Commit
·
f498ce7
1
Parent(s):
298cb2c
use <|im_start> as bos and <|im_end|> as eos tokens
Browse files- tokenizer_config.json +2 -2
tokenizer_config.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"bos_token": {
|
| 3 |
"__type": "AddedToken",
|
| 4 |
-
"content": "
|
| 5 |
"lstrip": false,
|
| 6 |
"normalized": false,
|
| 7 |
"rstrip": false,
|
|
@@ -10,7 +10,7 @@
|
|
| 10 |
"clean_up_tokenization_spaces": false,
|
| 11 |
"eos_token": {
|
| 12 |
"__type": "AddedToken",
|
| 13 |
-
"content": "
|
| 14 |
"lstrip": false,
|
| 15 |
"normalized": false,
|
| 16 |
"rstrip": false,
|
|
|
|
| 1 |
{
|
| 2 |
"bos_token": {
|
| 3 |
"__type": "AddedToken",
|
| 4 |
+
"content": "<|im_start|>",
|
| 5 |
"lstrip": false,
|
| 6 |
"normalized": false,
|
| 7 |
"rstrip": false,
|
|
|
|
| 10 |
"clean_up_tokenization_spaces": false,
|
| 11 |
"eos_token": {
|
| 12 |
"__type": "AddedToken",
|
| 13 |
+
"content": "<|im_end|>",
|
| 14 |
"lstrip": false,
|
| 15 |
"normalized": false,
|
| 16 |
"rstrip": false,
|