Mixed-scale Dense Networks for PyTorch

An implementation of Mixed-Scale Dense networks in PyTorch.

• Free software: GNU General Public License v3
• Documentation: [https://ahendriksen.github.io/msd_pytorch]

Getting Started

It takes a few steps to setup Mixed-scale Dense Networks for PyTorch on your machine. We recommend installing Anaconda package manager for Python 3.

Requirements

This package requires

• Linux x64
• CUDA 10.0 and/or 11.0 compatible graphics card
• Anaconda package manager

This package is compatible with python 3.7, 3.8, and 3.9.

Installing with Conda

The following instructions install msd_pytorch with pytorch version 1.8.1:

conda install msd_pytorch=0.10.1 cudatoolkit=11.1 -c aahendriksen -c pytorch -c defaults -c conda-forge
conda install msd_pytorch=0.10.1 cudatoolkit=10.2 -c aahendriksen -c pytorch -c defaults -c conda-forge

Note: The order of the channels is important. If you install pytorch from the default conda channel or from conda-forge, installation might fail.

Installing from source

To install msd_pytorch from source, you need to have the CUDA toolkit installed. Specifically, you need nvcc and a compatible C++ compiler. Moreover, you need to have a working installation of PyTorch.

To get the source code, simply clone this GitHub project.

git clone https://github.com/ahendriksen/msd_pytorch.git
cd msd_pytorch

Using pip to install the package automatically triggers the compilation of the native C++ and CUDA code. So you need to direct the installer to a CUDA-compatible C++ compiler in this way:

CC=/path/to/compatible/cpp/compiler pip install -e .[dev]

Or, if the standard C++ compiler is compatible with CUDA:

pip install -e .[dev]

Running the examples

To learn more about the functionality of the package check out our examples folder.

Cite

If you find our work useful, please cite as:

@software{hendriksen-2019-msd-pytor,
  author       = {Hendriksen, Allard A.},
  title        = {ahendriksenh/msd\_pytorch: v0.7.2},
  month        = dec,
  year         = 2019,
  publisher    = {Zenodo},
  version      = {v0.7.2},
  doi          = {10.5281/zenodo.3560114},
  url          = {https://doi.org/10.5281/zenodo.3560114}
}

Authors and contributors

• Allard Hendriksen - Initial work
• Ryan Pollitt - Port CUDA convolution code from 2D to 3D!
• Jonas Adler - Discussions and code
• Richard Schoonhoven - Testing and patches

See also the list of contributors who participated in this project.

How to contribute

Contributions are always welcome. Please submit pull requests against the dev branch.

If you have any issues, questions, or remarks, then please open an issue on GitHub.

License

This project is licensed under the GNU General Public License v3 - see the LICENSE.md file for details.