/dfno

Distributed Fourier Neural Operators

Primary LanguagePythonMIT LicenseMIT

DOI

Distributed Fourier Neural Operators

Implements a distributed version of Fourier Neural Operators based off of the code written by Zongyi Li, Shuhao Cao, and Jack Griffiths using DistDL.

Installation

To install this package you can either clone it and install it in dvelopper mode:

git clone https://github.com/slimgroup/dfno.git
cd dfno
pip install -e .

or directly install it with pip as a standard pythn package pip install git+https://github.com/slimgroup/dfno.git

Author

This package is written by Thomas Grady tgrady@gatech.edu at Georgia Institute of Technology

Citation

If you like and use our package, please cite our preprint:

@techreport {grady2022SCtll,
	title = {Model-Parallel Fourier Neural Operators as Learned Surrogates for Large-Scale Parametric PDEs},
	number = {TR-CSE-2022-1},
	year = {2022},
	month = {04},
	keywords = {CCS, deep learning, Fourier neural operators, HPC, large-scale, Model Parallelism, Multiphase Flow, Operator Learning},
	url = {https://arxiv.org/pdf/2204.01205.pdf},
	software = {https://github.com/slimgroup/dfno},
	author = {Thomas J. Grady II and Rishi Khan and Mathias Louboutin and Ziyi Yin and Philipp A. Witte and Ranveer Chandra and Russell J. Hewett and Felix J. Herrmann}
}