Fixed tokenizer_config.json
Browse filesFixed tokenizer_config.json's inconsistent BOS, UNK token contents.
- tokenizer_config.json +2 -2
tokenizer_config.json
CHANGED
@@ -3,7 +3,7 @@
|
|
3 |
"add_prefix_space": false,
|
4 |
"bos_token": {
|
5 |
"__type": "AddedToken",
|
6 |
-
"content": "
|
7 |
"lstrip": false,
|
8 |
"normalized": true,
|
9 |
"rstrip": false,
|
@@ -25,7 +25,7 @@
|
|
25 |
"tokenizer_class": "GPT2Tokenizer",
|
26 |
"unk_token": {
|
27 |
"__type": "AddedToken",
|
28 |
-
"content": "
|
29 |
"lstrip": false,
|
30 |
"normalized": true,
|
31 |
"rstrip": false,
|
|
|
3 |
"add_prefix_space": false,
|
4 |
"bos_token": {
|
5 |
"__type": "AddedToken",
|
6 |
+
"content": "<||bos||>",
|
7 |
"lstrip": false,
|
8 |
"normalized": true,
|
9 |
"rstrip": false,
|
|
|
25 |
"tokenizer_class": "GPT2Tokenizer",
|
26 |
"unk_token": {
|
27 |
"__type": "AddedToken",
|
28 |
+
"content": "<||unk||>",
|
29 |
"lstrip": false,
|
30 |
"normalized": true,
|
31 |
"rstrip": false,
|