LoopCache: Accelerating Long-Context Self-Speculative Decoding for Looped Transformerswith Zizhuo Fu, Shengxuan Qiu, Linye Wei, Ziyu Tang +21 versionW_yf7hd2nvv1 · current