Update README.md

README.md

@@ -155,54 +155,84 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
    <td><strong>Recovery</strong>
    </td>
   </tr>
+  <tr>
+   <td>MMLU (5-shot)
+   </td>
+   <td>62.98
+   </td>
+   <td>62.95
+   </td>
+   <td>100.0%
+   </td>
+  </tr>
   <tr>
    <td>MMLU-cot (0-shot)
    </td>
-   <td>55.22
+   <td>65.40
    </td>
-   <td>55.28
+   <td>65.23
    </td>
-   <td>100.1%
+   <td>99.7%
    </td>
   </tr>
   <tr>
    <td>ARC Challenge (0-shot)
    </td>
-   <td>77.39
+   <td>77.13
    </td>
-   <td>76.62
+   <td>76.71
    </td>
-   <td>99.0%
+   <td>99.4%
    </td>
   </tr>
   <tr>
    <td>GSM-8K-cot (8-shot, strict-match)
    </td>
-   <td>77.56
+   <td>77.94
    </td>
-   <td>76.12
+   <td>76.72
    </td>
-   <td>98.1%
+   <td>98.4%
    </td>
   </tr>
   <tr>
    <td>Winogrande (5-shot)
    </td>
-   <td>70.2
+   <td>71.11
+   </td>
+   <td>71.11
+   </td>
+   <td>100.0%
+   </td>
+  </tr>
+  <tr>
+   <td>Hellaswag (10-shot)
+   </td>
+   <td>73.62
    </td>
-   <td>69.3
+   <td>73.54
    </td>
-   <td>98.7%
+   <td>99.9%
    </td>
   </tr>
+  <tr>
+   <td>TruthfulQA (0-shot, mc2)
+   </td>
+   <td>51.47
+   </td>
+   <td>51.06
+   </td>
+   <td>99.2%
+   </td>
+  </tr>  
   <tr>
    <td><strong>Average</strong>
    </td>
-   <td><strong>70.09</strong>
+   <td><strong>68.52</strong>
    </td>
-   <td><strong>69.33</strong>
+   <td><strong>68.19</strong>
    </td>
-   <td><strong>98.92%</strong>
+   <td><strong>99.5%</strong>
    </td>
   </tr>
 </table>
@@ -212,11 +242,23 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
 The results were obtained using the following commands:
 
 
-#### MMLU-cot
+#### MMLU
 ```
 lm_eval \
   --model vllm \
-  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=False,max_model_len=4096,tensor_parallel_size=1 \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,max_model_len=3850,max_gen_toks=10,tensor_parallel_size=1 \
+  --tasks mmlu_llama_3.1_instruct \
+  --fewshot_as_multiturn \
+  --apply_chat_template \
+  --num_fewshot 5 \
+  --batch_size auto
+```
+
+#### MMLU-CoT
+```
+lm_eval \
+  --model vllm \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,max_model_len=4064,max_gen_toks=1024,tensor_parallel_size=1 \
   --tasks mmlu_cot_0shot_llama_3.1_instruct \
   --apply_chat_template \
   --num_fewshot 0 \
@@ -227,7 +269,7 @@ lm_eval \
 ```
 lm_eval \
   --model vllm \
-  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=False,max_model_len=4096,tensor_parallel_size=1 \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,max_model_len=3940,max_gen_toks=100,tensor_parallel_size=1 \
   --tasks arc_challenge_llama_3.1_instruct \
   --apply_chat_template \
   --num_fewshot 0 \
@@ -238,21 +280,40 @@ lm_eval \
 ```
 lm_eval \
   --model vllm \
-  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=False,max_model_len=4096,tensor_parallel_size=1 \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,max_model_len=4096,max_gen_toks=1024,tensor_parallel_size=1 \
   --tasks gsm8k_cot_llama_3.1_instruct \
-  --apply_chat_template \
   --fewshot_as_multiturn \
+  --apply_chat_template \
   --num_fewshot 8 \
   --batch_size auto
 ```
 
+#### Hellaswag
+```
+lm_eval \
+  --model vllm \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=True,max_model_len=4096,tensor_parallel_size=1 \
+  --tasks hellaswag \
+  --num_fewshot 10 \
+  --batch_size auto
+```
+
 #### Winogrande
 ```
 lm_eval \
   --model vllm \
-  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=False,max_model_len=4096,tensor_parallel_size=1 \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=True,max_model_len=4096,tensor_parallel_size=1 \
   --tasks winogrande \
   --num_fewshot 5 \
   --batch_size auto
 ```
 
+#### TruthfulQA
+```
+lm_eval \
+  --model vllm \
+  --model_args pretrained="neuralmagic/Llama-3.2-3B-Instruct-FP8-dynamic",dtype=auto,add_bos_token=True,max_model_len=4096,tensor_parallel_size=1 \
+  --tasks truthfulqa \
+  --num_fewshot 0 \
+  --batch_size auto
+```