"Accelerating Production LLMs with Combined Token/Embedding Speculators."

Davis Wertheimer et al. (2024)

Details and statistics

DOI: 10.48550/ARXIV.2404.19124

access: open

type: Informal or Other Publication

metadata version: 2024-05-27

a service of  Schloss Dagstuhl - Leibniz Center for Informatics