关于GCU、沐曦GPGPU、MLU、0卡V100资源4月7日恢复上架的公告>>> 关于共建具身智能开源数据集的倡议>>> 关于云脑任务中统一路径访问方式的公告>>> 关于将启智集群GPU资源迁移至智算集群的公告>>>
sy_liang
/
CAS_FGAC_CL
forked from zhounuoyan/CAS_CL
Not watched
1
0
Fork 3
Code
Releases 0 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain
You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
260 Commits
4 Branches
2.2 MiB
6 Download
 
Branch: master
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'master'
${ noResults }
sy_liang 20cd7aff08 更新 'MART/models/preactresnet5.py' 1 year ago
ActivationAnalysis 上传文件至 'ActivationAnalysis/models' 1 year ago
MART 更新 'MART/models/preactresnet5.py' 1 year ago
Standard_Adv_Training 上传文件至 'Standard_Adv_Training/utils' 1 year ago
Trades 更新 'Trades/train_trades_NPsfc_two_layer.py' 1 year ago
README.md 上传文件至 '' 1 year ago
fd,.log 更新 'fd,.log' 1 year ago

README.md

Improving Adversarial Robustness via Channel-wise Activation Suppressing (ICLR2021)

This is the official Pytorch implementation of CAS.

Training

  1. Train AT+CAS

    python ./Standard_Adv_Training/train.py --gpu 0 --adv_train --affix AT_CAS

  2. Train Trades+CAS

    python ./Trades/train_trades_cifar10_cas.py

  3. Train MART+CAS

    python ./MART/train_mart_cas.py

Test

Specify the checkpoint path as argument

python ./Trades/pgd_attack_cifar10_cas.py --model_path xxx
Delete task

Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.

cancel
confirm

CAS_CL

Python

Contributors (2)
All

Please read the following content carefully:

Dear OpenI User

Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.

For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》

Agree and continue
Disagree, exit