Skip to content

3PCv{1,2,3,4,5}, EF21, MARINA, GD on quadratic loss and CLAG on logistic loss.

License

Notifications You must be signed in to change notification settings

IgorSokoloff/3pc_experiments_source_code

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation

This repository contains Python 3 source code for the experimental part of the paper titled "3PC: Three Point Compressors for Communication-Efficient Distributed Training and a Better Theory for Lazy Aggregation" (arXiv, ICML).

Reference

In case you find the paper or code useful for your research, please consider citing it in your work.

@inproceedings{richtarik20223pc,
 title={3PC: Three point compressors for communication-efficient distributed training and a better theory for lazy aggregation},
 author={Richt{\'a}rik, Peter and Sokolov, Igor and Gasanov, Elnur and Fatkhullin, Ilyas and Li, Zhize and Gorbunov, Eduard},
 booktitle={International Conference on Machine Learning},
 pages={18596--18648},
 year={2022},
 organization={PMLR}
}

License

License: MIT

About

3PCv{1,2,3,4,5}, EF21, MARINA, GD on quadratic loss and CLAG on logistic loss.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published