Update app.py

app.py

@@ -206,7 +206,7 @@ def model_inference(input_dict, history):
 
         # Process the full segment (e.g., remove <|im_end|>)
         processed_segment = current_model_output_segment.split("<|im_end|>", 1)[0] if "<|im_end|>" in current_model_output_segment else current_model_output_segment
-        messages.append(dict(role='assistant', content=processed_segment))
+        # messages.append(dict(role='assistant', content=processed_segment))
         # Append this processed segment to the cumulative display string for Gradio
         complete_assistant_response_for_gradio += [processed_segment + "\n\n"]
         yield complete_assistant_response_for_gradio # Ensure the fully processed segment is yielded to Gradio