S2D: Sorted Speculative Decoding For More Efficient Deployment of Nested Large Language Models

Add code
Jul 02, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: