ray.train.sklearn.SklearnCheckpoint.from_estimator#

classmethod SklearnCheckpoint.from_estimator(estimator: sklearn.base.BaseEstimator, *, path: os.PathLike, preprocessor: Optional[Preprocessor] = None) → SklearnCheckpoint[source]#

Create a Checkpoint that stores an sklearn Estimator.

Parameters

  • estimator – The Estimator to store in the checkpoint.

  • path – The directory where the checkpoint will be stored.

  • preprocessor – A fitted preprocessor to be applied before inference.

Returns

An SklearnCheckpoint containing the specified Estimator.

Examples

>>> from ray.train.sklearn import SklearnCheckpoint
>>> from sklearn.ensemble import RandomForestClassifier
>>>
>>> estimator = RandomForestClassifier()
>>> checkpoint = SklearnCheckpoint.from_estimator(estimator, path=".")

You can use a SklearnCheckpoint to create an SklearnPredictor and preform inference.

>>> from ray.train.sklearn import SklearnPredictor
>>>
>>> predictor = SklearnPredictor.from_checkpoint(checkpoint)