Kangaroo: Lossless Self-Speculative Decoding via Double Early Exiting

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Liu, Fangcheng, Tang, Yehui, Liu, Zhenhua, Ni, Yunsheng, Han, Kai, Wang, Yunhe
Format:	Preprint
Published:	2024
Subjects:	Computation and Language Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2404.18911