No declaration of self.conv_stride

[apyskir]

Hi,
these 2 lines cannot work:
https://github.com/neptune-ml/steppy-toolkit/blob/master/toolkit/pytorch_transformers/architectures/unet.py#L251
https://github.com/neptune-ml/steppy-toolkit/blob/master/toolkit/pytorch_transformers/architectures/unet.py#L299
There is no place, where self.conv_stride is declared. These UNets work only on assumption, that stride of convolutional layers is equal to 1, so I think that it should be stride = 1.
Cheers