# Remote execution using Ray

machinable comes with support for seamless remote execution using Ray.

Ray supports local as well as remote execution on AWS and Google Cloud. Please refer to Ray's documentation to learn how to setup and start Ray.

machinable will automatically detect when Ray has been initialised via ray.init() and execute on the connected Ray backend.

import ray
ray.init()   # initialise Ray or connect to remote server

# detects and executes on the Ray cluster
# executes locally only in spite Ray being available
ml.execute(ml.Task().component('demo'), local=True)