Providing methods to efficiently benchmark hyperparameter optimization algorithms
Python Shell
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.

Surrogate Benchmarks: Cheap-to-evaluate, but realistic hyperparameter optimization benchmarks

This library provides tools to build and use surrogate benchmarks to evaluate and compare hyperparameter optimization methods. You can find the documentation here:

This software is licensed under the GPL license. You find the license text shipped with this package.

The code used for the Gaussian Process is part of the spearmint package, which can be found here:

Master branch

Documentation Documentation Status

Status: Build Status Code Health

Development branch

Documentation: Documentation Status

Status: Build Status Code Health