Ray is an open source framework that provides a simple, flexible, and universal API for building and running distributed applications.
Learn how to get started with RLlib for scalable reinforcement learning, from training to serving.Watch the video
Missed Ray Summit? All the sessions and tutorials are now available on-demand.Watch now
New to Ray? Jump-start your learning with this free hands-on tutorial on Ray Core.Watch the tutorial
With simple primitives, Ray makes converting your single-machine code to run in distributed mode intuitive, flexible, and easy.
Run the same code — from prototyping on your laptop to running a petabyte-scale system in production.
Ray handles all the tricky details of distributed execution — compute orchestration, scheduling, autoscaling, fault tolerance, and more — letting you enjoy seamless access to infinite compute.
Ray includes a rich set of data processing and machine learning libraries, as well as hooks into popular ones like Tensorflow, PyTorch, XGBoost and more. Scale your end-to-end machine learning application on a single distributed compute substrate, eliminating architectural complexity and simplifying operations.
Seamlessly scale workloads on infrastructure of your choosing — public cloud, private data centers, bare metal, Kubernetes cluster, etc.
Or choose Anyscale, and leave the infrastructure to us for bonus operational happiness.
Follow along these short tutorials to begin your Ray journey.