Using Kubernetes for Machine Learning Frameworks

GOTO Chicago 2019

Kubernetes provides isolation, auto-scaling, load balancing, flexibility and GPU support. These features are critical to run computationally and data intensive and hard to parallelize machine learning models. Declarative syntax of Kubernetes deployment descriptors make it easy for non-operationally focused engineers to easily train machine learning models on Kubernetes.

This talk will explain why and how Kubernetes is well suited for training and running your machine learning models in production. Specifically it will show how to setup a variety of open source machine learning frameworks such as TensorFlow, Apache MXNet and Pytorch on a Kubernetes cluster.

Attendees will learn training, massaging and inference phases of setting up a Machine Learning framework on Kubernetes. Attendees will leave with a GitHub repo of fully working samples.

Who should attend this talk: Developers, architects

Academic level: Intermediate

What is the take away in this talk: Attendees will learn why Kubernetes is well suited to run machine learning framework. They will see how the most popular open source ML frameworks can run on Kubernetes. They will leave with a working sample using GitHub repos