Searching by Generating: Flexible and Efficient One-Shot NAS with Architecture Generator

1 code implementation • CVPR 2021 • Sian-Yao Huang, Wei-Ta Chu

With the pre-trained supernt, the search time of SGNAS for $N$ different hardware constraints is only 5 GPU hours, which is $4N$ times faster than previous SOTA single-path methods.

PONAS: Progressive One-shot Neural Architecture Search for Very Efficient Deployment

1 code implementation • 11 Mar 2020 • Sian-Yao Huang, Wei-Ta Chu

During search, we evaluate candidate blocks in different layers and construct the accuracy table that is to be used in deployment.

Neural Architecture Search