Update inference_fine_tune.py

inference_fine_tune.py

@@ -43,8 +43,8 @@ def generate_response(prompt: str, history):
         return
 
     input_tokens = torch.tensor(input_tokens).unsqueeze(0).to(device)
-    temperature = 0.7
-    top_k = 50
+    temperature = 0.5
+    top_k = 20
     generated_text = ""
     i = 0