Update README.md
Browse files
README.md
CHANGED
|
@@ -12,8 +12,8 @@ license: apache-2.0
|
|
| 12 |
<tr>
|
| 13 |
<th>Category</th>
|
| 14 |
<th>Metric</th>
|
| 15 |
-
<th>meta-llama/Llama-4-
|
| 16 |
-
<th>nm-testing/Llama-4-
|
| 17 |
<th>Recovery (%)</th>
|
| 18 |
</tr>
|
| 19 |
</thead>
|
|
@@ -22,89 +22,89 @@ license: apache-2.0
|
|
| 22 |
<tr>
|
| 23 |
<td rowspan="7"><b>OpenLLM V1</b></td>
|
| 24 |
<td>ARC-Challenge (Acc-Norm, 25-shot)</td>
|
| 25 |
-
<td>
|
| 26 |
-
<td>
|
| 27 |
-
<td>
|
| 28 |
</tr>
|
| 29 |
<tr>
|
| 30 |
<td>GSM8K (Strict-Match, 5-shot)</td>
|
| 31 |
-
<td>
|
| 32 |
-
<td>
|
| 33 |
-
<td>
|
| 34 |
</tr>
|
| 35 |
<tr>
|
| 36 |
<td>HellaSwag (Acc-Norm, 10-shot)</td>
|
| 37 |
-
<td>
|
| 38 |
-
<td>
|
| 39 |
-
<td>
|
| 40 |
</tr>
|
| 41 |
<tr>
|
| 42 |
<td>MMLU (Acc, 5-shot)</td>
|
| 43 |
-
<td>
|
| 44 |
-
<td>
|
| 45 |
-
<td>100.
|
| 46 |
</tr>
|
| 47 |
<tr>
|
| 48 |
<td>TruthfulQA (MC2, 0-shot)</td>
|
| 49 |
-
<td>
|
| 50 |
-
<td>
|
| 51 |
-
<td>100.
|
| 52 |
</tr>
|
| 53 |
<tr>
|
| 54 |
<td>Winogrande (Acc, 5-shot)</td>
|
| 55 |
-
<td>
|
| 56 |
-
<td>
|
| 57 |
-
<td>99.
|
| 58 |
</tr>
|
| 59 |
<tr>
|
| 60 |
<td><b>Average Score</b></td>
|
| 61 |
-
<td><b>
|
| 62 |
-
<td><b>
|
| 63 |
-
<td><b>99.
|
| 64 |
</tr>
|
| 65 |
<!-- OpenLLM Leaderboard V2 -->
|
| 66 |
<tr>
|
| 67 |
<td rowspan="7"><b>OpenLLM V2</b></td>
|
| 68 |
<td>IFEval (Inst Level Strict Acc, 0-shot)</td>
|
| 69 |
-
<td>89.
|
| 70 |
<td>90.89</td>
|
| 71 |
-
<td>
|
| 72 |
</tr>
|
| 73 |
<tr>
|
| 74 |
<td>BBH (Acc-Norm, 3-shot)</td>
|
| 75 |
-
<td>
|
| 76 |
-
<td>
|
| 77 |
-
<td>100.
|
| 78 |
</tr>
|
| 79 |
<tr>
|
| 80 |
<td>Math-Hard (Exact-Match, 4-shot)</td>
|
| 81 |
-
<td>
|
| 82 |
-
<td>
|
| 83 |
-
<td>100.
|
| 84 |
</tr>
|
| 85 |
<tr>
|
| 86 |
<td>GPQA (Acc-Norm, 0-shot)</td>
|
| 87 |
-
<td>
|
| 88 |
-
<td>
|
| 89 |
-
<td>
|
| 90 |
</tr>
|
| 91 |
<tr>
|
| 92 |
<td>MUSR (Acc-Norm, 0-shot)</td>
|
| 93 |
-
<td>
|
| 94 |
-
<td>
|
| 95 |
-
<td>
|
| 96 |
</tr>
|
| 97 |
<tr>
|
| 98 |
<td>MMLU-Pro (Acc, 5-shot)</td>
|
| 99 |
-
<td>
|
| 100 |
-
<td>
|
| 101 |
-
<td>99.
|
| 102 |
</tr>
|
| 103 |
<tr>
|
| 104 |
<td><b>Average Score</b></td>
|
| 105 |
-
<td><b>
|
| 106 |
-
<td><b>
|
| 107 |
-
<td><b>100.
