TokenTiming: A Dynamic Alignment Method for Universal Speculative Decoding Model Pairs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Xiao, Sibo, Fu, Jinyuan, Xie, Zhongle, Shou, Lidan
Format:	Preprint
Published:	2025
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2510.15545