LIT: Learned Intermediate Representation Training for Model Compression

Animesh Koratana, Daniel Kang, Peter Bailis, Matei Zaharia

Abstract

Researchers have proposed a range of model compression techniques to reduce the computational and memory footprint of deep neural networks (DNNs). In this work, we introduce Learned Intermediate re...
BibTeX
@inproceedings{koratana2019lit,
  author = {Animesh Koratana and Daniel Kang and Peter Bailis and Matei Zaharia},
  title = {{LIT: Learned Intermediate Representation Training for Model Compression}},
  booktitle = {International Conference on Machine Learning},
  year = {2019},
}
Figure