The Baler project is a collaboration between 12 research physicists, computer scientists, and machine learning experts at the universities of Lund, Manchester, and Uppsala. The goal of this project is to provide a tool for ML-based data compression, that is modular, easy-to-use and cross-disciplinary.

My contributions here started as a side mission my overlord sent me on at the start of my PhD and projects like these add the kind of mental refreshment you need when you hit artistic blocks in academia!

Baler has now been presented at multiple scientific conferences, garnering interest from researchers in a broad set of fields such as Particle Physics, Computational Fluid Dynamics, Meteorology, Medical Imaging along with industrial interest from fields such as Autonomous Vehicles, Computing Accelerators (for eg. TPUs, IPUs), Weather Forecasting, Mobile Service Providers and more!

Learn more about us here, and click here to try Baler out to compress your own data! If you are interested in collaborating and have an idea, or, you find a missing feature that you would find useful, open an issue on our github and we will be happy to discuss! 

Currently, we are pursuing active research on:
  • Online Compression (think fast inference, low power, continual learning etc.)
  • Further Applications (expanding to neutrino experiments, video compression, QML studies)
  • Sustainability studies (evaluating environmental impact of Baler, mitigation strategies via sustainable code)