KVSwap: Disk-aware KV Cache Offloading for Long-Context On-device Inference

Mr Huawei Zhang, Chunwei Xia, Zheng Wang
November 25250
Code Dataset Project Project
Type
Conference paper
Publication
The 24th Annual International Conference on Mobile Systems, Applications and Services (MobiSys)(Artifacts Evaluated!)
Mr Huawei Zhang
Mr Huawei Zhang
PhD Student

PhD student under the supervision of Zheng Wang.

Chunwei Xia
Chunwei Xia
Lecturer (Assistant Professor)
Zheng Wang
Zheng Wang
Professor of Intelligent Software Technology