Zero-Shot Neural Architecture Search: Challenges, Solutions, and Opportunities

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Guihong, Hoang, Duc, Bhardwaj, Kartikeya, Lin, Ming, Wang, Zhangyang, Marculescu, Radu
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914838634037248
author Li, Guihong
Hoang, Duc
Bhardwaj, Kartikeya
Lin, Ming
Wang, Zhangyang
Marculescu, Radu
author_facet Li, Guihong
Hoang, Duc
Bhardwaj, Kartikeya
Lin, Ming
Wang, Zhangyang
Marculescu, Radu
contents Recently, zero-shot (or training-free) Neural Architecture Search (NAS) approaches have been proposed to liberate NAS from the expensive training process. The key idea behind zero-shot NAS approaches is to design proxies that can predict the accuracy of some given networks without training the network parameters. The proxies proposed so far are usually inspired by recent progress in theoretical understanding of deep learning and have shown great potential on several datasets and NAS benchmarks. This paper aims to comprehensively review and compare the state-of-the-art (SOTA) zero-shot NAS approaches, with an emphasis on their hardware awareness. To this end, we first review the mainstream zero-shot proxies and discuss their theoretical underpinnings. We then compare these zero-shot proxies through large-scale experiments and demonstrate their effectiveness in both hardware-aware and hardware-oblivious NAS scenarios. Finally, we point out several promising ideas to design better proxies. Our source code and the list of related papers are available on https://github.com/SLDGroup/survey-zero-shot-nas.
format Preprint
id arxiv_https___arxiv_org_abs_2307_01998
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Zero-Shot Neural Architecture Search: Challenges, Solutions, and Opportunities
Li, Guihong
Hoang, Duc
Bhardwaj, Kartikeya
Lin, Ming
Wang, Zhangyang
Marculescu, Radu
Machine Learning
Computer Vision and Pattern Recognition
Recently, zero-shot (or training-free) Neural Architecture Search (NAS) approaches have been proposed to liberate NAS from the expensive training process. The key idea behind zero-shot NAS approaches is to design proxies that can predict the accuracy of some given networks without training the network parameters. The proxies proposed so far are usually inspired by recent progress in theoretical understanding of deep learning and have shown great potential on several datasets and NAS benchmarks. This paper aims to comprehensively review and compare the state-of-the-art (SOTA) zero-shot NAS approaches, with an emphasis on their hardware awareness. To this end, we first review the mainstream zero-shot proxies and discuss their theoretical underpinnings. We then compare these zero-shot proxies through large-scale experiments and demonstrate their effectiveness in both hardware-aware and hardware-oblivious NAS scenarios. Finally, we point out several promising ideas to design better proxies. Our source code and the list of related papers are available on https://github.com/SLDGroup/survey-zero-shot-nas.
title Zero-Shot Neural Architecture Search: Challenges, Solutions, and Opportunities
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2307.01998