PyTorch Adversarial Attackers
A PyTorch-based library for evaluating model robustness against adversarial attacks: FGSM, PGD, and CW attacks. Designed to work with all PyTorch models like ResNet, VGG16, MobileNet, etc.
Installation
pip install pytorch-adversarial-attackers
Usage Example
import torch
import torchvision.models as models
from adversarial_eval.attacks import FGSM, PGD, CW
from adversarial_eval.evaluate import evaluate
# Load pre-trained model
model = models.resnet18(pretrained=True).eval()
# Load image
image = torch.rand((1, 3, 224, 224)) # Example input
label = torch.tensor([0]) # Example label
# Apply attacks
fgsm_attack = FGSM(model, epsilon=0.03)
adv_image = fgsm_attack(image, label)
# Evaluate robustness
evaluate(model, image, label)
Attacks Implemented
- FGSM (Fast Gradient Sign Method)
- PGD (Projected Gradient Descent)
- CW (Carlini & Wagner Attack)
- GITHUB LINK https://github.com/santhosh1705kumar/pytorch-adversarial-attackers
1 maintainer
Metadata
1 maintainer
Release files for pytorch-adversarial-attackers 0.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| pytorch_adversarial_attackers-0.1.tar.gz | 1.8 kB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| pytorch_adversarial_attackers-0.1-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 3.7 kB
Release files / pytorch_adversarial_attackers-0.1.tar.gz
| Download URL | pytorch_adversarial_attackers-0.1.tar.gz |
|---|---|
| Size | 1.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
5a7d4930d8dcb51945467b29b47314b5de7b9f6faf4f0c40e9ba30d338b82d02
|
|
BLAKE2b-256 checksum How to use checksums |
4b24e881878d2916f372cd61f7d3e57e00e4f61935c33a1bfa10060df29befc2
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.12.4
|
Release files / pytorch_adversarial_attackers-0.1-py3-none-any.whl
| Download URL | pytorch_adversarial_attackers-0.1-py3-none-any.whl |
|---|---|
| Size | 1.9 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
f5b9bf6c104e4ae190ee9d0a434b3d8daef22b58de0ed43dec4dc1ece82940f8
|
|
BLAKE2b-256 checksum How to use checksums |
53fdda14ec5fd8bedbd79f12478331e023bc6b231c535077d5bba798a932a80e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.1.0 CPython/3.12.4
|
Security for pytorch-adversarial-attackers 0.1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.
