Back To Schedule
Monday, May 20 • 5:40pm - 6:00pm
Transfer Learning for Performance Modeling of Deep Neural Network System

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Modern deep neural network (DNN) systems are highly configurable with large a number of options that significantly affect their non-functional behavior, for example inference time and energy consumption. Performance models allow to understand and predict the effects of such configuration options on system behavior, but are costly to build because of large configuration spaces. Performance models from one environment cannot be transferred directly to another; usually models are rebuilt from scratch for different environments, for example different hardware. Recently, transfer learning methods have been applied to reuse knowledge from performance models trained in one environment in another. In this paper, we perform an empirical study to understand the effectiveness of different transfer learning strategies for building performance models of DNN systems. Our results show that transferring information on the most influential configuration options and their interactions is an effective way of reducing the cost to build performance models in new environments.


Md Shahriar Iqbal

University of South Carolina

Lars Kotthoff

University of Wyoming

Pooyan Jamshidi

University of South Carolina

Monday May 20, 2019 5:40pm - 6:00pm PDT
Lawrence/San Tomas/Lafayette Rooms