PatrickHaller commited on
Commit
d8d7eca
·
1 Parent(s): d59f795

Upload tokenizer

Browse files
Files changed (3) hide show
  1. special_tokens_map.json +5 -0
  2. tokenizer_config.json +15 -0
  3. vocab.json +0 -0
special_tokens_map.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "eos_token": "\n",
3
+ "pad_token": "\n",
4
+ "unk_token": "<unk>"
5
+ }
tokenizer_config.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "auto_map": {
3
+ "AutoTokenizer": [
4
+ "tokenization_ngme.NGMETokenizer",
5
+ null
6
+ ]
7
+ },
8
+ "clean_up_tokenization_spaces": true,
9
+ "eos_token": "\n",
10
+ "model_max_length": 1000000000000000019884624838656,
11
+ "ngram": null,
12
+ "pad_token": "\n",
13
+ "tokenizer_class": "NGMETokenizer",
14
+ "unk_token": "<unk>"
15
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff