ "h": 8,
12
+ "i": 9,
13
+ "j": 10,
14
+ "k": 11,
15
+ "l": 12,
16
+ "m": 13,
17
+ "n": 14,
18
+ "o": 15,
19
+ "p": 16,
20
+ "q": 17,
21
+ "r": 18,
22
+ "s": 19,
23
+ "t": 20,
24
+ "u": 21,
25
+ "v": 22,
26
+ "w": 23,
27
+ "x": 24,
28
+ "y": 25,
29
+ "z": 26,
30
+ "|": 0,
31
+ "ç": 27,
32
+ "ö": 28,
33
+ "ü": 29,
34
+ "ğ": 30,
35
+ "ı": 31,
36
+ "ş": 32,
37
+ "̇": 33
 
 
 
 
 
 
38
  }