danielhanchen commited on
Commit
75cdc99
1 Parent(s): 6e54b93

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +7 -1
  2. tokenizer.json +1 -10
  3. tokenizer_config.json +2 -10
special_tokens_map.json CHANGED
@@ -13,7 +13,13 @@
13
  "rstrip": false,
14
  "single_word": false
15
  },
16
- "pad_token": "<pad>",
 
 
 
 
 
 
17
  "unk_token": {
18
  "content": "<unk>",
19
  "lstrip": false,
 
13
  "rstrip": false,
14
  "single_word": false
15
  },
16
+ "pad_token": {
17
+ "content": "[control_14]",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
  "unk_token": {
24
  "content": "<unk>",
25
  "lstrip": false,
tokenizer.json CHANGED
@@ -154,7 +154,7 @@
154
  "lstrip": false,
155
  "rstrip": false,
156
  "normalized": false,
157
- "special": false
158
  },
159
  {
160
  "id": 17,
@@ -6941,15 +6941,6 @@
6941
  "rstrip": false,
6942
  "normalized": false,
6943
  "special": false
6944
- },
6945
- {
6946
- "id": 32768,
6947
- "content": "<pad>",
6948
- "single_word": false,
6949
- "lstrip": false,
6950
- "rstrip": false,
6951
- "normalized": false,
6952
- "special": true
6953
  }
6954
  ],
6955
  "normalizer": {
 
154
  "lstrip": false,
155
  "rstrip": false,
156
  "normalized": false,
157
+ "special": true
158
  },
159
  {
160
  "id": 17,
 
6941
  "rstrip": false,
6942
  "normalized": false,
6943
  "special": false
 
 
 
 
 
 
 
 
 
6944
  }
6945
  ],
6946
  "normalizer": {
tokenizer_config.json CHANGED
@@ -137,7 +137,7 @@
137
  "normalized": false,
138
  "rstrip": false,
139
  "single_word": false,
140
- "special": false
141
  },
142
  "17": {
143
  "content": "[control_15]",
@@ -6170,14 +6170,6 @@
6170
  "rstrip": false,
6171
  "single_word": false,
6172
  "special": false
6173
- },
6174
- "32768": {
6175
- "content": "<pad>",
6176
- "lstrip": false,
6177
- "normalized": false,
6178
- "rstrip": false,
6179
- "single_word": false,
6180
- "special": true
6181
  }
6182
  },
6183
  "bos_token": "<s>",
@@ -6186,7 +6178,7 @@
6186
  "eos_token": "</s>",
6187
  "legacy": true,
6188
  "model_max_length": 1000000000000000019884624838656,
6189
- "pad_token": "<pad>",
6190
  "padding_side": "left",
6191
  "sp_model_kwargs": {},
6192
  "spaces_between_special_tokens": false,
 
137
  "normalized": false,
138
  "rstrip": false,
139
  "single_word": false,
140
+ "special": true
141
  },
142
  "17": {
143
  "content": "[control_15]",
 
6170
  "rstrip": false,
6171
  "single_word": false,
6172
  "special": false
 
 
 
 
 
 
 
 
6173
  }
6174
  },
6175
  "bos_token": "<s>",
 
6178
  "eos_token": "</s>",
6179
  "legacy": true,
6180
  "model_max_length": 1000000000000000019884624838656,
6181
+ "pad_token": "[control_14]",
6182
  "padding_side": "left",
6183
  "sp_model_kwargs": {},
6184
  "spaces_between_special_tokens": false,