329-neural-architecture-search-for-spiking-neural-networks
https://github.com/szu-advtech-2023/329-neural-architecture-search-for-spiking-neural-networks
Science Score: 28.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
○codemeta.json file
-
○.zenodo.json file
-
○DOI references
-
✓Academic publication links
Links to: arxiv.org -
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (13.7%) to scientific vocabulary
Last synced: 9 months ago
·
JSON representation
·
Repository
Basic Info
- Host: GitHub
- Owner: SZU-AdvTech-2023
- Language: Python
- Default Branch: main
- Size: 1.09 MB
Statistics
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
- Releases: 0
Created over 2 years ago
· Last pushed over 2 years ago
Metadata Files
Citation
https://github.com/SZU-AdvTech-2023/329-Neural-Architecture-Search-for-Spiking-Neural-Networks/blob/main/
# Neural Architecture Search for Spiking Neural Networks ## Prerequisites * Python 3.9 * PyTorch 1.10.0 * NVIDIA GPU (>= 12GB) * CUDA 10.2 (optional) ## Getting Started ### Conda Environment Setting ``` conda create -n SNASNet conda activate SNASNet conda install pytorch torchvision torchaudio cudatoolkit=10.2 -c pytorch pip install scipy ``` ### Spikingjelly Installation (ref: https://github.com/fangwei123456/spikingjelly) ``` git clone https://github.com/fangwei123456/spikingjelly.git cd spikingjelly python setup.py install ``` ## Training and testing ### Training * Run the following command ``` python search_snn.py --exp_name 'cifar100_backward' --dataset 'cifar100' --celltype 'backward' --batch_size 32 --num_search 5000 ``` simple argument instruction --exp_name: savefile name --dataset: dataset for experiment --celltype: find backward connections or forward connections --num_search: number of architecture candidates for searching ## Testing with pretrained models (CIFAR10 & CIFAR100) (1) Run the following command ``` python search_snn.py --dataset 'cifar10' --cnt_mat 0303 0030 2002 0200 --savemodel_pth './savemodel/save_cifar10_bw.pth.tar' --celltype 'backward' --second_avgpooling 4 ``` ``` python search_snn.py --dataset 'cifar100' --cnt_mat 0302 0030 3003 0000 --savemodel_pth './savemodel/save_cifar100_bw.pth.tar' --celltype 'backward' ``` --cnt_mat: searched network architecture --savemodel_pth: network parameters saved during training
Owner
- Name: SZU-AdvTech-2023
- Login: SZU-AdvTech-2023
- Kind: organization
- Repositories: 1
- Profile: https://github.com/SZU-AdvTech-2023
Citation (citation.txt)
@inproceedings{REPO329,
author = "Kim, Youngeun and Li, Yuhang and Park, Hyoungseob and Venkatesha, Yeshwanth and Panda, Priyadarshini",
booktitle = "European Conference on Computer Vision",
organization = "Springer",
pages = "36--56",
title = "{Neural Architecture Search for Spiking Neural Networks}",
year = "2022"
}