padding="same" and groups=in_channels in 3D convolution not supported

tensorly / torch

TensorLy-Torch: Deep Tensor Learning with TensorLy and PyTorch

http://tensorly.org/torch/

BSD 3-Clause "New" or "Revised" License

70 stars 18 forks source link

padding="same" and groups=in_channels in 3D convolution not supported #36

Open GDLmazari opened 2 months ago

GDLmazari commented 2 months ago

I have tried to make a 3D convolutional layer with the following parameters: padding='same' and groups='in_channels' and l get an AssertionError. Either with tltorch.FactorizedConv() or tltorch.FactorizedConv.from_conv()

It seems that padding='same' with groups is not supported in tensorly.

padding='same'in pytorch pytorch doc on padding
groups= in_channels in pytorch groups in conv

Thank you for your help.

JeanKossaifi commented 4 weeks ago

Thanks for opening the issue! Could you share a minimal snippet of code to reproduce the issue?