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.
Thank you for your help.