intellsion commited on
Commit
3ba493f
·
verified ·
1 Parent(s): 6661313

Upload tokenizer

Browse files
Files changed (2) hide show
  1. tokenizer_config.json +1 -1
  2. vocab.json +28 -28
tokenizer_config.json CHANGED
@@ -34,7 +34,7 @@
34
  }
35
  },
36
  "bos_token": "<s>",
37
- "clean_up_tokenization_spaces": true,
38
  "do_lower_case": false,
39
  "eos_token": "</s>",
40
  "model_max_length": 1000000000000000019884624838656,
 
34
  }
35
  },
36
  "bos_token": "<s>",
37
+ "clean_up_tokenization_spaces": false,
38
  "do_lower_case": false,
39
  "eos_token": "</s>",
40
  "model_max_length": 1000000000000000019884624838656,
vocab.json CHANGED
@@ -1,32 +1,32 @@
1
  {
2
- "'": 3,
3
  "[PAD]": 29,
4
  "[UNK]": 28,
5
- "a": 15,
6
- "b": 16,
7
- "c": 4,
8
- "d": 2,
9
- "e": 9,
10
- "f": 0,
11
- "g": 25,
12
- "h": 12,
13
- "i": 23,
14
- "j": 14,
15
- "k": 18,
16
- "l": 21,
17
- "m": 20,
18
- "n": 10,
19
- "o": 11,
20
- "p": 26,
21
- "q": 13,
22
- "r": 8,
23
- "s": 27,
24
- "t": 24,
25
- "u": 6,
26
- "v": 7,
27
- "w": 19,
28
- "x": 17,
29
- "y": 5,
30
- "z": 22,
31
- "|": 1
32
  }
 
1
  {
2
+ "'": 8,
3
  "[PAD]": 29,
4
  "[UNK]": 28,
5
+ "a": 25,
6
+ "b": 19,
7
+ "c": 2,
8
+ "d": 27,
9
+ "e": 14,
10
+ "f": 26,
11
+ "g": 21,
12
+ "h": 13,
13
+ "i": 12,
14
+ "j": 23,
15
+ "k": 9,
16
+ "l": 20,
17
+ "m": 5,
18
+ "n": 0,
19
+ "o": 7,
20
+ "p": 17,
21
+ "q": 10,
22
+ "r": 11,
23
+ "s": 16,
24
+ "t": 18,
25
+ "u": 15,
26
+ "v": 22,
27
+ "w": 1,
28
+ "x": 3,
29
+ "y": 6,
30
+ "z": 4,
31
+ "|": 24
32
  }