Jai1432002 commited on
Commit
325f9c5
·
verified ·
1 Parent(s): f0b81ab

Upload tokenizer

Browse files
Files changed (3) hide show
  1. added_tokens.json +2 -2
  2. tokenizer_config.json +4 -4
  3. vocab.json +43 -42
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
- "</s>": 43,
3
- "<s>": 42
4
  }
 
1
  {
2
+ "</s>": 44,
3
+ "<s>": 43
4
  }
tokenizer_config.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "40": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
@@ -8,7 +8,7 @@
8
  "single_word": false,
9
  "special": false
10
  },
11
- "41": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,7 +16,7 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "42": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
@@ -24,7 +24,7 @@
24
  "single_word": false,
25
  "special": true
26
  },
27
- "43": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "41": {
4
  "content": "[UNK]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
11
+ "42": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
 
16
  "single_word": false,
17
  "special": false
18
  },
19
+ "43": {
20
  "content": "<s>",
21
  "lstrip": false,
22
  "normalized": false,
 
24
  "single_word": false,
25
  "special": true
26
  },
27
+ "44": {
28
  "content": "</s>",
29
  "lstrip": false,
30
  "normalized": false,
vocab.json CHANGED
@@ -1,44 +1,45 @@
1
  {
2
- "/": 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
- "[": 11,
13
- "[PAD]": 41,
14
- "[UNK]": 40,
15
- "]": 12,
16
- "_": 13,
17
- "a": 14,
18
- "b": 15,
19
- "c": 16,
20
- "d": 17,
21
- "e": 18,
22
- "f": 19,
23
- "g": 20,
24
- "h": 21,
25
- "i": 22,
26
- "j": 23,
27
- "k": 24,
28
- "l": 25,
29
- "m": 26,
30
- "n": 27,
31
- "o": 28,
32
- "p": 29,
33
- "q": 30,
34
- "r": 31,
35
- "s": 32,
36
- "t": 33,
37
- "u": 34,
38
- "v": 35,
39
- "w": 36,
40
- "x": 37,
41
- "y": 38,
42
- "z": 39,
43
- "|": 0
 
44
  }
 
1
  {
2
+ "\u001b": 0,
3
+ "/": 2,
4
+ "1": 3,
5
+ "2": 4,
6
+ "3": 5,
7
+ "4": 6,
8
+ "5": 7,
9
+ "6": 8,
10
+ "7": 9,
11
+ "8": 10,
12
+ "9": 11,
13
+ "[": 12,
14
+ "[PAD]": 42,
15
+ "[UNK]": 41,
16
+ "]": 13,
17
+ "_": 14,
18
+ "a": 15,
19
+ "b": 16,
20
+ "c": 17,
21
+ "d": 18,
22
+ "e": 19,
23
+ "f": 20,
24
+ "g": 21,
25
+ "h": 22,
26
+ "i": 23,
27
+ "j": 24,
28
+ "k": 25,
29
+ "l": 26,
30
+ "m": 27,
31
+ "n": 28,
32
+ "o": 29,
33
+ "p": 30,
34
+ "q": 31,
35
+ "r": 32,
36
+ "s": 33,
37
+ "t": 34,
38
+ "u": 35,
39
+ "v": 36,
40
+ "w": 37,
41
+ "x": 38,
42
+ "y": 39,
43
+ "z": 40,
44
+ "|": 1
45
  }