PiLLM: Resource-Efficient LLM Inference Using Workload Prediction.

Yunqian Fan, Shihao Bai, Ruihao Gong, Zaijun Wang, Rui Fan

Browse the full EuroSys paper archive.