Nice and inspiring work! However, I want to know if any chance to use negative prompts during inference? Looking forward to your reply.