WindVChen

That's interesting. From your testing, S2I-FGSM is approximately 4 points less transferable. If there are no fundamental errors in running their source code, this discrepancy might be attributed to differences...

Ah, you may need to include the normalization with L150.

You may have a try with this (incorporate the normalization operation into the line 150): ```python model = torch.nn.Sequential(Normalize(opt.mean, opt.std), models.inception_v3(pretrained=True).eval().cuda()) ```

Hi, For HGD, you should run `defense.py`. You can find more information [here](https://github.com/lfz/Guided-Denoise/tree/master/nips_deploy). Regarding DiffPure, you can refer to DiffPure's source code for details. In the adversarial purification evaluation, they...

Hi @Rajesh1215 , Please recheck your version of **diffusers** package. Version 0.9.0 should work well. Hope this helps!

Hi @Rajesh1215 , Apology for the delayed response. For the error above, the answer [here](https://stackoverflow.com/questions/78302031/stable-diffusion-attributeerror-module-jax-random-has-no-attribute-keyarray) may help. Hope it helps.

Hello @Donus-S, Thank you for your recognition of our work. As the code was written quite some time ago, I tried to recall the exact reason behind that comment but...

Hi @wxzzu , It seems there might be a conflict between the `argparse.ArgumentParser()` in [main.py](https://github.com/WindVChen/DiffAttack/blob/163f0f7f2bbfd7e57ed99c0baf6523d95167a3e4/main.py) and [fid_score.py](https://github.com/WindVChen/DiffAttack/blob/163f0f7f2bbfd7e57ed99c0baf6523d95167a3e4/pytorch_fid/fid_score.py). Could you please share the input command you're using? This will help me...

Hi, it should be working now. You can update your code to the latest version of the repo, or check out the modifications [here](https://github.com/WindVChen/DiffAttack/commit/d66b1e79f1cd706b8deeedbe83e07925b7216e53).

Hi @HONG-0122 , Could you share more details about the issue? For instance, did you modify any code before encountering this error? The original code in the repository should be...