Bright deploys and manages the entire stack from bare metal all the way to the application layer, and it provides an intuitive installer that makes it simple to deploy both Docker and Kubernetes. Why wade through endless documentation?
This paper shows how to run three popular deep learning frameworks PyTorch, Tensorflow, and MXNet in Docker containers scheduled by Kubernetes, and in Singularity containers scheduled through an HPC workload manager.
Download the white paper to learn how to run them as Kubernetes pods and jobs, how to pass an input script to a job using a configMap, and how to build a custom version of an NGC container image.
To access the white paper, please fill out the form to the right.