Lossy and Lossless (L$^2$) Post-training Model Size Compression

TL;DR

This paper introduces a unified post-training compression method combining lossy and lossless techniques to significantly reduce neural network size with minimal accuracy loss.

Contribution

It proposes a novel parametric weight transformation and differentiable counter for joint lossy and lossless compression, enabling controlled and efficient model size reduction.

Findings

Achieves up to 20x compression with minor accuracy loss

Stable 10x compression without accuracy sacrifice

Allows adaptive layer-wise compression ratios

Abstract

Deep neural networks have delivered remarkable performance and have been widely used in various visual tasks. However, their huge size causes significant inconvenience for transmission and storage. Many previous studies have explored model size compression. However, these studies often approach various lossy and lossless compression methods in isolation, leading to challenges in achieving high compression ratios efficiently. This work proposes a post-training model size compression method that combines lossy and lossless compression in a unified way. We first propose a unified parametric weight transformation, which ensures different lossy compression methods can be performed jointly in a post-training manner. Then, a dedicated differentiable counter is introduced to guide the optimization of lossy compression to arrive at a more suitable point for later lossless compression. Additionally, our method can easily control a desired global compression ratio and allocate adaptive ratios for different layers. Finally, our method can achieve a stable $10\times$ compression ratio without sacrificing accuracy and a $20\times$ compression ratio with minor accuracy loss in a short time. Our code is available at https://github.com/ModelTC/L2_Compression .