Update app.py

app.py

@@ -72,31 +72,26 @@ def query_model(model_name: str, messages: List[Dict[str, str]]) -> str:
         return f"{model_name} error: {str(e)}"
 
 def respond(message: str, history: List[List[str]]) -> str:
-    """Handle sequential model responses with collaboration"""
+    """Handle sequential model responses with progressive context"""
     messages = [{"role": "user", "content": message}]
     responses = []
     
-    # Define processing order
-    processing_order = [
-        "Qwen2.5-Coder-32B-Instruct",
-        "Qwen2.5-72B-Instruct", 
-        "Llama3.3-70B-Instruct"
-    ]
+    # First model: Initial response
+    reply1 = query_model("Qwen2.5-Coder-32B-Instruct", messages)
+    responses.append(f"**Qwen2.5-Coder-32B-Instruct**:\n{reply1}")
+    messages.append({"role": "assistant", "content": reply1})
     
-    # Process models in sequence with accumulating context
-    for model_name in processing_order:
-        # Get current model's response
-        response = query_model(model_name, messages)
-        responses.append(f"**{model_name}**:\n{response}")
-        
-        # Add model's response to message history for next model
-        messages.append({
-            "role": "assistant",
-            "content": f"{model_name} response: {response}"
-        })
+    # Second model: Contextualize with first response
+    reply2 = query_model("Qwen2.5-72B-Instruct", messages)
+    responses.append(f"\n\n**Qwen2.5-72B-Instruct**:\n{reply2}")
+    messages.append({"role": "assistant", "content": reply2})
     
-    # Format output with collaboration timeline
-    return "\n\n→→→ Next Model Builds Upon This →→→\n\n".join(responses)
+    # Third model: Final synthesis
+    reply3 = query_model("Llama3.3-70B-Instruct", messages)
+    responses.append(f"\n\n**Llama3.3-70B-Instruct**:\n{reply3}")
+    
+    # Return progressive responses with clear separation
+    return "\n\n".join(responses)
 
 # Create the Gradio interface
 chat_interface = gr.ChatInterface(