Starting a run
MLflow uses the concept of a run as a way to organize model training. Before metrics, parameters or artifacts can be logged to MLflow Tracking, a run must become active. Each run must also belong to an existing experiment.
In the following exercise, you will start a new run so that you can begin logging a model. You will also set the experiment in which you would like the run to be logged. The mlflow
module will already be imported for you.
This exercise is part of the course
Introduction to MLflow
Exercise instructions
- Set the experiment so that the active run tracks to
"Unicorn Sklearn Experiment"
experiment. - Start a new run with
mlflow
module.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Set the experiment
____.____("____")
# Start a run
____.____()