AdrianSosic

Interesting, thanks for sharing. So it seems that some of the backprop graphs keep lying around, right? I mean – if I understand correctly what `detach_test_caches` is supposed to do...

I also thought about this problem a while ago and agree that we probably need a solution for it. I guess we all have the same kind of uses cases...

Unclear, but we can go incrementally, if you prefer. A simple half-way approach would be to at least type the context, perhaps. Something like adding an additional field, e.g. `constructor`...

I'm on it

@Waschenbacher @Scienfitz @AVHopp: I'm closing the PR now because we there are other things regarding cardinality constraints we should tackle first, i.e.: * finishing #394 * adding an example *...

> [@AdrianSosic](https://github.com/AdrianSosic) any conclusion on [#668](https://github.com/emdgroup/baybe/pull/668) and its connection to point 5? Not yet. Tried a bit yesterday but wasn't even sure what to I'm even looking for exactly. Do...

Hi @saitcakmak, thanks for the feedback. I can try to free some time for this in the next few weeks 👍🏼 Regarding backwards compatibility, perhaps it's worth discussing what the...

@AVHopp, @Scienfitz: Before accepting, please verify that running the `campaign_stopping` example now produces stable results (you can set the MC simulations to a smaller number, of course)

@AVHopp, @Scienfitz Finally, the epic is completed 😎 Last gate before merging into main

Alright, can do. Before, quick comment on what you said: > After having given several people the metadata approach for workarounds, no one has described the approach as `inelegant` or...