inject assistant placeholder at right time

app.py

@@ -138,9 +138,6 @@ def chat_response(user_msg, chat_history, system_prompt,
     else:
         debug = 'Web search disabled.'
 
-    # Prepare assistant placeholder
-    # history.append({'role': 'assistant', 'content': ''})
-
     try:
 
         # merge any fetched search results into the system prompt
@@ -187,6 +184,8 @@ def chat_response(user_msg, chat_history, system_prompt,
         gen_thread.start()
 
         assistant_text = ''
+        # Prepare assistant placeholder
+        history.append({'role': 'assistant', 'content': ''})
         for chunk in streamer:
             if cancel_event.is_set():
                 break