SaulLu commited on
Commit
df95ac1
1 Parent(s): 16b5869

add tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +15 -0
  2. tokenizer.json +0 -0
  3. tokenizer_config.json +19 -0
special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "eos_token": "[SEP]",
4
+ "unk_token": "<unk>",
5
+ "sep_token": "[SEP]",
6
+ "pad_token": "<pad>",
7
+ "cls_token": "[CLS]",
8
+ "mask_token": {
9
+ "content": "[MASK]",
10
+ "single_word": false,
11
+ "lstrip": true,
12
+ "rstrip": false,
13
+ "normalized": false
14
+ }
15
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,19 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "eos_token": "[SEP]",
4
+ "unk_token": "<unk>",
5
+ "sep_token": "[SEP]",
6
+ "cls_token": "[CLS]",
7
+ "pad_token": "<pad>",
8
+ "mask_token": {
9
+ "content": "[MASK]",
10
+ "single_word": false,
11
+ "lstrip": true,
12
+ "rstrip": false,
13
+ "normalized": false,
14
+ "__type": "AddedToken"
15
+ },
16
+ "model_max_length": 512,
17
+ "special_tokens_map_file": null,
18
+ "name_or_path": "SaulLu/bengali-tokenizer-v2"
19
+ }