Using Containers for Deep Learning Workflows

Level: 300

With different versions of frameworks, libraries, and drivers for CPUs and GPUs, developers and data scientists spend a lot of time ensuring deep learning software stacks work well together during upgrades and system changes. In this tech talk, we'll take a look at how container technologies can address these challenges by providing training and inference environments that are lightweight, portable, consistent, and scalable. Through code examples, we'll take a closer look at how to integrate AWS Deep Learning Containers into your development and deployment workflows, as well as how to run large scale deep learning workloads on Amazon EKS.