Pytorch-Single-Path-One-Shot-NAS

Single Path One-Shot NAS Pytorch implementation, include Supernet training and searching(sample)

Thanks for the great work of MXNet-Single-Path-One-Shot-NAS, i basiclly follow it.

Model	Official	MXNet repo	this repo
Subnet Training	√	√	√
Block Selection	×	√	√
Channel Selection	×	√	√
Supernet Training	×	√	×
FLOPs & Parameters Counting Tool	×	√	√
BN Statistics Update	×	√	√
Random Search	×	√	√
Genetic Search	×	√	√
SE	-	√	√
Efficient Last Conv Block	-	√	×
Op to Op Profiling Tool	-	√	×
Merge BN	-	√	×
Int8 Quantization	-	√	×

Citation

If you use these models in your research, please cite:

@article{guo2019single,
        title={Single path one-shot neural architecture search with uniform sampling},
        author={Guo, Zichao and Zhang, Xiangyu and Mu, Haoyuan and Heng, Wen and Liu, Zechun and Wei, Yichen and Sun, Jian},
        journal={arXiv preprint arXiv:1904.00420},
        year={2019}
}

@misc{yan2019sposmxnet,
      title={single-path-one-shot-mxnet},
      author={Yan, Kang},
      howpublished={\url{https://github.com/CanyonWind/Single-Path-One-Shot-NAS-MXNet}},
      year={2019}
}