Get More with LESS: Synthesizing Recurrence with KV Cache Compression for Efficient LLM Inference

Add code
Feb 14, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: