Response Length Perception and Sequence Scheduling: An LLM-Empowered LLM Inference Pipeline

Large language models (LLMs) have revolutionized the field of AI, demonstrating unprecedented capacity across various tasks.