Dovetail: A CPU/GPU Heterogeneous Speculative Decoding for LLM inference

Add code
Dec 25, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: