Gist originally from Introducing Ray Lightning: Multi-node PyTorch Lightning Training made easy: https://www.anyscale.com/blog/introducing-ray-lightning-multi-node-gpu-training-for-pytorch-lightning-made

cluster.yaml

cluster_name: ml

# Cloud-provider specific configuration.
provider:
    type: aws
    region: us-west-2
    availability_zone: us-west-2a,us-west-2b

# How Ray will authenticate with newly launched nodes.
auth:
    ssh_user: ubuntu

head_node:
    InstanceType: p3.8xlarge
    ImageId: latest_dlami

    # You can provision additional disk space with a conf as follows
    BlockDeviceMappings:
        - DeviceName: /dev/sda1
          Ebs:
              VolumeSize: 100
worker_nodes:
    InstanceType: p3.2xlarge
    ImageId: latest_dlami

file_mounts: {
    "/path1/on/remote/machine": "/path1/on/local/machine",
}

# List of shell commands to run to set up nodes.
setup_commands:
    - pip install -U ray-lightning