SWIFT: On-the-Fly Self-Speculative Decoding for LLM Inference AccelerationHeming Xia,Yongqi Li, Jun Zhang,Cunxiao Du,Wenjie LiICLR 2025(2025)引用 0|浏览10AI 理解论文溯源树样例生成溯源树,研究论文发展脉络Chat Paper正在生成论文摘要