opendatahub-io / distributed-workloads

Artifacts for the Distributed Workloads stack as part of ODH
Apache License 2.0
22 stars 42 forks source link

Add e2e tutorial using KubeRay alongside ODH #101

Open anishasthana opened 1 year ago

anishasthana commented 1 year ago

This should include supporting artifacts such as notebooks, templates, etc.

anishasthana commented 1 year ago

fyi @goern

goern commented 1 year ago

From my point of view the story of the tutorial should be something like

As a Data Scientists,
I want to create a Ray cluster and distribute my training process, 
and I want to utilize more than one Python process in a jupyter notebook,
and I know how to use an OpenShift cluster using the command line,
so that I utilize more resources on the RHODS cluster,
and so that I can prove that my training process was executed faster.

wdyt?

tedhtchang commented 1 year ago

Here is an example we may use https://developer.ibm.com/tutorials/awb-run-distributed-model-training-workloads-ray-codeflare/