Update README.md

README.md

@@ -31,8 +31,9 @@ base_model: meta-llama/Meta-Llama-3.1-405B-Instruct
 - **License(s):** Llama3.1
 - **Model Developers:** Neural Magic
 
-Quantized version of [Meta-Llama-3.1-405B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-405B-Instruct).
-It achieves scores within 1% of the scores of the unquantized model for MMLU, ARC-Challenge, GSM-8k, Hellaswag, Winogrande, and TruthfulQA.
+This model is a quantized version of [Meta-Llama-3.1-405B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3.1-405B-Instruct).
+It was evaluated on a several tasks to assess the its quality in comparison to the unquatized model, including multiple-choice, math reasoning, and open-ended text generation.
+Meta-Llama-3.1-405B-Instruct-quantized.w4a16 achieves 98.7% recovery for the Arena-Hard evaluation, 100.0% for OpenLLM v1 (using Meta's prompting when available), 99.0% for OpenLLM v2, 98.0% for HumanEval pass@1, and 98.5% for HumanEval+ pass@1.
 
 ### Model Optimizations
 
@@ -128,9 +129,19 @@ model.save_pretrained("Meta-Llama-3.1-405B-Instruct-quantized.w4a16")
 
 ## Evaluation
 
-The model was evaluated on MMLU, ARC-Challenge, GSM-8K, Hellaswag, Winogrande and TruthfulQA.
-Evaluation was conducted using the Neural Magic fork of [lm-evaluation-harness](https://github.com/neuralmagic/lm-evaluation-harness/tree/llama_3.1_instruct) (branch llama_3.1_instruct) and the [vLLM](https://docs.vllm.ai/en/stable/) engine.
-This version of the lm-evaluation-harness includes versions of ARC-Challenge, GSM-8K, and MMLU that match the prompting style of [Meta-Llama-3.1-Instruct-evals](https://huggingface.co/datasets/meta-llama/Meta-Llama-3.1-405B-Instruct-evals).
+This model was evaluated on the well-known Arena-Hard, OpenLLM v1, OpenLLM v2, HumanEval, and HumanEval+ benchmarks.
+In all cases, model outputs were generated with the [vLLM](https://docs.vllm.ai/en/stable/) engine.
+
+Arena-Hard evaluations were conducted using the [Arena-Hard-Auto](https://github.com/lmarena/arena-hard-auto) repository.
+The model generated a single answer for each prompt form Arena-Hard, and each answer was judged twice by GPT-4.
+We report below the scores obtained in each judgement and the average.
+
+OpenLLM v1 and v2 evaluations were conducted using Neural Magic's fork of [lm-evaluation-harness](https://github.com/neuralmagic/lm-evaluation-harness/tree/llama_3.1_instruct) (branch llama_3.1_instruct).
+This version of the lm-evaluation-harness includes versions of MMLU, ARC-Challenge and GSM-8K that match the prompting style of [Meta-Llama-3.1-Instruct-evals](https://huggingface.co/datasets/meta-llama/Meta-Llama-3.1-405B-Instruct-evals) and a few fixes to OpenLLM v2 tasks.
+
+HumanEval and HumanEval+ evaluations were conducted using Neural Magic's fork of the [EvalPlus](https://github.com/neuralmagic/evalplus) repository.
+
+Detailed model outputs are available as HuggingFace datasets for [Arena-Hard](https://huggingface.co/datasets/neuralmagic/quantized-llama-3.1-arena-hard-evals), [OpenLLM v2](https://huggingface.co/datasets/neuralmagic/quantized-llama-3.1-leaderboard-v2-evals), and [HumanEval](https://huggingface.co/datasets/neuralmagic/quantized-llama-3.1-humaneval-evals).
 
