Gummybear05 commited on
Commit
3bdc67a
·
verified ·
1 Parent(s): ce45f6f

Upload tokenizer

Browse files
Files changed (3) hide show
  1. added_tokens.json +1 -1
  2. tokenizer_config.json +1 -17
  3. vocab.json +2 -2
added_tokens.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
  "</s>": 1204,
3
- "<s>": 1203
4
  }
 
1
  {
2
  "</s>": 1204,
3
+ "[PAD]": 1203
4
  }
tokenizer_config.json CHANGED
@@ -1,14 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "1201": {
4
- "content": "[UNK]",
5
- "lstrip": true,
6
- "normalized": false,
7
- "rstrip": true,
8
- "single_word": false,
9
- "special": false
10
- },
11
- "1202": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,14 +8,6 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "1203": {
20
- "content": "<s>",
21
- "lstrip": false,
22
- "normalized": false,
23
- "rstrip": false,
24
- "single_word": false,
25
- "special": true
26
- },
27
  "1204": {
28
  "content": "</s>",
29
  "lstrip": false,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "1203": {
 
 
 
 
 
 
 
 
4
  "content": "[PAD]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
 
 
 
 
 
 
 
 
11
  "1204": {
12
  "content": "</s>",
13
  "lstrip": false,
vocab.json CHANGED
@@ -7,8 +7,8 @@
7
  "O": 6,
8
  "P": 7,
9
  "S": 8,
10
- "[PAD]": 1202,
11
- "[UNK]": 1201,
12
  "|": 0,
13
  "가": 9,
14
  "각": 10,
 
7
  "O": 6,
8
  "P": 7,
9
  "S": 8,
10
+ "[PAD]": 1203,
11
+ "[UNK]": 1203,
12
  "|": 0,
13
  "가": 9,
14
  "각": 10,