Ray logo


Runtime engine
Pipeline orchestration
Model serving

Ray is an open-source tool that makes it simple to scale compute-intensive Python workloads.

Use it when

  • You want to parallelize your machine learning computation across several machines.
  • You want a general-purpose distributed computing library that supports heterogeneous workloads and is not restricted to structured data.
  • You want to seamlessly scale your code from your local machine to a cluster.
  • You want high-level libraries that support model training, hyperparameter tuning, building pipelines, and serving models.

Example stacks

Airflow + MLflow stack


pip install ray