BootstrapNAS (1) takes a pre-trained model as input. (2) It uses this model to generate a weight-sharing super-network. (3) BootstrapNAS then applies a training strategy, and once the super-network has been trained, (4) it searches for efficient subnetworks that satisfy the user's requirements. (5) The configuration of the discovered sub-network(s) is returned to the user.
Please follow the instructions here.
If you already have a super-network trained with BootstrapNAS, please follow the instructions to search for sub-networks here.
More information about BootstrapNAS is available in our papers:
Automated Super-Network Generation for Scalable Neural Architecture Search.
@inproceedings{
munoz2022automated,
title={Automated Super-Network Generation for Scalable Neural Architecture Search},
author={Muñoz, J. Pablo and Lyalyushkin, Nikolay and Lacewell, Chaunte and Senina, Anastasia and Cummings, Daniel and Sarah, Anthony and Kozlov, Alexander and Jain, Nilesh},
booktitle={First Conference on Automated Machine Learning (Main Track)},
year={2022},
url={https://openreview.net/pdf?id=HK-zmbTB8gq}
}Enabling NAS with Automated Super-Network Generation
@article{
bootstrapNAS,
author = {Muñoz, J. Pablo and Lyalyushkin, Nikolay and Akhauri, Yash and Senina, Anastasia and Kozlov, Alexander and Jain, Nilesh},
title = {Enabling NAS with Automated Super-Network Generation},
journal = {1st International Workshop on Practical
Deep Learning in the Wild at AAAI},
year = {2022},
url = {https://practical-dl.github.io/2022/short_paper/21.pdf},
}Please follow the contribution guidelines in NNCF.