Update app.py

app.py

@@ -668,9 +668,9 @@ def generate_answer(message, choice, retrieval_mode, selected_model):
                 logging.debug(f"Generated LM-2 prompt: {prompt}")
 
                 response = selected_model(prompt, **{
-                    "max_new_tokens": 400,
+                    "max_new_tokens": 150,
                     "return_full_text": True,
-                    "temperature": 0.7,
+                    "temperature": 0.3,
                     "do_sample": True,
                 })