scripts

Helper scripts

This folder contains helper scripts for common actions.

Building the docker image(s) and deploying to GCR

Within build_docker.sh, make sure you replace line 3 which sets the GCR_PATH variable.

Let's say you set your GCR_PATH=gcr.io/my-project/my-image.

$ ./scripts/build_docker.sh all

will build and deploy these images to gcr.io/my-project/my-image:train and gcr.io/my-project/my-image:serve, which you would then need to put within your cluster/train.yaml and cluster/serve.yaml files.

Connecting to the dev cluster in Client mode

$ ./scripts/get_cluster_ip.sh

External IP address of the head node: 35.186.77.141
Internal IP address of the head node: 10.130.0.107


If developing from a notebook, connect to the Ray cluster as follows:

import ray

ray.init("ray://10.130.0.107:10001")

Submitting the Training Job

Run

$ ./scripts/submit_train.sh

to submit the main training job, attach to the dashboard via

$ ray dashboard cluster/train.yaml

and go to http://localhost:8265 to view the Job logs.

Submitting the Serving job

Run

$ ./scripts/start_gradio.sh

to submit the serve deployment, attach to the GradIO deployment via

$ ray attach -p 8000 cluster/serve.yaml

and go to http://localhost:8000 to view the GradIO deployment.