SparDA

Python ★ 22 updated 17d ago

Sparse Decoupled Attention for Efficient Long-Context LLM Inference

No plain-English explanation yet — one is being written right now. Check back in a minute.