Update README.md

README.md

@@ -148,71 +148,71 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge and
   <tr>
    <td>MMLU (5-shot)
    </td>
-   <td>67.94
+   <td>82.21
    </td>
-   <td>68.09
+   <td>82.12
    </td>
-   <td>100.2%
+   <td>99.9%
    </td>
   </tr>
   <tr>
    <td>ARC Challenge (0-shot)
    </td>
-   <td>83.19
+   <td>95.05
    </td>
-   <td>82.68
+   <td>93.60
    </td>
-   <td>99.4%
+   <td>98.5%
    </td>
   </tr>
   <tr>
    <td>GSM-8K (CoT, 8-shot, strict-match)
    </td>
-   <td>82.79
+   <td>93.10
    </td>
-   <td>82.64
+   <td>92.27
    </td>
-   <td>99.8%
+   <td>99.1%
    </td>
   </tr>
   <tr>
    <td>Hellaswag (10-shot)
    </td>
-   <td>80.01
+   <td>86.40
    </td>
-   <td>80.21
+   <td>86.11
    </td>
-   <td>100.3%
+   <td>99.7%
    </td>
   </tr>
   <tr>
    <td>Winogrande (5-shot)
    </td>
-   <td>77.90
+   <td>85.00
    </td>
-   <td>77.27
+   <td>84.14
    </td>
-   <td>99.2%
+   <td>99.0%
    </td>
   </tr>
   <tr>
    <td>TruthfulQA (0-shot, mc2)
    </td>
-   <td>54.04
+   <td>59.83
    </td>
-   <td>54.15
+   <td>58.90
    </td>
-   <td>100.2%
+   <td>98.5%
    </td>
   </tr>
   <tr>
    <td><strong>Average</strong>
    </td>
-   <td><strong>74.31</strong>
+   <td><strong>83.60</strong>
    </td>
-   <td><strong>74.17</strong>
+   <td><strong>82.66</strong>
    </td>
-   <td><strong>99.8%</strong>
+   <td><strong>99.1%</strong>
    </td>
   </tr>
 </table>