search

MadryLab / blackbox-bandits

Code for "Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors"
https://arxiv.org/abs/1807.07978
MIT License
61 stars 14 forks source link

Code running slow #9

Open maithal opened 3 years ago

maithal commented 3 years ago

Hi,

I am trying to run the attack method for ImageNET data and for some reason the code is running extremely slow. It is printing tensor(1.)tensor(0.) for quite a long time. Is it expected to take longer time when you run for the first time?

Any help will be highly appreciated!

momo1986 commented 11 months ago

Same problem when I tried to attack on InceptionV3 model.

python -u src/main.py --json-config src/configs/linf.json | tee imagenet_incepetionv2_bandit_linf_eval.log Downloading: "https://download.pytorch.org/models/inception_v3_google-1a9a5a14.pth" to /root/.cache/torch/hub/checkpoints/inception_v3_google-1a9a5a14.pth 100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 104M/104M [00:05<00:00, 18.3MB/s] tensor(1.) tensor(0.) tensor(1.) tensor(0.) tensor(1.) tensor(0.)