Jeremy Lewi

Installing papermill works on `tensorflow/tensorflow:2.1.2-jupyter` even thought is python 3.6.9

It looks like we need to also look at the datasets and see if there is a model training in progress.

#182 auto PR created for a model trained by manually running the notebook. Need to verify that a new model is trained automatically and then deployed.

kubeflow/code-intelligence#184 opened a PR to update to the same model. It doesn't look like a new model got trained.

Looks like this issue got labeled so it might be a fluke.

@GauravSarkar I appreciate your willingness to help. I will try to review your PRs and answer questions but I don't have a whole bunch of time.

New image was built. https://hub.docker.com/r/kubeflow/kubeflow-triage Still failing. ``` Update Kanban6s ##[error]Docker run failed with exit code 1 Run kubeflow/code-intelligence/Issue_Triage/action@master /usr/bin/docker run --name kubeflowkubeflowtriage_345db4 --label 671ee6 --workdir /github/workspace --rm -e INPUT_ISSUE_NUMBER...

@hamelsmu Do you know offhand if there is a good way to encode metadata in a PR?

I didn't see anything in the payload. https://developer.github.com/v4/object/pullrequest/ I think we will just have to put data into the title and comment and use that to look for PRs.

@hamelsmu I'd like to be able to programmatically identify if the bot had previously opened a PR to update a model to a specific version. So the easiest thing would...