Graph is all you need? Lightweight data-agnostic neural architecture search without training

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Huang, Zhenhan, Pedapati, Tejaswini, Chen, Pin-Yu, Jiang, Chunheng, Gao, Jianxi
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908413906124800
author Huang, Zhenhan
Pedapati, Tejaswini
Chen, Pin-Yu
Jiang, Chunheng
Gao, Jianxi
author_facet Huang, Zhenhan
Pedapati, Tejaswini
Chen, Pin-Yu
Jiang, Chunheng
Gao, Jianxi
contents Neural architecture search (NAS) enables the automatic design of neural network models. However, training the candidates generated by the search algorithm for performance evaluation incurs considerable computational overhead. Our method, dubbed nasgraph, remarkably reduces the computational costs by converting neural architectures to graphs and using the average degree, a graph measure, as the proxy in lieu of the evaluation metric. Our training-free NAS method is data-agnostic and light-weight. It can find the best architecture among 200 randomly sampled architectures from NAS-Bench201 in 217 CPU seconds. Besides, our method is able to achieve competitive performance on various datasets including NASBench-101, NASBench-201, and NDS search spaces. We also demonstrate that nasgraph generalizes to more challenging tasks on Micro TransNAS-Bench-101.
format Preprint
id arxiv_https___arxiv_org_abs_2405_01306
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Graph is all you need? Lightweight data-agnostic neural architecture search without training
Huang, Zhenhan
Pedapati, Tejaswini
Chen, Pin-Yu
Jiang, Chunheng
Gao, Jianxi
Machine Learning
Neural architecture search (NAS) enables the automatic design of neural network models. However, training the candidates generated by the search algorithm for performance evaluation incurs considerable computational overhead. Our method, dubbed nasgraph, remarkably reduces the computational costs by converting neural architectures to graphs and using the average degree, a graph measure, as the proxy in lieu of the evaluation metric. Our training-free NAS method is data-agnostic and light-weight. It can find the best architecture among 200 randomly sampled architectures from NAS-Bench201 in 217 CPU seconds. Besides, our method is able to achieve competitive performance on various datasets including NASBench-101, NASBench-201, and NDS search spaces. We also demonstrate that nasgraph generalizes to more challenging tasks on Micro TransNAS-Bench-101.
title Graph is all you need? Lightweight data-agnostic neural architecture search without training
topic Machine Learning
url https://arxiv.org/abs/2405.01306