Success: Thank you for
subscribing! You will begin receiving emails as new content is posted. You may unsubscribe any time by
clicking the link in the email.
Kubeflow news
Your weekly update of curated kubeflow news from across the web.
Canonical Kubeflow Operators
5 November 2020
Canonical, the publisher of Ubuntu, announces Kubeflow operators and packages.
Within the last week, Canonical announced two new technologies that aim at improving the Kubeflow experience:
Charmed Kubeflow – A set of Kubeflow charm operators, that leverage Juju OLM technology for lifecycle management of the applications inside Kubeflow. Read the announcement.
Lightweight Kubeflow bundles – two new packages of pre-selected applications from the Kubeflow bundle to fit desktop (Kubeflow lite) and edge scenarios (Kubeflow edge). Read blog post.
Subscribe to our weekly newsletter
Kubeflow 1.1 is out!
24 August 2020
New release, increased capabilities.
After 6 months since the release of 1.0, Kubeflow releases a new version with increased capabilities.
This new version has focused on improving ML Workflow Productivity, Isolation and Security, and GitOps. Here is a list of the enhanced features:
Fairing and Kale (Kubeflow Automated pipeLines Engine) for end-to-end workflows.
Katib hyperparameter tuning:
New frameworks & algorithms (goptuna with CMA-ES, DARTS, chocolate, hyperopt, skopt).
Flexible config & tuning options (new python SDK, experiments with undefined goal, new UI, new resume policy).
Install and operations to support GitOps, using blueprints and kpt primitives.
Isolation and security: multi-user Kubeflow pipelines, CVE scanning, and support for Google’s Private GKE and Anthos.
MXNet and XGBoost distributed training operators, simplify training on multiple nodes, and speeds model creation.
The Kubeflow 101 series of short videos is a great way to quickly get up to speed on Kubeflow concepts. This week, Stephanie Wong guides us through Hyperparameters and how you can use Katib to achieve the
Kubeflow with Amazon Sagemaker
Shashank Prasanna, Senior Developer Advocate at AWS walks viewers on how to get the best out of Kubeflow and Amazon Sagemaker, in the same stack.
Security breach
12 June 2020
Microsoft exposes attacks to Kubeflow deployments
Microsoft publishes report detailing series of attacks against clusters running Kubeflow with the purpose of mining cryptocurrencies. To ensure that you are on the safe side, follow the steps below:
1. When deploying Kubeflow, make sure that its dashboard isn’t exposed to the internet: check the type of the Istio ingress service by the following command and make sure that it is not a load balancer with a public IP:
kubectl get service istio-ingressgateway -n istio-system
2. Verify that the malicious container is not deployed in the cluster, through the following command:
kubectl get pods –all-namespaces -o jsonpath=”{.items[*].spec.containers[*].image}” | grep -i ddsfdfsaadfs
Amazon announced this week the possibility to configure Kubeflow Pipelines to run ML jobs with Amazon SageMaker.
This is yet another validation of Kubeflow as a widespread solution and reinforces the idea of ML workflows on Kubernetes. Read the post here.
Kubeflow 101
27 May 2020
Kubeflow in 3 minutes
Google launches Kubeflow 101, a series of short videos to demystify Kubeflow. Intro to Kubeflow. Watch an overview of Kubeflow in the first episode:
Intro to Kubeflow pipelines
In the second episode of Kubeflow 101, watch an introduction to Kubeflow pipelines, and the presentation of a new hosted version of Kubeflow
Weekend watch list
6 March 2020
CNCF introduction to Kubeflow
Arun Gupta, Senior Engineering Manager at Apple walks viewers through the benefits of Kubeflow and how to get started.
Securing Kubeflow
Talk by Barton Rhodes, Senior Machine Learning Engineer at DaVita Inc. This talk is motivated by Ian Coldwater during the Kubernetes Podcast ‘Attacking and Defending Kubernetes:
“Containers are only as secure as their runtimes, and their orchestration frameworks, and their kernels, and their operating systems, and everything else”
Building and monitoring Kubeflow
Kirk Kaiser, APM at Datadog, talks about how to build and monitor Kubeflow machine learning Services with examples for an art project.
Kubeflow 1.0 is here!
2 March 2020
The wait is over, it’s official, Kubeflow 1.0 is here.
Many of you have been waiting for Kubeflow to reach 1.0 to suggest it to your managers, put it in production or use it more often in business critical applications. With Kubeflow reaching the 1.0 stage you can now do this with confidence and knowledge that Kubeflow is ‘here to stay’.
Read more from Thea Lamkin, Open Source Strategist for AI/ML at Google, talking about this release in her blog post.
