Anyscale Academy

Ray was created to make it easier to scale diverse computation tasks and distributed state across a cluster, with a minimum of distributed systems expertise and knowledge required. This crash course discusses the particular challenges Ray meets, how Ray works under the hood, how to use Ray in your own applications, and how to debug problems when they occur.

Many users of Ray will actually use a higher-level, domain-specific libraries, rather than the core API. The other tutorials cover these libraries.

Ray RLlib is the Ray-based library for implementing reinforcement learning applications, supporting all the popular, state-of-the-art libraries, including integrations with TensorFlow and PyTorch for deep reinforcement learning. This tutorial explains RL principles and common algorithms, including multi-armed bandits, using a series of example problems.

This tutorial covers Ray Tune for hyperparameter tuning and Ray Serve for scalable, intuitive model serving.