Sktime Example

This example trains a Sktime NaiveForecaster model using the Longley dataset for forecasting with exogenous variables. It shows a custom model type implementation that logs the training hyper-parameters, evaluation metrics and the trained model as an artifact.

Running the code

Run the train.py module to create a new MLflow experiment and to compute interval forecasts loading the trained model in native sktime flavor and pyfunc flavor:

python train.py

To view the newly created experiment and logged artifacts open the MLflow UI:

mlflow server

Model serving

This section illustrates an example of serving the pyfunc flavor to a local REST API endpoint and subsequently requesting a prediction from the served model. To serve the model run the command below where you substitute the run id printed during execution of the train.py module:

mlflow models serve -m runs:/<run_id>/model --env-manager local --host 127.0.0.1

Open a new terminal and run the score_model.py module to request a prediction from the served model (for more details read the MLflow deployment API reference):

python score_model.py

Running the code as a project

You can also run the code as a project as follows:

mlflow run .

Running unit tests

The test_sktime_model_export.py module includes a number of tests that can be executed as follows:

pytest test_sktime_model_export.py

While these tests will depend on the specifics of each individual flavor and in particular the design of the model wrapper interface (e.g. _SktimeModelWrapper), the above module can provide some orientation for the type of tests that can be useful when creating a new custom model flavor.