Update README.md

README.md

@@ -17,13 +17,13 @@ language:
 ### Benchmark Metrics
 
 
-| Metric                | llama-2-7b-hf_open-platypus | meta-llama/Llama-2-7b-hf  (base) |
-|-----------------------|-------|-------|
-| Avg.                  | **54.35** | 54.32 |
-| ARC (25-shot)         | 51.45 | **53.07** |
-| HellaSwag (10-shot)   | **78.63** | 78.59 |
-| MMLU (5-shot)         | 43.6 | **46.87** |
-| TruthfulQA (0-shot)   | **43.71** | 38.76 |
+| Metric                | llama-2-7b-hf_open-platypus | garage-bAInd/Platypus2-7B| meta-llama/Llama-2-7b-hf  (base) |
+|-----------------------|-------|-------|-------|
+| Avg.                  | 54.35|**56.13** | 54.32 |
+| ARC (25-shot)         | 51.45 |**55.2**| 53.07 |
+| HellaSwag (10-shot)   | 78.63 |**78.84**| 78.59 |
+| MMLU (5-shot)         | 43.6 |**49.83**| 46.87 |
+| TruthfulQA (0-shot)   | **43.71** |40.64| 38.76 |
 
 
 We use state-of-the-art [Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness) to run the benchmark tests above, using the same version as the HuggingFace LLM Leaderboard. Please see below for detailed instructions on reproducing benchmark results.