17:00 - 17:45
Track С
ABOUT THE SPEAKER
Is an engineer and researcher in the fields of deep learning and computer vision with over 10 years of experience. Dmitrii has been designing neural networks and training pipelines for the last 5 years. He has a record of deploying neural networks for edge
analytics, robotics, and autonomous driving to production in high-performance and simultaneously high-reliability scenarios.
Talk: How to launch neural network training on Microsoft Azure
Modern deep learning tasks require lots of GPU compute for neural network training. Assembling an on-premise GPU server is not nearly fast as launching a cloud GPU instance. In this talk, we will have a step-by-step walk through the process of launching a GPU machine on Azure and train our first neural network on it.