|
| 108 |
</tr>
|
| 109 |
<td rowspan="4" ><strong>Coding</strong>
|
| 110 |
</td>
|
|
|
|
| 12 |
<tr>
|
| 13 |
<th>Category</th>
|
| 14 |
<th>Metric</th>
|
| 15 |
+
<th>meta-llama/Llama-4-Maverick-17B-128E-Instruct</th>
|
| 16 |
+
<th>nm-testing/Llama-4-Maverick-17B-128E-Instruct-block-FP8</th>
|
| 17 |
<th>Recovery (%)</th>
|
| 18 |
</tr>
|
| 19 |
</thead>
|
|
|
|
| 22 |
<tr>
|
| 23 |
<td rowspan="7"><b>OpenLLM V1</b></td>
|
| 24 |
<td>ARC-Challenge (Acc-Norm, 25-shot)</td>
|
| 25 |
+
<td>73.38</td>
|
| 26 |
+
<td>73.38</td>
|
| 27 |
+
<td>100.00</td>
|
| 28 |
</tr>
|
| 29 |
<tr>
|
| 30 |
<td>GSM8K (Strict-Match, 5-shot)</td>
|
| 31 |
+
<td>93.03</td>
|
| 32 |
+
<td>92.72</td>
|
| 33 |
+
<td>99.67</td>
|
| 34 |
</tr>
|
| 35 |
<tr>
|
| 36 |
<td>HellaSwag (Acc-Norm, 10-shot)</td>
|
| 37 |
+
<td>87.39</td>
|
| 38 |
+
<td>87.33</td>
|
| 39 |
+
<td>99.93</td>
|
| 40 |
</tr>
|
| 41 |
<tr>
|
| 42 |
<td>MMLU (Acc, 5-shot)</td>
|
| 43 |
+
<td>86.03</td>
|
| 44 |
+
<td>86.15</td>
|
| 45 |
+
<td>100.13</td>
|
| 46 |
</tr>
|
| 47 |
<tr>
|
| 48 |
<td>TruthfulQA (MC2, 0-shot)</td>
|
| 49 |
+
<td>62.76</td>
|
| 50 |
+
<td>62.90</td>
|
| 51 |
+
<td>100.23</td>
|
| 52 |
</tr>
|
| 53 |
<tr>
|
| 54 |
<td>Winogrande (Acc, 5-shot)</td>
|
| 55 |
+
<td>79.56</td>
|
| 56 |
+
<td>79.40</td>
|
| 57 |
+
<td>99.80</td>
|
| 58 |
</tr>
|
| 59 |
<tr>
|
| 60 |
<td><b>Average Score</b></td>
|
| 61 |
+
<td><b>80.36</b></td>
|
| 62 |
+
<td><b>80.31</b></td>
|
| 63 |
+
<td><b>99.94</b></td>
|
| 64 |
</tr>
|
| 65 |
<!-- OpenLLM Leaderboard V2 -->
|
| 66 |
<tr>
|
| 67 |
<td rowspan="7"><b>OpenLLM V2</b></td>
|
| 68 |
<td>IFEval (Inst Level Strict Acc, 0-shot)</td>
|
| 69 |
+
<td>89.93</td>
|
| 70 |
<td>90.89</td>
|
| 71 |
+
<td>101.07</td>
|
| 72 |
</tr>
|
| 73 |
<tr>
|
| 74 |
<td>BBH (Acc-Norm, 3-shot)</td>
|
| 75 |
+
<td>70.53</td>
|
| 76 |
+
<td>71.03</td>
|
| 77 |
+
<td>100.71</td>
|
| 78 |
</tr>
|
| 79 |
<tr>
|
| 80 |
<td>Math-Hard (Exact-Match, 4-shot)</td>
|
| 81 |
+
<td>64.73</td>
|
| 82 |
+
<td>65.26</td>
|
| 83 |
+
<td>100.82</td>
|
| 84 |
</tr>
|
| 85 |
<tr>
|
| 86 |
<td>GPQA (Acc-Norm, 0-shot)</td>
|
| 87 |
+
<td>31.29</td>
|
| 88 |
+
<td>30.54</td>
|
| 89 |
+
<td>97.59</td>
|
| 90 |
</tr>
|
| 91 |
<tr>
|
| 92 |
<td>MUSR (Acc-Norm, 0-shot)</td>
|
| 93 |
+
<td>46.56</td>
|
| 94 |
+
<td>46.03</td>
|
| 95 |
+
<td>98.86</td>
|
| 96 |
</tr>
|
| 97 |
<tr>
|
| 98 |
<td>MMLU-Pro (Acc, 5-shot)</td>
|
| 99 |
+
<td>64.11</td>
|
| 100 |
+
<td>63.95</td>
|
| 101 |
+
<td>99.75</td>
|
| 102 |
</tr>
|
| 103 |
<tr>
|
| 104 |
<td><b>Average Score</b></td>
|
| 105 |
+
<td><b>61.19</b></td>
|
| 106 |
+
<td><b>61.28</b></td>
|
| 107 |
+
<td><b>100.15</b></td>
|
| 108 |
</tr>
|
| 109 |
<td rowspan="4" ><strong>Coding</strong>
|
| 110 |
</td>
|