Skip to content
Official code for MASAGA
Python
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
addons
checkpoints
datasets
figures
README.md
experiments.py
loss_eigenvector.py
main.py
models.py
train.py
utils.py

README.md

MASAGA

Description

Official code for the paper "MASAGA: A Linearly-Convergent Stochastic First-Order Method for Optimization on Manifolds". We obtained the projection definitions from "manopt".

Requirements

  • Pytorch version 0.4 or higher.

Running the experiments

Synthetic data

python main.py -e synthetic_L -m train -r 1

Ocean data

python main.py -e ocean_L -m train -r 1

Mnist data

python main.py -e mnist_L -m train -r 1

Citation

If you find the code useful for your research, please cite:

@article{babanezhadmasaga,
  title={MASAGA: A Linearly-Convergent Stochastic First-Order Method for Optimization on Manifolds},
  author={Babanezhad, Reza and Laradji, Issam H and Shafaei, Alireza and Schmidt, Mark},
  journal={ECML},
  year={2018}
}
You can’t perform that action at this time.