About

This is the supplementary code (in Python) for Decentralized Algorithm 1 from the paper “Optimal Algorithms for Decentralized Stochastic Variational Inequalities” by Dmitry Kovalev, Aleksandr Beznosikov, Abdurakhmon Sadiev, Michael Persiianov, Peter Richtarik and Alexander Gasnikov.

Usage

Just open Synthetic Jupyter notebook for synthetic data or Real for LIBSVM data at decentralized setup or Synthetic at centralized.

Acknowledgement

I am deeply thankful to Alexander Rogozin, whose code influenced decentralized part of this repo and Yura Malitsky for centralized part.

Reference

If you find this code useful, please cite our paper:

@article{kovalev2022optimal,
  title={Optimal Algorithms for Decentralized Stochastic Variational Inequalities},
  author={Kovalev, Dmitry and Beznosikov, Aleksandr and Sadiev, Abdurakhmon and Persiianov, Michael and Richt{\'a}rik, Peter and Gasnikov, Alexander},
  journal={arXiv preprint arXiv:2202.02771},
  year={2022}
}