I've been trying to convince support that there is an issue regarding this for 2 weeks now and have been gaslit repeatedly. I've gone as far as to archive+delete every single task, pipeline, dataset, and model (including hidden ones, yes I waited 24h) so the workspace is completely empty, and it's still showing 910MB of metrics usage (+200MB artifacts) for an empty workspace. Support still won't look into it (workspace id 385ec3f32cfe4feda5b2c3db610db744 if anyone wants to actually be of help).

I would very much prefer to pay for the enterprise plan than self-host, but I will not be doing that if there's a chance that I could be paying for up to $250/month in fees due to some internal bug, like @<1898906633770110976:profile|MinuteFlamingo30> might be. I saw this was on the roadmap as far back as 2023, and IMO it really needs to be prioritized. Not overly enthused about how long confirmed issues on github stay unaddressed either.

Hi @<1898906633770110976:profile|MinuteFlamingo30> , from my understanding this is actually on the roadmap. Currently there is no easy way to check it. Basically any experiment with a lot of scalars or console logs (think like experiments that ran for very long)