Update README.md
#1
by
karnakar
- opened
README.md
CHANGED
|
@@ -112,24 +112,24 @@ We report in the following table our internal pipeline benchmarks:
|
|
| 112 |
<tbody>
|
| 113 |
<tr>
|
| 114 |
<td>MMLU</td>
|
| 115 |
-
<td>43.
|
| 116 |
-
<td>42.
|
| 117 |
-
<td>43.
|
| 118 |
-
<td>
|
| 119 |
</tr>
|
| 120 |
<tr>
|
| 121 |
<td>MMLU-PRO</td>
|
| 122 |
-
<td>18.
|
| 123 |
-
<td>17.
|
| 124 |
-
<td>18.
|
| 125 |
-
<td>17.
|
| 126 |
</tr>
|
| 127 |
<tr>
|
| 128 |
<td>IFEval</td>
|
| 129 |
-
<td>54.
|
| 130 |
-
<td>51.
|
| 131 |
-
<td>56.
|
| 132 |
-
<td>51.
|
| 133 |
</tr>
|
| 134 |
</tbody>
|
| 135 |
</table>
|
|
|
|
| 112 |
<tbody>
|
| 113 |
<tr>
|
| 114 |
<td>MMLU</td>
|
| 115 |
+
<td>43.6</td>
|
| 116 |
+
<td>42.6</td>
|
| 117 |
+
<td>43.5</td>
|
| 118 |
+
<td>43.0</td>
|
| 119 |
</tr>
|
| 120 |
<tr>
|
| 121 |
<td>MMLU-PRO</td>
|
| 122 |
+
<td>18.5</td>
|
| 123 |
+
<td>17.7</td>
|
| 124 |
+
<td>18.5</td>
|
| 125 |
+
<td>17.3</td>
|
| 126 |
</tr>
|
| 127 |
<tr>
|
| 128 |
<td>IFEval</td>
|
| 129 |
+
<td>54.9</td>
|
| 130 |
+
<td>51.4</td>
|
| 131 |
+
<td>56.1</td>
|
| 132 |
+
<td>51.2</td>
|
| 133 |
</tr>
|
| 134 |
</tbody>
|
| 135 |
</table>
|