1 paper · 1 filter
Yongchang Hao, Lili Mou
Speculative sampling (SpS) has been successful in accelerating the decoding throughput of auto-regressive large language models by leveraging smaller draft models. SpS strictly enf…