A drop-in replacement for Scikit-Learn’s GridSearchCV / RandomizedSearchCV -- but with cutting edge hyperparameter tuning techniques.
tune-sklearn
is no longer being maintainedThis release (0.5.0
) is the last version of the library that will be released by the Ray team. It has been updated to be compatible
with ray>=2.7.x, ray<=2.9.x
. The library will not be guaranteed to work with future Ray versions.
The recommended alternative to keep up with the latest features of Ray is to migrate tune-sklearn
usage
to the Ray Tune APIs to accomplish the same thing.
Feel free to post an issue on the Ray Github if you run into any issues in migrating.
ExperimentAnalysis
usage for Ray 2.7 by @justinvyu in https://github.com/ray-project/tune-sklearn/pull/272
0.5.0
and update Ray Tune lower bound to 2.7.1
by @justinvyu in https://github.com/ray-project/tune-sklearn/pull/281
Full Changelog: https://github.com/ray-project/tune-sklearn/compare/v0.4.6...v0.5.0
error_score
argument being ignored by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/259
mlflow
integration from new air.integrations
module by @justinvyu in https://github.com/ray-project/tune-sklearn/pull/264
0.4.6
by @justinvyu in https://github.com/ray-project/tune-sklearn/pull/268
Full Changelog: https://github.com/ray-project/tune-sklearn/compare/v0.4.5...v0.4.6
fit_params
, use with_parameters
for it by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/256
Full Changelog: https://github.com/ray-project/tune-sklearn/compare/v0.4.4...v0.4.5
early_stopping=True
with BOHB not setting the correct scheduler by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/254
The minimum supported Ray version is 2.0.0 now.
Full Changelog: https://github.com/ray-project/tune-sklearn/compare/v0.4.3...v0.4.4
build_doc
by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/242
Full Changelog: https://github.com/ray-project/tune-sklearn/compare/v0.4.2...v0.4.3
max_concurrency
from BOHB kwargs by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/220
sk_n_jobs
, fix cloning issues by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/225
early_stopping
by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/229
os.path.expanduser
(defer to Ray Tune) by @krfricke in https://github.com/ray-project/tune-sklearn/pull/231
ray_context
by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/236
tune.with_parameters
by @Yard1 in https://github.com/ray-project/tune-sklearn/pull/237
Full Changelog: https://github.com/ray-project/tune-sklearn/compare/v0.4.1...v0.4.2
Releasing a new version of tune-sklearn! This version should be compatible with both Ray master and Ray 1.6.
This is a maintenance release without any public-facing changes.
Changelog:
TuneBaseSearchCV.fit
(https://github.com/ray-project/tune-sklearn/pull/215)Releasing a new version of tune-sklearn! This version should be compatible with both Ray master and Ray 1.4.
Changelog:
tune.run
params in fit
using the tune_params
argument (https://github.com/ray-project/tune-sklearn/pull/212)_is_param_distributions_all_tune_domains
(https://github.com/ray-project/tune-sklearn/pull/209)Releasing a new version of tune-sklearn! This version should be compatible with both Ray master and Ray 1.3.
Changelog:
name
to your run args (#200) (thanks, @rspeare!)