Dream Distillation: A Data-Independent Model Compression Framework

05/17/2019

∙

by   Kartikeya Bhardwaj, et al.

∙

5

∙

share

Model compression is eminently suited for deploying deep learning on IoT-devices. However, existing model compression techniques rely on access to the original or some alternate dataset. In this paper, we address the model compression problem when no real data is available, e.g., when data is private. To this end, we propose Dream Distillation, a data-independent model compression framework. Our experiments show that Dream Distillation can achieve 88.5 original data!