changed temperature to 0.5 for more focused response

app.py

@@ -44,7 +44,7 @@ def llama_generation(input_text: str,
     # llama generation looks for the numeric vectors not the tensors so there is no need for **input_ids rather just input_ids
     output_ids = llama_model.generate(input_ids=input_ids,
                                       max_new_tokens=256,
-                                      temperature=0.7,
+                                      temperature=0.5,
                                       top_p=0.9,
                                       repetition_penalty=2.0)