Repolode
Submit
ray-project

ray-project/ray

Ray is a unified framework for scaling Python and AI applications from a laptop to a cluster, with a distributed runtime and AI libraries.

View on GitHubWebsite
Stars
44,087
Forks
8,174
Open issues
3,610
Last commit
6 hours ago

What is it?

Ray is a distributed compute framework for Python and AI workloads. Its core runtime lets you turn ordinary Python functions into tasks and classes into stateful actors that run across a cluster, with shared objects passed between them.

On top of the core, Ray provides AI libraries for common ML pipeline stages: Data for scalable datasets, Train for distributed training, Tune for hyperparameter tuning, RLlib for reinforcement learning, and Serve for programmable model serving. The project also includes a dashboard for monitoring apps and clusters, and a distributed debugger.

Ray runs on a single machine, a cluster, major cloud providers, and Kubernetes. Installation is done with pip, and the same code can move from a laptop to a larger cluster without changes to the application structure.

Key features

  • 01Turns Python functions and classes into distributed tasks and stateful actors.
  • 02Provides Data, Train, Tune, RLlib and Serve libraries for ML pipeline stages.
  • 03Includes a Ray Dashboard for monitoring apps and clusters.
  • 04Offers a distributed debugger for Ray applications.
  • 05Runs on laptops, clusters, major clouds and Kubernetes.

Use cases

  • Scaling a Python data processing or ML job beyond one machine.
  • Distributed training or hyperparameter search across many workers.
  • Serving machine learning or LLM models behind a scalable endpoint.
  • Running reinforcement learning experiments on a cluster.

Quick start

  1. 1

    Install Ray with pip

    pip install ray

Topics

#distributed#machine-learning#python#distributed-training#model-serving#hyperparameter-tuning#reinforcement-learning#llm-inference#data-science#deep-learning

FAQ

What is ray?
Ray is a unified framework for scaling Python and AI applications from a laptop to a cluster, with a distributed runtime and AI libraries. Ray is a distributed compute framework for Python and AI workloads. Its core runtime lets you turn ordinary Python functions into tasks and classes into stateful actors that run across a cluster, with shared objects passed between them.
Is ray free to use? What license does it have?
ray is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
Is ray still maintained?
Actively maintained. The last commit to ray was on Oct 20, 2026.
What can I use ray for?
Scaling a Python data processing or ML job beyond one machine. Distributed training or hyperparameter search across many workers. Serving machine learning or LLM models behind a scalable endpoint. Running reinforcement learning experiments on a cluster.
How do I install ray?
Install Ray with pip: pip install ray
Where can I get ray?
The source code and installation instructions are on GitHub: https://github.com/ray-project/ray

Summary generated by AI from the project's README. GitHub stats refresh daily.

Related projects