Update README.md

README.md

@@ -75,8 +75,8 @@ The training utilized the Falcon architecture and was fine-tuned using a Causal-
 
 - **Training regime:** CausalLM fine-tuning
 - **Batch size:** 1
-- **Learning rate:** 2e-4
-- **Epochs:** 5
+- **Learning rate:** 2e-5
+- **Epochs:** 3
 
 ## Evaluation