The Kubeflow makes AI/ML on Kubernetes simple, portable and scalable.
Please refer to the official docs at kubeflow.org.
The Kubeflow ecosystem is composed of multiple open-source projects for each stage in the ML lifecycle.
Learn more about each project in the Kubeflow documentation.
Please use the following GitHub repositories to open issues and pull requests for the different Kubeflow components:
| Component | Source Code |
|---|---|
| KServe | kserve/kserve |
| Kubeflow Katib | kubeflow/katib |
| Kubeflow Model Registry | kubeflow/model-registry |
| Kubeflow MPI Operator | kubeflow/mpi-operator |
| Kubeflow Notebooks | kubeflow/notebooks |
| Kubeflow Pipelines | kubeflow/pipelines |
| Kubeflow Spark Operator | kubeflow/spark-operator |
| Kubeflow Training Operator | kubeflow/training-operator |
If you want to open issue or pull request for the Kubeflow Platform components:
- Use the
kubeflow/manifestsGitHub repository for the Kubeflow Manifests. - Use the
kubeflow/dashboardGitHub repository for the Kubeflow Profile Controller, Central Dashboard, CRUD Web Apps, PVC Viewer, PodDefault, and Access Management components.
If you have questions about Kubeflow community or Kubeflow ecosystem, please use the
kubeflow/community GitHub repository.
Every Kubeflow component is maintained by Kubeflow working group (WG) leads.
Learn more about Kubeflow community and WG in this guide.