Speaker: Timothy Liu
Managing the hardware, drivers, libraries and packages that make up a ML development environment can be hard. In this talk, I will introduce how Docker can be used to simplify the process of setting up a local ML development environment, and how we can use Kubernetes and Kubeflow to scale that standardised environment to provide scalable, web-based Jupyter environments for a large number of users, that can be served from both public cloud providers and from on-premise clusters.
Timothy is an undergraduate student at the Singapore University of Technology and Design and an intern at NVIDIA. He is passionate about making ML development more accessible and easier to manage.
Produced by Engineers.SG
Recorded by: Ambrose Chua
Help us caption & translate this video!