itqop commited on
Commit
cc6ad2a
·
verified ·
1 Parent(s): 9ceb0de

Upload tokenizer

Browse files
special_tokens_map.json CHANGED
@@ -1,4 +1,20 @@
1
  {
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  "boi_token": "<start_of_image>",
3
  "bos_token": {
4
  "content": "<bos>",
 
1
  {
2
+ "additional_special_tokens": [
3
+ {
4
+ "content": "<start_of_sum>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false
9
+ },
10
+ {
11
+ "content": "<end_of_sum>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ }
17
+ ],
18
  "boi_token": "<start_of_image>",
19
  "bos_token": {
20
  "content": "<bos>",
tokenizer.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b84ad421f379485dfd23db043f807e1d1cb27c5f1e645f9151067ab2b98d55b0
3
  size 33384948
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62b641e9926d7fe6186c86bc16101b5c2a543a9fe71ca8b73a6d3a999b048dbe
3
  size 33384948
tokenizer_config.json CHANGED
@@ -51325,20 +51325,24 @@
51325
  "262145": {
51326
  "content": "<start_of_sum>",
51327
  "lstrip": false,
51328
- "normalized": true,
51329
  "rstrip": false,
51330
  "single_word": false,
51331
- "special": false
51332
  },
51333
  "262146": {
51334
  "content": "<end_of_sum>",
51335
  "lstrip": false,
51336
- "normalized": true,
51337
  "rstrip": false,
51338
  "single_word": false,
51339
- "special": false
51340
  }
51341
  },
 
 
 
 
51342
  "boi_token": "<start_of_image>",
51343
  "bos_token": "<bos>",
51344
  "clean_up_tokenization_spaces": false,
 
51325
  "262145": {
51326
  "content": "<start_of_sum>",
51327
  "lstrip": false,
51328
+ "normalized": false,
51329
  "rstrip": false,
51330
  "single_word": false,
51331
+ "special": true
51332
  },
51333
  "262146": {
51334
  "content": "<end_of_sum>",
51335
  "lstrip": false,
51336
+ "normalized": false,
51337
  "rstrip": false,
51338
  "single_word": false,
51339
+ "special": true
51340
  }
51341
  },
51342
+ "additional_special_tokens": [
51343
+ "<start_of_sum>",
51344
+ "<end_of_sum>"
51345
+ ],
51346
  "boi_token": "<start_of_image>",
51347
  "bos_token": "<bos>",
51348
  "clean_up_tokenization_spaces": false,