Skip to main content
Article·anyscale.com
raydistributed-computinginferencefine-tuningenterpriseanyscale

Anyscale

Anyscale is an enterprise platform built on Ray for scaling AI. It offers managed Ray clusters, model serving, and fine-tuning infrastructure, streamlining production AI workloads.

beginner15 min5 steps
The play
  1. Create an Anyscale Account
    Sign up for a free Anyscale account at their website. This provides access to their managed Ray clusters.
  2. Install the Anyscale CLI
    Install the Anyscale CLI using pip to interact with your Anyscale account from your terminal.
  3. Configure Anyscale CLI
    Configure the Anyscale CLI with your Anyscale account credentials.
  4. Launch a Ray Cluster
    Launch a Ray cluster on Anyscale using the CLI. Specify the cluster configuration based on your workload.
  5. Connect to the Cluster
    Connect to the Ray cluster you created using the Anyscale CLI.
Starter code
# Sample Ray script to run on Anyscale
import ray

@ray.remote
def square(x):
    return x * x

if __name__ == '__main__':
    ray.init()
    results = ray.get([square.remote(i) for i in range(4)])
    print(results)
    ray.shutdown()
Source
Anyscale — Action Pack