Deep Learning Energy Measurement and Optimization
Project News ⚡
- [2024/05] Zeus is now a PyTorch ecosystem project. Read the PyTorch blog post here!
- [2024/02] Zeus was selected as a 2024 Mozilla Technology Fund awardee!
- [2023/12] We released Perseus, an energy optimizer for large model training: Preprint | Blog | Optimizer
- [2023/07] We used the
ZeusMonitor
to profile GPU time and energy consumption for the ML.ENERGY leaderboard & Colosseum.
Zeus is a library for (1) measuring the energy consumption of Deep Learning workloads and (2) optimizing their energy consumption.
Zeus is part of The ML.ENERGY Initiative.
Documentation Organization
- Getting Started: Instructions on installation and setup.
- Measuring Energy: How to measure time and energy programmatically and on the command line.
- Optimizing Energy: How to optimize energy.
- Research Overview: Overview of the research papers Zeus is rooted on.
- Source Code Reference: Auto-generated source code reference for the entire codebase.
We also provide usage examples in our GitHub repository.
Contact
Jae-Won Chung (jwnchung@umich.edu)