khederwaaOne commited on
Commit
b93da2d
·
verified ·
1 Parent(s): 7abe53f

Upload tokenizer

Browse files
Files changed (3) hide show
  1. added_tokens.json +2 -2
  2. tokenizer_config.json +4 -4
  3. vocab.json +50 -39
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 40,
3
- "<s>": 39
4
  }
 
1
  {
2
+ "</s>": 51,
3
+ "<s>": 50
4
  }
tokenizer_config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "37": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
@@ -8,7 +8,7 @@
8
  "single_word": false,
9
  "special": false
10
  },
11
- "38": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,7 +16,7 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "39": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
@@ -24,7 +24,7 @@
24
  "single_word": false,
25
  "special": true
26
  },
27
- "40": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "48": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
11
+ "49": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
 
16
  "single_word": false,
17
  "special": false
18
  },
19
+ "50": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
 
24
  "single_word": false,
25
  "special": true
26
  },
27
+ "51": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
vocab.json CHANGED
@@ -1,41 +1,52 @@
1
  {
2
- "0": 1,
3
- "1": 2,
4
- "2": 3,
5
- "3": 4,
6
- "4": 5,
7
- "5": 6,
8
- "6": 7,
9
- "7": 8,
10
- "8": 9,
11
- "9": 10,
12
- "[PAD]": 38,
13
- "[UNK]": 37,
14
- "a": 11,
15
- "b": 12,
16
- "c": 13,
17
- "d": 14,
18
- "e": 15,
19
- "f": 16,
20
- "g": 17,
21
- "h": 18,
22
- "i": 19,
23
- "j": 20,
24
- "k": 21,
25
- "l": 22,
26
- "m": 23,
27
- "n": 24,
28
- "o": 25,
29
- "p": 26,
30
- "q": 27,
31
- "r": 28,
32
- "s": 29,
33
- "t": 30,
34
- "u": 31,
35
- "v": 32,
36
- "w": 33,
37
- "x": 34,
38
- "y": 35,
39
- "z": 36,
40
- "|": 0
 
 
 
 
 
 
 
 
 
 
 
41
  }
 
1
  {
2
+ "!": 1,
3
+ "\"": 2,
4
+ "%": 3,
5
+ "'": 4,
6
+ ",": 5,
7
+ "-": 6,
8
+ ".": 7,
9
+ "0": 8,
10
+ "1": 9,
11
+ "2": 10,
12
+ "3": 11,
13
+ "4": 12,
14
+ "5": 13,
15
+ "6": 14,
16
+ "7": 15,
17
+ "8": 16,
18
+ "9": 17,
19
+ ":": 18,
20
+ ";": 19,
21
+ "[PAD]": 49,
22
+ "[UNK]": 48,
23
+ "a": 20,
24
+ "b": 21,
25
+ "c": 22,
26
+ "d": 23,
27
+ "e": 24,
28
+ "f": 25,
29
+ "g": 26,
30
+ "h": 27,
31
+ "i": 28,
32
+ "j": 29,
33
+ "k": 30,
34
+ "l": 31,
35
+ "m": 32,
36
+ "n": 33,
37
+ "o": 34,
38
+ "p": 35,
39
+ "q": 36,
40
+ "r": 37,
41
+ "s": 38,
42
+ "t": 39,
43
+ "u": 40,
44
+ "v": 41,
45
+ "w": 42,
46
+ "x": 43,
47
+ "y": 44,
48
+ "z": 45,
49
+ "|": 0,
50
+ "‘": 46,
51
+ "”": 47
52
  }