BigNAS: Scaling Up Neural Architecture Search with Big Single-Stage Models

Yu Jiahui

[0]

Jin Pengchong

[0]

Liu Hanxiao (刘寒骁)

[0]

Bender Gabriel

[0]

Kindermans Pieter-Jan

Tan Mingxing

[0]

Huang Thomas (黄煦涛)

[0]

Song Xiaodan

Pang Ruoming (庞若鸣)

[0]

Le Quoc

[0]

european conference on computer vision, pp. 702-717, 2020.

Cited by: 26|Views157

Keywords:

different network architecturenas methodshot modelsuper networkneural networkMore(8+)

Weibo:

We presented a novel paradigm for neural architecture search by training a single-stage model, from which high-quality child models of different sizes can be induced for instant deployment without retraining or finetuning

Abstract:

Neural architecture search (NAS) has shown promising results discovering models that are both accurate and fast. For NAS, training a one-shot model has become a popular strategy to rank the relative quality of different architectures (child models) using a single set of shared weights. However, while one-shot model weights can effective...More

Code:

Data:

Full Text

View via Publisher