Commit
·
0315ff8
1
Parent(s):
b78fd6b
Update README.md
Browse files
README.md
CHANGED
|
@@ -40,9 +40,9 @@ The benchmarks and corresponding scores listed in the table below are taken dire
|
|
| 40 |
|TriviaQA|5-shot|69.9|54.97|-21.36%|
|
| 41 |
|HumanEval|0-shot|30.5|68.3|+123.93%|
|
| 42 |
|MBPP|3-shot|47.5|60.3|+26.95%|
|
| 43 |
-
|MATH|4-shot, maj@4|13.1|40.2*|+
|
| 44 |
|GSM8K|8-shot, maj@8|52.2|77.71|+48.87%|
|
| 45 |
-
||||**Average**|**+33.
|
| 46 |
|
| 47 |
\* : We report the 4-shot score instead of the 4-shot, maj@4.
|
| 48 |
|
|
|
|
| 40 |
|TriviaQA|5-shot|69.9|54.97|-21.36%|
|
| 41 |
|HumanEval|0-shot|30.5|68.3|+123.93%|
|
| 42 |
|MBPP|3-shot|47.5|60.3|+26.95%|
|
| 43 |
+
|MATH|4-shot, maj@4|13.1|40.2*|+206.87%|
|
| 44 |
|GSM8K|8-shot, maj@8|52.2|77.71|+48.87%|
|
| 45 |
+
||||**Average**|**+33.77%**|
|
| 46 |
|
| 47 |
\* : We report the 4-shot score instead of the 4-shot, maj@4.
|
| 48 |
|