Max_tuning_runs – The maximum number of child Mlflow runs created for hyperparameter If False, show all events and warnings during scikit-learn Silent – If True, suppress all event logs and warnings from MLflow during scikit-learnĪutologging. Scikit-learn that have not been tested against this version of the MLflow If False, autologged content is logged to the active fluent run,ĭisable_for_unsupported_versions – If True, disable autologging for versions of If False,Įnables the scikit-learn autologging integration.Įxclusive – If True, autologged content is not logged to user-created fluent runs. Input examples and model signatures, which are attributes of MLflow models,Īre also omitted when log_models is False.ĭisable – If True, disables the scikit-learn autologging integration. Log_models – If True, trained models are logged as MLflow model artifacts. Note: Model signatures are MLflow model attributes
With scikit-learn model artifacts during training. Note: Input examples are MLflow model attributesĪnd are only collected if log_models is also True.ĭescribing model inputs and outputs are collected and logged along Logged along with scikit-learn model artifacts during training. Log_input_examples – If True, input examples from training datasets are collected and Metric APIs defined in the trics moduleįor post training metrics autologging, the metric key format is: Types of scikit-learn metric APIs are supported: Results and log them as MLflow metrics to the Run associated with the model. When users call metric APIs after model training, MLflow tries to capture the metric API
If the classifier has method predict_proba, we additionally log:
Note that the training score isĬomputed using parameters given to fit().
This means when you fit a meta estimator that chainsĪ series of estimators, the parameters of these child estimators are also logged.Ī training score obtained by estimator.score. Parameters obtained by estimator.get_params(deep=True). Autologging may not succeed when used with package versions outside of this range.Įnables (or disables) and configures autologging for scikit-learn estimators. Autologging is known to be compatible with the following package versions: 0.22.1 <= scikit-learn <= 1.1.2.
0 kommentar(er)
