sklearn 0.24 compatibility

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

sklearn 0.24 compatibility #171

Closed Yard1 closed 3 years ago

Yard1 commented 3 years ago

This changes the name of self.num_samples attribute to self.n_trials in TuneSearchCV in order to fix an exception with scikit-learn>=0.24.0. Backwards compatible with previous scikit-learn releases. API compatibility is technically broken, but I don't think anyone would access num_samples from a TuneSearchCV object directly.

The CI fails due to an issue with CatBoost and not tune-sklearn (looks like they were taken by surprise by the sklearn update too!).

Yard1 commented 3 years ago

Looks like the catboost example is now causing CI to fail