EdgeServe: A Streaming System for Decentralized Model Serving

arXiv:2303.08028v3 Announce Type: replace-cross
Abstract: The relevant features for a machine learning task may arrive as one or more continuous streams of data. Serving machine learning models over streams of data creates a number of interesting systems challenges in managing data routing, time-synchronization, and rate control. This paper presents EdgeServe, a distributed streaming system that can serve predictions from machine learning models in real time. We evaluate EdgeServe on three streaming prediction tasks: (1) human activity recognition, (2) autonomous driving, …

abstract arxiv challenges continuous control cs.db cs.dc cs.lg data decentralized distributed features machine machine learning machine learning models paper rate routing streaming synchronization systems type