[EPIC] Training pipeline

[Shreyanand]

This issue will be used to track all outstanding tasks required to complete the training pipeline. Demo 2 (issue #58) currently supports inference using pre-trained models. For a complete demo, we should have the training pipeline in place as well.

Issues:

• Write training notebooks Write notebooks for the training pipeline #11
• Run in a kubeflow pipeline Run training in a kubeflow pipeline #154
• Integrate training pipeline and the inference pipeline so that they work together #179
• Update README to reflect the training pipeline content #180

[JeremyGohBNP]

@Shreyanand @chauhankaranraj Hello! I hope you enjoyed your days off! I am following up on the training pipeline implementation. How has it been going? Do you need additional info or material? Thanks for the update.

[chauhankaranraj]

@Shreyanand @chauhankaranraj Hello! I hope you enjoyed your days off! I am following up on the training pipeline implementation. How has it been going? Do you need additional info or material? Thanks for the update.

Hi @JeremyGohBNP, I have opened PRs (#155, #158) for both the training notebooks required in the pipeline, but they haven't been merged yet coz of some recent issues with the CI system. As soon as these are merged, we'll assemble them into a kubeflow pipeline and that would conclude the initial training pipeline implementation.

[JeremyGohBNP]

Hi @chauhankaranraj! I see that #155 and #158 are now closed (?). Does this mean the issues with the CI system were resolved? Are we now able to assemble the PRs into a training pipeline? Thanks!

[chauhankaranraj]

Hi @chauhankaranraj! I see that #155 and #158 are now closed (?). Does this mean the issues with the CI system were resolved? Are we now able to assemble the PRs into a training pipeline? Thanks!

Hello @JeremyGohBNP! Yes, #155 just got merged yesterday afternoon. So now we're working on assembling the notebooks in an elyra pipeline :)