Update README.md

README.md

@@ -32,7 +32,7 @@ base_model: meta-llama/Meta-Llama-3.1-8B-Instruct
 - **Model Developers:** Neural Magic
 
 Quantized version of [Meta-Llama-3.1-8B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-8B-Instruct).
-It achieves scores within 2.3% of the scores of the unquantized model for MMLU, ARC-Challenge, GSM-8k, Hellaswag and Winogrande, and within 7.4% for TruthfulQA.
+It achieves scores within 3.1% of the scores of the unquantized model for MMLU, ARC-Challenge, GSM-8k, Hellaswag and Winogrande, and TruthfulQA.
 
 ### Model Optimizations
 
@@ -200,9 +200,9 @@ This version of the lm-evaluation-harness includes versions of MMLU, ARC-Challen
    </td>
    <td>78.06
    </td>
-   <td>76.40
+   <td>77.98
    </td>
-   <td>97.9%
+   <td>99.9%
    </td>
   </tr>
   <tr>
@@ -210,9 +210,9 @@ This version of the lm-evaluation-harness includes versions of MMLU, ARC-Challen
    </td>
    <td>54.48
    </td>
-   <td>50.46
+   <td>52.81
    </td>
-   <td>92.6%
+   <td>96.9%
    </td>
   </tr>
   <tr>
@@ -220,9 +220,9 @@ This version of the lm-evaluation-harness includes versions of MMLU, ARC-Challen
    </td>
    <td><strong>74.25</strong>
    </td>
-   <td><strong>72.58</strong>
+   <td><strong>73.24</strong>
    </td>
-   <td><strong>97.7%</strong>
+   <td><strong>98.6%</strong>
    </td>
   </tr>
 </table>