Commit
·
4ad3073
1
Parent(s):
db7ad2f
Update README.md
Browse files
README.md
CHANGED
|
@@ -45,10 +45,12 @@ PROMPT 格式: [chatml](https://github.com/openai/openai-python/blob/main/chatml
|
|
| 45 |
|
| 46 |
当前的 MMLU: 53.48
|
| 47 |
|
|
|
|
|
|
|
| 48 |
```
|
| 49 |
MMLU - stem ACC: 46.40 Humanities ACC: 47.61 other ACC: 61.31 social ACC: 61.78 AVERAGE ACC:53.48
|
| 50 |
|
| 51 |
CEval (val) - STEM acc: 45.28 Social Science acc: 66.19 Humanities acc: 58.76 Other acc: 54.62 Hard acc:28.64 AVERAGE acc:54.13
|
| 52 |
```
|
| 53 |
|
| 54 |
-
问题:相比原本的 Qwen-7B-Chat 的 MMLU 分数 53.90 和 CEval (val) 分数 54.
|
|
|
|
| 45 |
|
| 46 |
当前的 MMLU: 53.48
|
| 47 |
|
| 48 |
+
当前的 CEval (val): 54.13
|
| 49 |
+
|
| 50 |
```
|
| 51 |
MMLU - stem ACC: 46.40 Humanities ACC: 47.61 other ACC: 61.31 social ACC: 61.78 AVERAGE ACC:53.48
|
| 52 |
|
| 53 |
CEval (val) - STEM acc: 45.28 Social Science acc: 66.19 Humanities acc: 58.76 Other acc: 54.62 Hard acc:28.64 AVERAGE acc:54.13
|
| 54 |
```
|
| 55 |
|
| 56 |
+
问题:相比原本的 Qwen-7B-Chat 的 MMLU 分数 53.90 和 CEval (val) 分数 54.18,由于不够充分的重新对齐,分数都略有下降(MMLU -0.42, CEval -0.05)。
|