registry

Note

See the Glossary for the meaning of the acronyms used in this guide.

art.py

A task plugin module for interfacing the Adversarial Robustness Toolbox with the MLFlow model registry.

load_wrapped_tensorflow_keras_classifier(name: str, version: int, imagenet_preprocessing: bool = False, classifier_kwargs: Optional[Dict[str, Any]] = None) → art.estimators.classification.TensorFlowV2Classifier

Loads and wraps a registered Keras classifier for compatibility with the Adversarial Robustness Toolbox.

Parameters

• name – The name of the registered model in the MLFlow model registry.

• version – The version number of the registered model in the MLFlow registry.

• classifier_kwargs – A dictionary mapping argument names to values which will be passed to the TensorFlowV2Classifier constructor.

Returns

A trained TensorFlowV2Classifier

object.

See also

• art.estimators.classification.TensorFlowV2Classifier

• mlflow.load_tensorflow_keras_classifier()

mlflow.py

A task plugin module for using the MLFlow model registry.

add_model_to_registry(active_run: mlflow.entities.Run, name: str, model_dir: str) → Optional[mlflow.entities.model_registry.ModelVersion]

Registers a trained model logged during the current run to the MLFlow registry.

Parameters

• active_run – The mlflow.ActiveRun object managing the current run’s state.

• name – The registration name to use for the model.

• model_dir – The relative artifact directory where MLFlow logged the model trained during the current run.

Returns

A ModelVersion object created by the backend.

get_experiment_name(active_run: mlflow.entities.Run) → str

Gets the name of the experiment for the current run.

Parameters

active_run – The mlflow.ActiveRun object managing the current run’s state.

Returns

The name of the experiment.

load_tensorflow_keras_classifier(name: str, version: int) → tensorflow.keras.models.Sequential

Loads a registered Keras classifier.

Parameters

• name – The name of the registered model in the MLFlow model registry.

• version – The version number of the registered model in the MLFlow registry.

Returns

A trained tf.keras.Sequential object.