A hyperparameter optimization library

Area: Artificial intelligence

CASS member: RAPIDS

Description

DeepHyper is first and foremost a hyperparameter optimization (HPO) library for single and multi-objective optimization. By leveraging this core HPO functionality, DeepHyper also provides neural architecture search, multi-fidelity and ensemble capabilities. With DeepHyper, users can easily perform these tasks on a single machine or distributed across multiple machines, making it ideal for use in a variety of environments.

Target audience

Machine Learning developers and researchers or engineers interested in black-box optimization (e.g., for software autotuning, automated algorithm configuration, automated experimental design).

License: BSD-3-Clause

Additional resources