DRLServe: Adaptive Prefill Chunking with Deep Reinforcement Learning for LLM Inference.

Chongxiang Sun, Han Bao, Yijie Wang

Browse the full ECAI paper archive.