ProudMosquito87 Just a few pointers on how we convert the TB histograms to awesome (but less accurate) 3D surfaces.
First I have to admit, I almost never use these histograms, maybe to detect a plateau of if something goes really wrong...
The 3D surface is basically grouping all the histograms and then bucketing them (I think the default is 50 buckets) so that you get a general feel of what's going on, not necessary a detailed view. Bottom line, you are correct, the TB is the source of truth here 🙂
From what i remember the bins in tb are wider. And the tapering off around zero cannot be real since this happens in super sparse modela. Overall if you are sure, than this is a nice issue to open on GitHub.
I looked at the values myself just in case. Tensorboard is right for sure 🙂
Honestly, it looks like the tensorboard representation is the wrong one. Only one way to find out - you need to plot the histogram on your own 😅