Update README.md

Update benchmarks

README.md

@@ -126,10 +126,10 @@ We report in the following table our internal pipeline benchmarks:
         <tr>
             <td rowspan="4">Reasoning</td>
             <td>Arc Challenge (25-shot)</td>
-            <td>53.4</td>
-            <td>59.0</td>
-            <td>63.7</td>
-            <td>59.6</td>
+            <td>58.2</td>
+            <td>63.2</td>
+            <td>67.5</td>
+            <td>63.1</td>
         </tr>
         <tr>
             <td>GPQA (0-shot)</td>
@@ -155,17 +155,17 @@ We report in the following table our internal pipeline benchmarks:
         <tr>
             <td rowspan="4">CommonSense Understanding</td>
             <td>PIQA (0-shot)</td>
-            <td>80.3</td>
-            <td>78.7</td>
-            <td>81.4</td>
-            <td>77.7</td>
+            <td>81.2</td>
+            <td>79.9</td>
+            <td>79.1</td>
+            <td>82.9</td>
         </tr>
         <tr>
             <td>SciQ (0-shot)</td>
-            <td>96.3</td>
-            <td>96.6</td>
-            <td>97.2</td>
-            <td>95.3</td>
+            <td>94.6</td>
+            <td>95.2</td>
+            <td>92.4</td>
+            <td>97.1</td>
         </tr>
         <tr>
             <td>Winogrande (0-shot)</td>
@@ -176,10 +176,10 @@ We report in the following table our internal pipeline benchmarks:
         </tr>
         <tr>
             <td>OpenbookQA (0-shot)</td>
-            <td>33.4</td>
-            <td>33.6</td>
-            <td>34.0</td>
-            <td>31.4</td>
+            <td>44.8</td>
+            <td>47.0</td>
+            <td>43.8</td>
+            <td>47.2</td>
         </tr>
     </tbody>
 </table>