Kubeflow

Kubeflow is an open-source platform for machine learning and MLOps on Kubernetes

introduced by Google. The different stages in a typical machine learning lifecycle are Kubeflow
represented with different software components in Kubeflow, including model
development (Kubeflow Notebooks[4]), model training (Kubeflow Pipelines,[5] Kubeflow
Training Operator[6]), model serving (KServe[a][7]), and automated machine learning
(Katib[8]).
Original author(s) Google
Each component of Kubeflow can be deployed separately, and it is not a requirement to
deploy every component.[9] Developer(s) Kubeflow
Contributors[1]
- AWS,
History Bloomberg,
Google, IBM,
The Kubeflow project was first announced at KubeCon + CloudNativeCon North
America 2017 by Google engineers David Aronchick, Jeremy Lewi, and Vishnu NVIDIA,
Kannan[10] to address a perceived lack of flexible options for building production-ready Nutanix, Red
machine learning systems.[11] The project has also stated it began as a way for Google to Hat, Arrikto,
open-source how they ran TensorFlow internally.[12] and others
Initial release April 5, 2018[2]
The first release of Kubeflow (Kubeflow 0.1) was announced at KubeCon +
Stable release 1.7[3] /
CloudNativeCon Europe 2018[13] with claims of having already become among the top
March 29,
2% of GitHub projects ever.[14] Kubeflow 1.0 was released in March 2020 via a public
blog post announcing that many Kubeflow components were graduating to a "stable 2023
status", indicating they were now ready for production usage.[15] Repository github.com
/kubeflow (http
Components s://github.com/
kubeflow)
Written in Go, Python
Kubeflow Notebooks for model development Platform Kubernetes

Machine learning models are developed in the notebooks component called Kubeflow Type Machine
Notebooks. The component runs web-based development environments inside a Learning
Kubernetes cluster, with native support for Jupyter Notebook, Visual Studio Code, and Platform
RStudio.[16] License Apache
License 2.0
Kubeflow Pipelines for model training Website kubeflow.org
(https://kubeflo
Once developed, models are trained in the Kubeflow Pipelines component. The w.org)
component acts as a platform for building and deploying portable, scalable machine
learning workflows based on Docker containers.[17] Google Cloud Platform has adopted the Kubeflow Pipelines DSL within its
Vertex AI Pipelines product.[18]

Kubeflow Training Operator for model training

For certain machine learning models and libraries, the Kubeflow Training Operator component provides Kubernetes custom
resources support. The component runs distributed or non-distributed TensorFlow, PyTorch, Apache MXNet, XGBoost, and MPI
training jobs on Kubernetes.[6]

KServe for model serving

The KServe component (previously named KFServing[19]) provides Kubernetes custom resources for serving machine learning
models on arbitrary frameworks including TensorFlow, XGBoost, scikit-learn, PyTorch, and ONNX.[20] KServe was developed
collaboratively by Google, IBM, Bloomberg, NVIDIA, and Seldon.[19] Publicly disclosed adopters of KServe include
Bloomberg,[21] Gojek,[22] and others.[23]

Katib for automated machine learning

Lastly, Kubeflow includes a component for automated training and development of machine learning models, the Katib
component. It is described as a Kubernetes-native project and features hyperparameter tuning, early stopping, and neural
architecture search.[24]

Release timeline
Release timeline

Release
Version Release Information Release Blog
Date

Kubeflow 5 April, https://kubernetes.io/blog/2018/05/04/announcing-kubeflow-

-
0.1 2018[2] 0.1/

Kubeflow 2 July, https://medium.com/kubeflow/kubeflow-0-2-offers-new-

-
0.2 2018[25] components-and-simplified-setup-735e4c56988d

Kubeflow 5 October, https://medium.com/kubeflow/kubeflow-0-3-simplifies-setup-

-
0.3 2018[26] improves-ml-development-98b8ca10bd69

8 January, https://medium.com/kubeflow/kubeflow-0-4-release-
Kubeflow
- enhancements-for-machine-learning-productivity-
0.4 2019[27] d77c54df07a9
https://medium.com/kubeflow/kubeflow-v0-5-simplifies-
Kubeflow 9 April,
- model-development-with-enhanced-ui-and-fairing-library-
0.5 2019[28] 78e19cdc9f50

19 July, https://medium.com/kubeflow/kubeflow-v0-6-a-robust-
Kubeflow https://www.kubeflow.org/docs/releases/kubeflow-
foundation-for-artifact-tracking-data-versioning-multi-user-
0.6 2019[29] 0.6/
support-9896d329412c

17
Kubeflow https://www.kubeflow.org/docs/releases/kubeflow- https://medium.com/kubeflow/kubeflow-v0-7-delivers-beta-
October,
0.7 0.7/ functionality-in-the-leadup-to-v1-0-1e63036c07b8
2019[30]
20
Kubeflow https://www.kubeflow.org/docs/releases/kubeflow- https://blog.kubeflow.org/releases/2020/03/02/kubeflow-1-0-
February,
1.0 1.0/ cloud-native-ml-for-everyone
2020[31]

Kubeflow 31 July, https://www.kubeflow.org/docs/releases/kubeflow- https://blog.kubeflow.org/release/official/2020/07/31/kubeflow-

1.1 2020[32] 1.1/ 1.1-blog-post

18
Kubeflow November, https://www.kubeflow.org/docs/releases/kubeflow- https://blog.kubeflow.org/release/official/2020/11/18/kubeflow-
1.2 1.2/ 1.2-blog-post
2020[33]

Kubeflow 23 April, https://www.kubeflow.org/docs/releases/kubeflow-

https://blog.kubeflow.org/kubeflow-1.3-release/
1.3 2021[34] 1.3/

12
Kubeflow October, https://www.kubeflow.org/docs/releases/kubeflow-
https://blog.kubeflow.org/kubeflow-1.4-release/
1.4 1.4/
2021[35]

Kubeflow 10 March, https://www.kubeflow.org/docs/releases/kubeflow-

https://blog.kubeflow.org/kubeflow-1.5-release/
1.5 2022[36] 1.5/

7
Kubeflow September, https://www.kubeflow.org/docs/releases/kubeflow-
https://blog.kubeflow.org/kubeflow-1.6-release/
1.6 1.6/
2022[37]

Kubeflow 29 March, https://www.kubeflow.org/docs/releases/kubeflow-

https://blog.kubeflow.org/kubeflow-1.7-release/
1.7 2023[3] 1.7/

Notes
a. KServe was previously known as KFServing[19]

References
 1. "Kubeflow Website - Working Groups" (https://kubeflow.org/docs/about/community/#kubeflow-working-groups).
2. "Kubeflow 0.1 - Release Tag" (https://github.com/kubeflow/kubeflow/releases/tag/v0.1.0/). GitHub.
3. "Kubeflow 1.7 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.7/).
4. "Kubeflow Website - Kubeflow Notebooks" (https://www.kubeflow.org/docs/components/notebooks/).
5. "Kubeflow Website - Kubeflow Pipelines" (https://www.kubeflow.org/docs/components/pipelines/).
6. "Kubeflow GitHub - Kubeflow Training Operator" (https://github.com/kubeflow/training-operator). GitHub.
7. "Kubeflow Website - KServe" (https://www.kubeflow.org/docs/external-add-ons/kserve/).
8. "Kubeflow Website - Katib" (https://www.kubeflow.org/docs/components/katib/).
9. "Kubeflow Website - Installing Kubeflow" (https://www.kubeflow.org/docs/started/installing-kubeflow/).
10. " "Hot Dogs or Not" - At Scale with Kubernetes [I] - Vish Kannan & David Aronchick, Google" (https://www.youtub
e.com/watch?v=R3dVF5wWz-g). YouTube.
11. "Introducing Kubeflow - A Composable, Portable, Scalable ML Stack Built for Kubernetes" (https://kubernetes.io/b
log/2017/12/introducing-kubeflow-composable/). 21 December 2017.
12. "Kubeflow Website - History" (https://www.kubeflow.org/docs/started/introduction/#history).
13. "Google-led Kubeflow, machine learning for Kubernetes, begins to take shape" (https://techcrunch.com/2018/05/
04/google-kubeflow-machine-learning-for-kubernetes-begins-to-take-shape/). 4 May 2018.
14. "Announcing Kubeflow 0.1" (https://kubernetes.io/blog/2018/05/04/announcing-kubeflow-0.1/). 4 May 2018.
15. "Kubeflow 1.0: Cloud-Native ML for Everyone" (https://blog.kubeflow.org/releases/2020/03/02/kubeflow-1-0-cloud
-native-ml-for-everyone.html). 2 March 2020.
16. "Kubeflow Website - Kubeflow Notebooks Overview" (https://www.kubeflow.org/docs/components/notebooks/ove
rview/).
17. "Kubeflow Website - Kubeflow Pipelines Introduction" (https://www.kubeflow.org/docs/components/pipelines/intr
oduction/).
18. "Vertex AI - Building a pipeline" (https://cloud.google.com/vertex-ai/docs/pipelines/build-pipeline).
19. "KServe: The next generation of KFServing" (https://blog.kubeflow.org/release/official/2021/09/27/kfserving-transi
tion). 27 September 2021.
20. "KServe GitHub" (https://github.com/kserve/kserve). GitHub.
21. "The journey to build Bloomberg's ML Inference Platform Using KServe (formerly KFServing)" (https://www.bloo
mberg.com/company/stories/the-journey-to-build-bloombergs-ml-inference-platform-using-kserve-formerly-kfservi
ng/). Bloomberg L.p. 12 October 2021.
22. "Merlin: Making ML Model Deployments Magical" (https://www.gojek.io/blog/merlin-making-ml-model-deploymen
ts-magical).
23. "KServe Website - Adopters of KServe" (https://kserve.github.io/website/master/community/adopters/).
24. "Kubeflow GitHub - Katib" (https://github.com/kubeflow/katib). GitHub.
25. "Kubeflow 0.2 - Release Tag" (https://github.com/kubeflow/kubeflow/releases/tag/v0.2.0/). GitHub.
26. "Kubeflow 0.3 - Release Tag" (https://github.com/kubeflow/kubeflow/releases/tag/v0.3.0/). GitHub.
27. "Kubeflow 0.4 - Release Tag" (https://github.com/kubeflow/kubeflow/releases/tag/v0.4.0/). GitHub.
28. "Kubeflow 0.5 - Release Tag" (https://github.com/kubeflow/kubeflow/releases/tag/v0.5.0/). GitHub.
29. "Kubeflow 0.6 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-0.6/).
30. "Kubeflow 0.7 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-0.7/).
31. "Kubeflow 1.0 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.0/).
32. "Kubeflow 1.1 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.1/).
33. "Kubeflow 1.2 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.2/).
34. "Kubeflow 1.3 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.3/).
35. "Kubeflow 1.4 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.4/).
36. "Kubeflow 1.5 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.5/).
37. "Kubeflow 1.6 - Release Information" (https://www.kubeflow.org/docs/releases/kubeflow-1.6/).

External links
Official website (https://www.kubeflow.org/)
Kubeflow (https://github.com/kubeflow) on GitHub

Retrieved from "https://en.wikipedia.org/w/index.php?title=Kubeflow&oldid=1155411071"