This repo provides the supernet of S1 and our confirmatory experiments on NAS-Bench-101.
Python >= 3.6, Pytorch >= 1.0.0, torchvision >= 0.2.0
CIFAR-10 can be automatically downloaded by torchvision
. It has 50,000 images for
training and 10,000 images for validation.
python S1/train_search.py \
--exp_name experiment_name \
--m number_of_paths[1,2,3,4]
--data_dir /path/to/dataset \
--seed 2020 \
python NasBench101/nas_train_search.py \
--exp_name experiment_name \
--m number_of_paths[1,2,3,4]
--data_dir /path/to/dataset \
--seed 2020 \
@article{chu2020mixpath,
title={MixPath: A Unified Approach for One-shot Neural Architecture Search},
author={Chu, Xiangxiang and Li, Xudong and Lu, Yi and Zhang, Bo and Li, Jixiang},
journal={arXiv preprint arXiv:2001.05887},
url={https://arxiv.org/abs/2001.05887},
year={2020}
}