Christopher Lee

@sandeepnmenon Hi! If you look at the definition of batch norm, you can see that we use the `fapply` function to run the `forward` with a SparseTensor. https://github.com/mit-han-lab/torchsparse/blob/74099d10a51c71c14318bce63d6421f698b24f24/torchsparse/nn/modules/norm.py#L13 I would...

@zhijian-liu Great! I will assign it to you for now and if I have time I'll take a look.

@evertonaleixo 500,000 epochs is way too many. In the original pix2pix paper, they use 200 epochs at most, for small datasets. If you're not getting reasonable results within 200 epochs,...

Hi @karantai ! Please reach out to me via linkedin or my email on my profile, I would love to chat about the repo.

This may be an upstream issue. It does not compile in debug mode but works when forcing release mode on the triplet.

@Haojia521 thanks for the report! Looks like some refactoring will be needed since we raise exceptions in a few methods called by constructors.

> > > For the radius query, there seems to be a minor bug. The python interface is returning square rooted distance, but the radius query takes in a squared...

Yeah, I agree it is confusing that the constructor radius is the regular distance but the `radius_neighbors` radius requires the squared distance, it would probably be good to make those...

Is this feature dependent on Visual Studio fixing the ticket linked above, or is there a path forward within VScode?

Hi, I was able to replicate the error on 2.2.0 and the latest master as well. ``` [3/4] C:\msys64\ucrt64\bin\c++.exe -DEIGEN_MPL2_ONLY -DGFLAGS_IS_A_DLL=1 -DGLOG_CUSTOM_PREFIX_SUPPORT -DGLOG_NO_ABBREVIATED_SEVERITIES -DNOMINMAX -Dceres_EXPORTS -IC:/Users/Chris/ceres-solver/include -IC:/Users/Chris/ceres-solver/internal -IC:/Users/Chris/ceres-solver/bin/include -IC:/msys64/ucrt64/include/eigen3 -g...