 **Note:** Results have been updated after Meta modified the chat template.
 
@@ -148,12 +159,26 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
    <td><strong>Recovery</strong>
    </td>
   </tr>
+  <tr>
+   <td><strong>Arena Hard</strong>
+   </td>
+   <td>67.4 (67.3 / 67.5)
+   </td>
+   <td>66.5 (66.5 / 66.4)
+   </td>
+   <td>98.7%
+   </td>
+  </tr>
+  <tr>
+   <td><strong>OpenLLM v1</strong>
+   </td>
+  </tr>
   <tr>
    <td>MMLU (5-shot)
    </td>
-   <td>87.38
+   <td>87.4
    </td>
-   <td>87.22
+   <td>87.2
    </td>
    <td>99.8%
    </td>
@@ -161,9 +186,9 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
   <tr>
    <td>ARC Challenge (0-shot)
    </td>
-   <td>94.97
+   <td>95.0
    </td>
-   <td>95.31
+   <td>95.3
    </td>
    <td>100.4%
    </td>
@@ -171,9 +196,9 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
   <tr>
    <td>GSM-8K (CoT, 8-shot, strict-match)
    </td>
-   <td>96.44
+   <td>96.4
    </td>
-   <td>96.29
+   <td>96.3
    </td>
    <td>99.8%
    </td>
@@ -181,9 +206,9 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
   <tr>
    <td>Hellaswag (10-shot)
    </td> 
-   <td>88.33
+   <td>88.3
    </td>
-   <td>88.27
+   <td>88.3
    </td>
    <td>99.9%
    </td>
@@ -191,9 +216,9 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
   <tr>
    <td>Winogrande (5-shot)
    </td>
-   <td>87.21
+   <td>87.2
    </td>
-   <td>87.37
+   <td>87.4
    </td>
    <td>100.2%
    </td>
@@ -201,9 +226,9 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
   <tr>
    <td>TruthfulQA (0-shot)
    </td>
-   <td>64.64
+   <td>64.6
    </td>
-   <td>65.26
+   <td>65.3
    </td>
    <td>101.0%
    </td>
@@ -211,13 +236,111 @@ This version of the lm-evaluation-harness includes versions of ARC-Challenge, GS
   <tr>
    <td><strong>Average</strong>
    </td>
-   <td><strong>86.75</strong>
+   <td><strong>86.8</strong>
    </td>
-   <td><strong>86.76</strong>
+   <td><strong>86.8</strong>
    </td>
    <td><strong>100.0%</strong>
    </td>
   </tr>
+  <tr>
+   <td><strong>OpenLLM v2</strong>
+   </td>
+  </tr>
+  <tr>
+   <td>MMLU-Pro (5-shot)
+   </td>
+   <td>59.7
+   </td>
+   <td>59.4
+   </td>
+   <td>99.3%
+   </td>
+  </tr>
+  <tr>
+   <td>IFEval (0-shot)
+   </td>
+   <td>87.7
+   </td>
+   <td>88.0
+   </td>
+   <td>100.4%
+   </td>
+  </tr>
+  <tr>
+   <td>BBH (3-shot)
+   </td>
+   <td>67.0
+   </td>
+   <td>67.5
+   </td>
+   <td>100.7%
+   </td>
+  </tr>
+  <tr>
+   <td>Math-|v|-5 (4-shot)
+   </td>
+   <td>39.0
+   </td>
+   <td>37.6
+   </td>
+   <td>96.5%
+   </td>
+  </tr>
+  <tr>
+   <td>GPQA (0-shot)
+   </td>
+   <td>19.5
+   </td>
+   <td>17.5
+   </td>
+   <td>89.8%
+   </td>
+  </tr>
+  <tr>
+   <td>MuSR (0-shot)
+   </td>
+   <td>19.5
+   </td>
+   <td>19.4
+   </td>
+   <td>99.5%
+   </td>
+  </tr>
+  <tr>
+   <td><strong>Average</strong>
+   </td>
+   <td><strong>48.7</strong>
+   </td>
+   <td><strong>48.2</strong>
+   </td>
+   <td><strong>99.0%</strong>
+   </td>
+  </tr>
+  <tr>
+   <td><strong>Coding</strong>
+   </td>
+  </tr>
+  <tr>
+   <td>HumanEval pass@1
+   </td>
+   <td>86.8
+   </td>
+   <td>85.1
+   </td>
+   <td>98.0%
+   </td>
+  </tr>
+  <tr>
+   <td>HumanEval+ pass@1
+   </td>
+   <td>80.1
+   </td>
+   <td>78.9
+   </td>
+   <td>98.5%
+   </td>
+  </tr>
 </table>
 
 ### Reproduction
@@ -287,4 +410,39 @@ lm_eval \
   --tasks truthfulqa \
   --num_fewshot 0 \
   --batch_size auto
-```
+```
+
+#### OpenLLM v2
+```
+lm_eval \
+  --model vllm \
+  --model_args pretrained="neuralmagic/Meta-Llama-3.1-405B-Instruct-quantized.w4a16",dtype=auto,max_model_len=4096,tensor_parallel_size=8,enable_chunked_prefill=True \
+  --apply_chat_template \
+  --fewshot_as_multiturn \
+  --tasks leaderboard \
+  --batch_size auto
+```
+
+#### HumanEval and HumanEval+
+##### Generation
+```
+python3 codegen/generate.py \
+  --model neuralmagic/Meta-Llama-3.1-405B-Instruct-quantized.w4a16 \
+  --bs 16 \
+  --temperature 0.2 \
+  --n_samples 50 \
+  --root "." \
+  --dataset humaneval \
+  --tp 8
+```
+##### Sanitization
+```
+python3 evalplus/sanitize.py \
+  humaneval/neuralmagic--Meta-Llama-3.1-405B-Instruct-quantized.w4a16_vllm_temp_0.2
+```
+##### Evaluation
+```
+evalplus.evaluate \
+  --dataset humaneval \
+  --samples humaneval/neuralmagic--Meta-Llama-3.1-405B-Instruct-quantized.w4a16_vllm_temp_0.2-sanitized
+```