Digest #2020.02.03 – AI-Created Medicine All Set for Human Trials
3 February 2020
-xkcd comics-
Artificial intelligence-created medicine to be used on humans for first time – For the first time, a drug created by AI and Machine Learning has been approved for human trials. The drug was created by British start-up Exscientia in collaboration with Japanese pharmaceutical firm Sumitomo Dainippon Pharma. While typical drug development takes five years to get to trials, this drug took 12 months! The molecule AKA DSP-1181 was created by using algorithms that sifted through potential compounds, checking them against a huge database of parameters. The drug will be used to treat patients who have obsessive-compulsive disorder (OCD).
DIY object tracking system with TensorFlow – Thinking of a great machine learning project on your spare Raspberry Pi(s) + some DIY hardware? Check out this interesting tutorial to build a real-time object tracker using TensorFlow. The article walks you through setting up the hardware and the TensorFlow libraries needed to get you started with machine learning; it’s gonna be a fun weekend project!
Hyperparameter tuning with Keras Tuner – If your ML projects have struggled with hyperparameter tuning you need to try Keras Tuner. This article is a great way to get started with Keras Tuner with a step-by-step walkthrough. Also, it’s an open-source project so check out their Github repo if you’d like to report issues, changes or contribute.
OpenAI goes all-in on PyTorch – Elon Musk’s AI research firm announced that it will be moving to PyTorch for its new machine learning projects. Many of the projects and engineers at OpenAI are transitioning to PyTorch as their main framework. “Going forward we’ll primarily use PyTorch as our deep learning framework but sometimes use other ones when there’s a specific technical reason to do so.” OpenAI is also expected to open-source its PyTorch binding for blocksparse kernels in the coming months.
Digest #2020.01.27 – Which Algorithm Should I Use?
27 January 2020
-Fossbytes-
Using AI to Enrich Digital Maps – Your commute may now be saved from the road closures, private routes and diversions! Researchers at MIT and Qatar Computing Research Institute (QCRI) have invented a model that could improve GPS navigation on digital maps.An interesting read about the research combining convolutional neural network (CNN) and graph neural network (GNN) to tag road features and predict the types of roads, lanes and obstruction along the route.
Solving Problems with TensorFlow – A good article on solving practical problems with TensorFlow. The article walks through solving an optimization problem, solving a linear regression problem, and a “Hello World” of Deep Learning classification projects with the MNIST Dataset. Check out for an interesting “getting started with tensorFlow” and train a neural network with visual representations and some fun.
Which Machine Learning Algorithm Should I Use? – One of the most common questions in the machine learning space; people are often confused about their ML needs and do not know the right qualifying questions to ask. The Microsoft team provides a walkthrough on how to decide the answer to which machine algorithm to use. It usually comes down to what are the requirements for your scenario? And, what do you want to do with your data? There are some neat cheat sheets for reference.
Data Sampling Methods for Imbalanced Classification – At some point we’ve all probably been thrown off by over-optimistic performance of machine learning solutions. The author walks through the challenges of machine learning with imbalanced classifications, a tutorial on sampling techniques to handle skewed class distributions, and explanations of undersampling, oversampling and a hybrid of both. A well-put tutorial for learning!
Digest #2020.01.20 – Machine Learning for Music, Video, Weather…
Microsoft released version 1.3 of their NNI project. NNI is Microsoft’s Neural Network Intelligence project, it lets you search for the best neural network architecture and hyperparameters.
Microsoft NNI supports Kubeflow, where NNI can take the place of Katib, the native hyperparameter tuner in Kubeflow. While there are clear efficiency related benefits of using Kubernetes native Katib, for users that are using Azure Kubernetes, NNI now gives another option.
NNI Docs – How NNI works with Kubeflow and Kubernetes
TensorFlow for Video Analytics – With the ever-increasing importance of object detection the application of machine learning is attracting a lot of interest. The article discusses how TensorFlow is used to identify objects like humans, puppies, faces, etc. The article is a good getting started tutorial for people looking to get their hands dirty with machine learning.
Using Machine Learning for Weather Prediction – Machine learning being prevalent is not going to miss out on something as important as the weather! Check out this AI research piece from Google on how to improve weather forecasts with real-time data using machine learning compared to existing techniques with ~3 hours of latency. The improvements can lead to significantly better forecasts on weather events that are time-sensitive. You may be able to take that picnic in the sun without rain showers.
Spotify Personalization with Machine Learning – Ever wondered how Spotify gets most of the song recommendation on your Spotify Home right? With machine learning, yes! Spotify engineering delivers and improves its recommendation system using a standardized ML framework including TensorFlow Extended, Kubeflow and the Google Cloud Platform Ecosystem. There is also a presentation by Tony Jebara on the topic that can be viewed here.
