Is it worth reconsidering the way data is packaged to be presented in the dashboard?

[HarrisonWilde]

As it stands, we need all four items requested to be uploaded by the user, might be worth wrapping them all up in a single pickle file. This would only really be useful in context of using the dashboard, rather than using parts of the package in pre-existing workflows and pipelines. We already have auto upload during a pipeline run / via command line experiment path, so I think things are fine as they are, but from a UX perspective there is probably a better way to do this. Will record any examples I find here.