khaoulaoub commited on
Commit
6faf679
1 Parent(s): ce7133a

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>": 36,
3
- "<s>": 35
4
  }
 
1
  {
2
  "</s>": 36,
3
+ "[PAD]": 35
4
  }
tokenizer_config.json CHANGED
@@ -1,14 +1,6 @@
1
  {
2
  "added_tokens_decoder": {
3
- "33": {
4
- "content": "[UNK]",
5
- "lstrip": true,
6
- "normalized": false,
7
- "rstrip": true,
8
- "single_word": false,
9
- "special": false
10
- },
11
- "34": {
12
  "content": "[PAD]",
13
  "lstrip": true,
14
  "normalized": false,
@@ -16,14 +8,6 @@
16
  "single_word": false,
17
  "special": false
18
  },
19
- "35": {
20
- "content": "<s>",
21
- "lstrip": false,
22
- "normalized": true,
23
- "rstrip": false,
24
- "single_word": false,
25
- "special": true
26
- },
27
  "36": {
28
  "content": "</s>",
29
  "lstrip": false,
 
1
  {
2
  "added_tokens_decoder": {
3
+ "35": {
 
 
 
 
 
 
 
 
4
  "content": "[PAD]",
5
  "lstrip": true,
6
  "normalized": false,
 
8
  "single_word": false,
9
  "special": false
10
  },
 
 
 
 
 
 
 
 
11
  "36": {
12
  "content": "</s>",
13
  "lstrip": false,
vocab.json CHANGED
@@ -1,6 +1,6 @@
1
  {
2
- "[PAD]": 34,
3
- "[UNK]": 33,
4
  "|": 16,
5
  "ء": 11,
6
  "ا": 26,
 
1
  {
2
+ "[PAD]": 35,
3
+ "[UNK]": 35,
4
  "|": 16,
5
  "ء": 11,
6
  "ا": 26,