Acceleration Multiple Heads Decoding for LLM via Dynamic Tree Attention

Add code
Feb 09, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: