Kubeflow (@kubeflow )


Bio Helping make ML on Kubernetes easy, portable and scalable, everywhere.
Tweets 564
Followers 5,0K
Following 50
Account created 20-10-2017 13:54:21
ID 921374026781556736

Twitter Web App : Have you been wondering how to continuously train and deploy your models? This post describes how we solved this problem 👏 without forcing our data scientists to give up Project Jupyter 👏 and still using common #devops patterns like #gitops.


TweetDeck : THREAD: Is it possible that Kubeflow pipeline is one of the best CI/CD tools for Kubernetes?

I spent some time playing with Kubernetes & Kubeflow pipelines, and they have one feature which is just great:

You can define the pipeline with real code!

TweetDeck : IBM is running a two day Kubeflow Dojo today and tomorrow, for folks who are looking to get started with Kubeflow, or looking for deeper dives in certain areas.

Please go through it virtually, or join live.


TweetDeck : Calling all MiniKF users! As we plan for Kubeflow 1.1, the Kubeflow Contributors are requesting input from MiniKF users using this 5 question survey. Your input will help us to develop & support MiniKF more effectively. We appreciate your insights! docs.google.com/forms/d/e/1FAI…

TweetDeck : 📢New series alert! In this first episode of #Kubeflow 101, I cover 3️⃣ principles that Kubeflow is built around: composability, portability, and scalability. @Kubeflow
is an OSS project containing ML tools and frameworks (and built using #kubernetes!) youtu.be/cTZArDgbIWw

TweetDeck : We have great news! A new version of #MiniKF on GCP and Vagrant is fresh out of the oven, featuring @Kubeflow 1.0 🎉

The new MiniKF comes with a brand new Volume Manager UI 🥳

Try it out now!


TweetDeck : Introducing Kubeflow's MPI Operator, which makes it easy to run allreduce-style distributed training on #Kubernetes:


by Yuan Tang
#machinelearning #mlops

TweetDeck : Excited to see KubeFlow 1.0 reach GA.

Kubeflow runs now on OpenShift 4 - kubeflow.org/docs/openshift/

IBM and Red Hat are contributing towards a Kubeflow operator for Operatorhub.io - developer.ibm.com/blogs/kubeflow…

TweetDeck : Kubeflow 1.0 is out 🎉 Congratulations to everyone! We are so proud to be part of the @Kubeflow community as a contributor from the early beginning! twitter.com/kubeflow/statu…

TweetDeck : ⚙️A look into the open source processes we created to manage the Kubeflow 1.0 release, from Arrikto's Josh Bottom and our Product Management group. medium.com/kubeflow/the-m…

TweetDeck : Learn how GPU-as-a-Service over #KubeFlow enables faster and fully automated #ML pipelines. Tightly integrated with
NVIDIA #DGX, dynamic GPU allocation and scale-out is now simple to use in your enterprise. Read yaron haviv's new post on Towards Data Science: towardsdatascience.com/gpu-as-a-servi…

TweetDeck : It's an honour to have been part of the Kubeflow community as a contributor since its inception. We're excited to see our joint effort on #KFServing with Google, IBM, Microsoft and Bloomberg form part of the 1.0 release. twitter.com/kubeflow/statu…