Does it makes sense to run AI workloads on Kubernetes? Pros, cons and how to start
Václav Pavlín, [email protected]
- https://docs.google.com/presentation/d/1ogop6lRQpY-4VGQ6VvoYQyL_E8WCfmo4114-tAjqLYs/edit?usp=sharing
Let's go over all the available components of Kubeflow and Open Data Hub projects and explore what technologies are available and what problems they can help you solve.
The journey of Open Data Hub and how we are turning KF into downstream enterprise grade distribution
JupyterHub vs. notebook-controller, multitenancy, extendability, scaling
An expert will walk through how to to use Jupyter Notebooks (not only on Kubernetes) for successful data analysis
Workflows and pipelines are core components of data analysis. What are our options on Kubernetes? How to create a simple pipeline? What are the caveats?
What tools do we have on Kubernetes to strategically track our experiments and build automated pipelines for dev to prod model deployment?
Open Source offers a great deal of AI/ML frameworks for model creation. Let’s see which of these have good support on Kubernetes and how to use them.
How can we move beyond a massive data center clusters to the edge for our AI workloads?