Anonymous_Submission

To train a backdoor model with "blend" attack with poison ratio of "10%"-

  python train_backdoor_cifar.py --poison-type blend --posion-rate 0.10 --output-dir /folder/to/save --gpuid 0 

To train a benign model-

python train_backdoor_cifar.py --poison-type benign --output-dir /folder/to/save --gpuid 0 

To remove backdoor with 1% clean validation data-

python Remove_Backdoor_SFT.py --poison-type blend --val-frac 0.01 --checkpoint path/to/backdoor/model --gpuid 0