This charter adheres to the conventions, roles and organization management outlined in wg-governance.
WG AutoML is responsible for all aspects of Automated Machine Learning technologies on Kubeflow. The WG covers researching, developing and operating various targets of ML automation for Kubeflow.
- APIs and tools used for running AutoML experiments (e.g. Experiment, Suggestion, Trial CRD/Controller).
- Web UI to enhance user experience with AutoML techniques.
- Database manager to analyse and store experiments metrics.
- Documentation for detailed description of using AutoML in Kubeflow.
- Coordinating with Training WG to make sure that all distributed training jobs can be used in AutoML experiments.
- Coordinating with Manifests WG to ensure that AutoML manifests are properly deployed with Kubeflow.
- Coordinating with Central Dashboard WG to correct integration with AutoML UI.
- Coordinating with release teams to ensure that the AutoML features can be released properly.
- APIs used for running Training Jobs (Trials) (this is related to Training WG).
This WG follows adheres to the Roles and Organization Management outlined in wg-governance and opts-in to updates and modifications to wg-governance.
WG Technical Leads.