Torch-layers

Bugfix Release

Fixed Bugs:
Implicitly copied batchsize if `View`-Module if `view` contained `None`. Solution: make a deepcopy of the tuple before

Currently the following layers are implemented:

* [`Conv2dWithSamePadding`](torch_layers/conv_padding_same) [[docs]](https://justusschock.github.io/torch_layers/_api/_build/torch_layers/conv_padding_same.html)
* [`ActivationConv`](torch_layers/activation_conv) [[docs]](https://justusschock.github.io/torch_layers/_api/_build/torch_layers/activation_conv.html)
* [`MaxPool2dSamePadding`](torch_layers/maxpool_padding_same) [[docs]](https://justusschock.github.io/torch_layers/_api/_build/torch_layers/maxpool_padding_same.html)
* [`Upsample`](torch_layers/upsample) [[docs]](https://justusschock.github.io/torch_layers/_api/_build/torch_layers/upsample.html)
* [`View`](torch_layers/view) [[docs]](https://justusschock.github.io/torch_layers/_api/_build/torch_layers/view.html)