Trained AzureML model prerequisite and azureml-container

[adelton]

In pipelines/README.md, Prerequisites start with

Trained AzureML model

The Deploy AzureML Container build pipeline section in oc apply -k tekton/azureml-container-pipeline creates an azureml-container Pipeline.

Yet, there is nothing about AzureML in the tensorflow-housing model and image that gets built -- those used TensorFlow and MLflow, IIUIC. And the bike-rentals image does not get built at all (#93).

The list of prerequisites, section names, and Pipeline names should match the output they produce. If AzureML-trained model is currently not deployed and tested, renaming the Pipeline and updating the README might be useful.

[piotrpdev]

One of our main focuses at the start of the PoC was to take a model trained on AzureML (bike-rentals-auto-ml) and do stuff with it, Stefan's pipeline code (which our current code derives from) also focused on this: https://github.com/stefan-bergstein/azureml-model-to-edge/tree/dev/tekton/azureml-container-pipeline

However we are now going to use models from S3 (#70), PVC (#87), Azure, and Git (#57).

Renaming the pipeline for a more general name is something we will probably do, however there are multiple people currently making changes to it.

[adelton]

So should this be marked as a duplicate of some other issue where the renaming is tracked?

[piotrpdev]

So should this be marked as a duplicate of some other issue where the renaming is tracked?

I guess this could maybe fall under #62? Restructuring the whole repo might be a task that needs to be split into multiple smaller issues anyway though. If you want you can keep this up and the pipeline can be renamed when nobody is working on it (hopefully soon?)

[adelton]

The Prerequisites entry was removed in #100.

Renaming the directories and files now being tested in #131.