Managing the ML Model

MLflow integration

In this lesson, we will learn about using MLflow models.

MLflow is an open-source platform for managing the ML lifecycle. Azure Machine Learning is integrated with MLflow. Therefore, we can save the models in MLflow format, track the models using MLflow URI, and log the metrics and artifacts.

Running an MLflow job

Let’s run a simple job to learn about three important use cases for MLflow integration.

• Creating the model parameters: We can create model parameters using the mlflow.log_param function.
• Logging metrics: We can log any custom metric using the mlflow.log_metric function.
• Logging artifacts: The function mlflow.log_artifact allows us to log the artifacts in a file.

Let’s run the job below to understand these functionalities.