Custom PyTorch layers

This repository implements various layers (either to replace deprecated layers of torch or to add useful new features or to encapsulate parts of the functional API into layers).

Currently the following layers are implemented:

• Conv2dWithSamePadding [docs]
• ActivationConv [docs]
• MaxPool2dSamePadding [docs]
• Upsample [docs]
• View [docs]

Installation

This package can be installed via pip with:

pip install torch_layers

or from source via:

pip install git+https://github.com/justusschock/torch_layers