Feature request: use number of samples as resource for early stopping

ray-project / tune-sklearn

A drop-in replacement for Scikit-Learn’s GridSearchCV / RandomizedSearchCV -- but with cutting edge hyperparameter tuning techniques.

https://docs.ray.io/en/master/tune/api_docs/sklearn.html

Apache License 2.0

465 stars 52 forks source link

Feature request: use number of samples as resource for early stopping #120

Open awqb opened 4 years ago

awqb commented 4 years ago

Successive halving has been introduced in scikit-learn, and the default resource is the number of data samples used to train the model. It would be great to have this option in tune-sklearn as well, because it would mean that early stopping could be used in combination with any scikit-learn compliant estimator.