Update README.md

README.md

@@ -138,6 +138,7 @@ lm_eval \
   --tasks openllm \
   --batch_size auto
 ```
+Certain benchmarks for the full precision model are still being acquired. Average recovery is calculated only with metrics that both models have been evaluated on.
 
 ### Accuracy
 
@@ -156,71 +157,71 @@ lm_eval \
   <tr>
    <td>MMLU (5-shot)
    </td>
-   <td>82.21
+   <td>*
    </td>
-   <td>82.13
+   <td>88.34
    </td>
-   <td>99.90%
+   <td>*
    </td>
   </tr>
   <tr>
    <td>ARC Challenge (25-shot)
    </td>
-   <td>70.65
+   <td>73.38
    </td>
-   <td>70.31
+   <td>72.61
    </td>
-   <td>99.52%
+   <td>98.95%
    </td>
   </tr>
   <tr>
    <td>GSM-8K (5-shot, strict-match)
    </td>
-   <td>87.95
+   <td>95.07
    </td>
-   <td>88.40
+   <td>95.00
    </td>
-   <td>100.5%
+   <td>99.93%
    </td>
   </tr>
   <tr>
    <td>Hellaswag (10-shot)
    </td>
-   <td>86.33
+   <td>*
    </td>
-   <td>86.27
+   <td>88.34
    </td>
-   <td>99.93%
+   <td>*
    </td>
   </tr>
   <tr>
    <td>Winogrande (5-shot)
    </td>
-   <td>85.00
+   <td>87.21
    </td>
-   <td>85.00
+   <td>87.45
    </td>
-   <td>100.0%
+   <td>100.2%
    </td>
   </tr>
   <tr>
    <td>TruthfulQA (0-shot)
    </td>
-   <td>59.90
+   <td>*
    </td>
-   <td>60.01
+   <td>64.71
    </td>
-   <td>100.1%
+   <td>*
    </td>
   </tr>
   <tr>
    <td><strong>Average</strong>
    </td>
-   <td><strong>78.67</strong>
+   <td><strong>*</strong>
    </td>
-   <td><strong>78.69</strong>
+   <td><strong>82.38</strong>
    </td>
-   <td><strong>100.0%</strong>
+   <td><strong>98.95%</strong>
    </td>
   </tr>
 </table>