Skip to content

stevelaskaridis/Maestro-LoD

 
 

Repository files navigation

Maestro: Uncovering Low-Rank Structures via Trainable Decomposition

Implementation of Maestro's LOD (Low-rank ordered Decomposition) for different type of NN layers in PyTorch.

The repo contains:

How to run

pip install -r requirements
python setup.py develop

Reference

If you find this repo useful, please cite the paper:

@inproceedings{
    horv{\'a}th2024maestro,
    title={Maestro: Uncovering Low-Rank Structures via Trainable Decomposition},
    author={Samuel Horv{\'a}th and Stefanos Laskaridis and Shashank Rajput and Hongyi Wang},
    booktitle={Forty-first International Conference on Machine Learning},
    year={2024},
    url={https://openreview.net/forum?id=7bjyambg4x}
}

About

No description, website, or topics provided.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 69.0%
  • Python 30.8%
  • Shell 0.2%