About#

Hyperactive is an optimization and data collection toolbox for convenient and fast prototyping of computationally expensive models.

About Hyperactive#

Hyperactive provides a unified interface for hyperparameter optimization using various gradient-free optimization algorithms. It supports optimization for scikit-learn, sktime, skpro, and PyTorch Lightning models, as well as custom objective functions.

Mission#

Hyperactive aims to make hyperparameter optimization accessible and practical for machine learning practitioners. By providing a unified API across many optimization algorithms and ML frameworks, it reduces the barrier to finding optimal model configurations.

Key Features#

  • 20+ Optimization Algorithms: From simple hill climbing to advanced Bayesian optimization, population methods, and Optuna integration.

  • Experiment-Based Architecture: Clean separation between what to optimize (experiments) and how to optimize (algorithms).

  • Framework Integrations: First-class support for scikit-learn, sktime, skpro, and PyTorch Lightning.

  • Flexible Search Spaces: Discrete, continuous, and mixed parameter spaces using familiar NumPy/list syntax.

  • Production Ready: Battle-tested since 2019 with comprehensive testing and active maintenance.

Sponsorship#

Hyperactive is sponsored by the German Center for Open Source AI (GC.OS).

GC.OS Sponsored

Citing Hyperactive#

If you use Hyperactive in your research, please cite it:

@Misc{hyperactive2021,
  author =   {{Simon Blanke}},
  title =    {{Hyperactive}: An optimization and data collection toolbox
              for convenient and fast prototyping of computationally
              expensive models.},
  howpublished = {\url{https://github.com/SimonBlanke}},
  year = {since 2019}
}

Community#