Update app.py

app.py

@@ -23,12 +23,13 @@ def get_response(user_message, history, verbosity):
     # Adjust the temperature based on the verbosity level
     temperature = 0.5 if verbosity == "Balanced" else 0.3 if verbosity == "Concise" else 0.7
 
-    # Make the API call
+    # Make the API call with a custom stop sequence
     response = client.chat.completions.create(
         model='ambrosfitz/llama-3-history',
         messages=history_openai_format,
         temperature=temperature,
-        max_tokens=512
+        max_tokens=150,
+        stop=["<|eot_id|>"]  # Add a custom stop sequence
     )
 
     # Access the text response