Dean Wampler
Author of O'Reilly's "What Is Ray?" report and "Programming Scala, Third Edition"
Dean is a Principal Software Engineer at Domino Data Lab, where he works on various aspects of the Domino platform for data scientists. He's the author of O'Reilly's "What Is Ray?" report and "Programming Scala, Third Edition" (forthcoming), which covers Scala 3. You can find Dean on Twitter: @deanwampler.
Past Activities
Code Mesh V
19.45 - 20.25
Ray: A System for High-performance, Distributed Computing
Ray is an open-source, distributed framework from U.C. Berkeley’s RISELab that easily scales Python and Java applications from a laptop to a cluster. While broadly applicable, it was developed to solve the unique performance challenges of ML/AI systems, such as the heterogeneous tasks and state management required for reinforcement learning (RL), everything from training neural networks to running simulators. It was also designed to be as easy to use as possible, reflecting the goals of the ML researchers who created it.
I'll explain the problems that Ray solves for cluster-wide scaling of general-purpose applications with specific examples from RLlib, the Ray-based library for Reinforcement Learning (RL). We’ll see that Ray’s features include rapid scheduling and execution of “tasks”, management of distributed state using a simple “actor” model, and an intuitive API.