What web browser should I use?
The Open edX platform works best with current versions of Chrome, Edge, Firefox, Internet Explorer, or Safari.
See our list of supported browsers for the most up-to-date information.
AutoML with Kubeflow and Kale
In this course, you will create an AutoML workflow that includes Machine Learning techniques, Hyperparameter Tuning and Meta-Learning, starting from your notebook. By completing this course, you will learn the basic concepts of AutoML. More specifically, you will discover and train models in an automated manner and optimize the best performing one using hyperparameter tuning, as steps of the same workflow. Then, you will serve your model from inside your notebook using Kale and KF Serving. As a final step, you will create a TensorBoard server to view the logs that the training of the model produced.
Course Number
CourseSelf-Paced
About This Course
In this course, you will start from creating a notebook, where you will call a Kale API providing a dataset and a task. Then Kale takes over, you can sit back and relax, and watch the following process unfold as Kale
- Uses Meta-Learning to analyze the input dataset and suggests some model architectures that should perform well.
- Starts KFP pipelines to train these model architectures.
- Collects all the results from the previous step. As soon as all pipelines are done, it retrieves the best performing one, based on the target metric.
- Starts a new Katib HP Tuning experiment on the best model, to further optimize its initialization parameters.
- Continually logs models, datasets, and TensorBoard reports to MLMD, using reproducible Rok snapshots. You can have a complete lineage of your experiment and everything is persisted into immutable Rok snapshots.
Finally, once all of this is done, you can go back to your notebook and, with a single Kale API call, you will
- Serve any of the trained models using KFServing.
- Start a TensorBoard server to analyze any of the trained models.