"Accelerating LLM Inference with Staged Speculative Decoding."

Benjamin Spector, Christopher Ré (2023)

Details and statistics

DOI: 10.48550/ARXIV.2308.04623

access: open

type: Informal or Other Publication

metadata version: 2023-10-30

a service of  Schloss Dagstuhl - Leibniz Center for Informatics