With support from the Kubeflow Project Steering Group, Google has just announced the application for Kubeflow to become an incubating project in the Cloud Native Computing Foundation (CNCF).
We in the Kubeflow community are excited about how far the project has come since its inception in 2017, and are blown away by the diverse community of contributors and users we have built together. Today, Kubeflow has close to 200 contributors from over 30 organizations. Over the past five years, the project has seen almost 20,000 GitHub stars, tens of thousands of commits, and hosted several user and contributor gatherings across the world.
What is Kubeflow
With the help of the project community, Kubeflow has developed into an end-to-end, extendable ML platform for Kubernetes, with mature components to address key stages of the ML lifecycle:
- model development (Kubeflow Notebooks)
- model training (Kubeflow Pipelines and Kubeflow Training Operator)
- model serving (KServe)
- automated machine learning (Katib)
Kubeflow’s powerful development experience helps data scientists all around the world build, train, and deploy their ML models, and gives MLOps teams the tools they need to deploy and scale advanced workflows in a variety of infrastructures.
In May 2020, with the v1.0 release, Kubeflow reached maturity across a core set of its stable applications. In 2022, we also graduated Kubeflow Serving as an independent project, KServe, which now, is incubating in Linux Foundation AI & Data.
The six releases since v1.0 have truly been a community effort. The community, collaborating together in 16 Working Groups, brought to project a number of key improvements in ML workflow simplicity, resource efficiency, security, and tooling. The wider Kubeflow ecosystem has also added 10 Kubeflow distributions across multiple Cloud and on-prem environments.
Our application for Kubeflow to become a CNCF incubating project is the next big step for the project. We’re excited to see Kubeflow get closer to its roots in the cloud-native ecosystem, and for the community to continue driving innovation in the MLOps and data science experiences.
The pull request we’ve opened today to join CNCF as an incubating project is only the first step. Google and the Kubeflow community will work with the CNCF and their Technical Oversight Committee (TOC), to meet the incubation stage requirements. While the due diligence and eventual TOC decision will take some time, the Kubeflow project will continue developing and releasing throughout this process.
If Kubeflow is accepted into CNCF, the project will transfer to the CNCF its source code, trademark, website, and other collaboration assets and social media accounts. As part of the donation, we will also work with the Kubeflow community to update the project governance to ensure diverse representation of the broader community. Open source comes with the responsibility to contribute, sustain, and improve the projects to make our ecosystem better. We at Google will continue to support Kubeflow and participate in the Kubeflow community as project maintainers, contributors, and users.
A huge thank you to everyone who has contributed to Kubeflow over the years! We are excited for what lays ahead.
By Thea Lamkin, Senior Program Manager, Google, on behalf of the Kubeflow Project Steering Group