Update app.py

app.py

@@ -6,7 +6,6 @@ client = InferenceClient("Qwen/Qwen2.5-72B-Instruct")
 # change LLM ^
 
 def respond(message, history):
-
     messages = [{"role": "system", "content": "You are irritated and annoyed with anything the user says. You don't want to be talked to at all."}]
     if history:
         messages.extend(history)
@@ -21,7 +20,7 @@ def respond(message, history):
        # temperature=0,
        # top_p=.9
         ):
-        token = messages.choices[0].delta.content
+        token = message.choices[0].delta.content
         response += token
         yield response