Accurate Neural Network Potential on PyTorch

Metrics:

Checks:

Deploy:

We only provide compatibility with nightly PyTorch, but you can check if stable PyTorch happens to be supported by looking at the following badge:

TorchANI is a pytorch implementation of ANI. It is currently under alpha release, which means, the API is not stable yet. If you find a bug of TorchANI, or have some feature request, feel free to open an issue on GitHub, or send us a pull request.

Install

TorchANI requires the latest preview version of PyTorch. Please install PyTorch before installing TorchANI.

Please see PyTorch's official site for instructions of installing latest preview version of PyTorch.

Note that if you updated TorchANI, you may also need to update PyTorch.

After installing the correct PyTorch, you can install TorchANI by pip or conda:

pip install torchani

conda install -c conda-forge torchani

See https://github.com/conda-forge/torchani-feedstock for more information about the conda package.

To run the tests and examples, you must manually download a data package

./download.sh

CUAEV (Optional)
To install AEV CUDA Extension (speedup for AEV forward and backward), please follow the instruction at torchani/cuaev.

Citation

Please cite the following paper if you use TorchANI

Xiang Gao, Farhad Ramezanghorbani, Olexandr Isayev, Justin S. Smith, and Adrian E. Roitberg. TorchANI: A Free and Open Source PyTorch Based Deep Learning Implementation of the ANI Neural Network Potentials. Journal of Chemical Information and Modeling 2020 60 (7), 3408-3415,

Please refer to isayev/ASE_ANI for ANI model references.

ANI model parameters

All the ANI model parameters including (ANI2x, ANI1x, and ANI1ccx) are accessible from the following repositories:

Develop

To install TorchANI from GitHub:

git clone https://github.com/aiqm/torchani.git
cd torchani
pip install -e .

After TorchANI has been installed, you can build the documents by running sphinx-build docs build. But make sure you install dependencies:

pip install -r docs_requirements.txt

To manually run unit tests, do

pytest -v

If you opened a pull request, you could see your generated documents at https://aiqm.github.io/torchani-test-docs/ after you docs check succeed. Keep in mind that this repository is only for the purpose of convenience of development, and only keeps the latest push. The CI runing for other pull requests might overwrite this repository. You could rerun the docs check to overwrite this repo to your build.

Note to TorchANI developers

Never commit to the master branch directly. If you need to change something, create a new branch, submit a PR on GitHub.

You must pass all the tests on GitHub before your PR can be merged.

Code review is required before merging pull request.