Ray is a flexible, high-performance distributed execution framework.

Ray comes with libraries that accelerate deep learning and reinforcement learning development:

• Ray.tune: Efficient Distributed Hyperparameter Search
• Ray RLlib: A Composable and Scalable Reinforcement Learning Library

• Ray can be installed on Linux and Mac with pip install ray.
• To build Ray from source, see the instructions for Ubuntu and Mac.

import time





def f():
    time.sleep(1)
    return 1

# Execute f serially.
results = [f() for i in range(4)]

import time
import ray

ray.init()

@ray.remote
def f():
    time.sleep(1)
    return 1

# Execute f in parallel.
object_ids = [f.remote() for i in range(4)]
results = ray.get(object_ids)

• Documentation
• Tutorial
• Blog
• Ray HotOS paper

• Ask questions on our mailing list ray-dev@googlegroups.com.
• Please report bugs by submitting a GitHub issue.
• Submit contributions using pull requests.