Currently we have been using the flip-coin example. troyer@redhat.com has working examples that are better indicative of actual ml workflows that we should incorporate, or provide as sample pipelines for both devs and users.

This ticket should also figure out where these samples and the documentation for how to leverage these must reside.

Trevor's demo: https://drive.google.com/file/d/14c5lmkdEUHN7r4QFoyoCluUqwQQFZkWK/view