Understanding Kubeflow Pipelines

[YannCoderre]

Roadmap

• [x] Create a new KF notebook: https://kubeflow.aaw-dev.cloud.statcan.ca/
• [x] Explore KFP and possible demo solutions
• [x] Keep track of any pain points/things you needed to google search or modify.
• [x] Update relevant demos/docs to reflect findings

End goal

• [x] be comfortable submitting a KF pipeline

  Reading Resources

• [ ] Kubeflow pipeline documentation
• [ ] ML Systems Course

[Souheil-Yazji]

Followed along with the available KFP documentation and the KFP examples repo provided by Christian. I think what we currently have in the slides is adequate. One thing that I couldn't find references to creating Images/packages which can be referenced by KF pipelines when creating a pipeline. The examples use images available in the StatCan ACR instance.