Update README.md

README.md

@@ -38,8 +38,8 @@ The model was fine-tuned using Low-Rank Adaption (LoRA). For this model, LoRA wa
 # Evaluation
 | Dataset                                    | LLM-As-a-Judge | HellaSwag |
 |--------------------------------------------|----------------|-----------|
-| meta-llama/Llama-3.2-1B Instruct           | ~0.30          | ~0.45     |
-| meta-llama/Llama-3.2-1B                    | ~0.54          | ~0.44     |
+| meta-llama/Llama-3.2-1B Instruct           | ~0.58          | ~0.47     |
+| meta-llama/Llama-3.2-1B                    | ~0.60          | ~0.44     |
 | meta-llama/Llama-3.2-1B with Prompt Tuning | ~0.70          | ~0.41     |
 | MH-Persona                                 | ~0.87          | ~0.44     |