nm-research commited on
Commit
bfc0121
·
verified ·
1 Parent(s): 7fc5898

Add reasoning evals

Browse files
Files changed (1) hide show
  1. README.md +25 -0
README.md CHANGED
@@ -168,6 +168,31 @@ lm_eval \
168
  </thead>
169
  <tbody>
170
  <tr>
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
171
  <td rowspan="7"><b>OpenLLM V1</b></td>
172
  <td>ARC-Challenge (Acc-Norm, 25-shot)</td>
173
  <td>64.59</td>
 
168
  </thead>
169
  <tbody>
170
  <tr>
171
+ <td rowspan="4"><b>Reasoning</b></td>
172
+ <td>AIME 2024 (pass@1)</td>
173
+ <td>69.75</td>
174
+ <td>68.17</td>
175
+ <td>97.73%</td>
176
+ </tr>
177
+ <tr>
178
+ <td>MATH-500 (pass@1)</td>
179
+ <td>95.09</td>
180
+ <td>94.98</td>
181
+ <td>99.88%</td>
182
+ </tr>
183
+ <tr>
184
+ <td>GPQA Diamond (pass@1)</td>
185
+ <td>64.05</td>
186
+ <td>64.75</td>
187
+ <td>101.09%</td>
188
+ </tr>
189
+ <tr>
190
+ <td><b>Average Score</b></td>
191
+ <td><b>76.3</b></td>
192
+ <td><b>75.97</b></td>
193
+ <td><b>99.57%</b></td>
194
+ </tr>
195
+ <tr>
196
  <td rowspan="7"><b>OpenLLM V1</b></td>
197
  <td>ARC-Challenge (Acc-Norm, 25-shot)</td>
198
  <td>64.59</td>