Update README.md

open llm leaderboard update

README.md

@@ -17,11 +17,11 @@ Platypus-30B is an instruction fine-tuned model based on the LLaMA-30B transform
 
 | Metric                | Value |
 |-----------------------|-------|
-| MMLU (5-shot)         | 64.2  |
-| ARC (25-shot)         | 64.6  |
-| HellaSwag (10-shot)   | 84.3  |
-| TruthfulQA (0-shot)   | 45.8  |
-| Avg.                  | 64.7  | 
+| MMLU (5-shot)         | 64.1  |
+| ARC (25-shot)         | 57.6  |
+| HellaSwag (10-shot)   | 81.9  |
+| TruthfulQA (0-shot)   | 45.3  |
+| Avg.                  | 62.2  | 
 
 We use state-of-the-art [Language Model Evaluation Harness](https://github.com/EleutherAI/lm-evaluation-harness) to run the benchmark tests above.