-
Notifications
You must be signed in to change notification settings - Fork 80
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
KubeDL 2022 Annual Review #265
Comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
KubeDL 2022 Annual Review
Table of Contents
Backgroud
KubeDL is a suite of Kubernentes controllers that enable running machine learning workloads on Kubernentes, such as model training and inferences, configuration tuning for model deployments.
Project website https://kubedl.io/
DevStats
CNCF devstats website
Docker Pulls: 6.7K
Stars: 344
Contributors and Maintainers
We currently have 18 contributors and 5 maintainers.
Adoption
KubeDL has been used at Alibaba to run machine learning training workloads at very large scale, and has also been used by many external companies. We cannot obtain official adoption statements at this point.
KubeDL has achieved great progress in adopting production usages. Many of the users use the project but have not contributed in community as much. We will aim to engage more community contributors next year.
KubeDL has also a number of contributors from research institutes and universities to innovate on forward-thinking features in ML infra to improve the scheduling and performance efficiency.
Project Past
KubeDL has added many significat features since last year.
KubeDL has also been under serveral summer of code projects, such as ASOC(link), xxx, where a number of students are actively participating.
KubeDL/Morphling project is honored to have a paper published at ACM SoCC 2021.
Current Goals
Project Features
Docs
Community
We focused a lot in enhancing the project quality to production grade, and continuously contribute our best practices in large production to the community. We have many users using KubeDL in production, but have not get many external contributors in community. We had short amount of resources. Next year, we will focus more on building more awareness of the project by writing more blogs, showcasing, and presentation in conferences, and hope to engage more users to contribute.
Incubation
KubeDL is not yet ready for incubation, though KubeDL has been used heavily at Alibaba in large scale. We would like to see more contributors and maintiners and more adopters.
The text was updated successfully, but these errors were encountered: