code-soup
code-soup copied to clipboard
Adversarial-Deep-Learning
This is a collection of algorithms and approaches used in the book adversarial deep learning
References: - https://github.com/QData/TextAttack/blob/9a67b9eff09a2b4382d3ecc0ddfe0afba5039162/textattack/attack_recipes/pwws_ren_2019.py - https://github.com/thunlp/OpenAttack/blob/master/OpenAttack/attackers/pwws/__init__.py Paper: https://aclanthology.org/P19-1103/ Will try to wrap https://github.com/Adversarial-Deep-Learning/code-soup/issues/66, https://github.com/Adversarial-Deep-Learning/code-soup/issues/97, https://github.com/Adversarial-Deep-Learning/code-soup/issues/98 issues in the PR which addresses this issue.
Make a wrapper class for Hugging Face models. The class will initialise the model (and the tokeniser). Perform a forward pass on an input list of text. We need functions...
Similar to https://github.com/Adversarial-Deep-Learning/code-soup/issues/97. Reference: https://github.com/QData/TextAttack/blob/master/textattack/models/wrappers/pytorch_model_wrapper.py
https://arxiv.org/pdf/1801.02610.pdf
For evasive whitebox or blackbox attacks, the objective of each attack is to fool the model to predict a different class but making it deceptive by making small changes, these...
This issue is a continuation of #19. The code for MNIST will be merged in #85. The original paper also mentions architectures for ImageNet models. It would be nice to...
The annual Hacktoberfest is underway! Register on Hacktoberfest's website and open 4 pull requests during October 2021 to plant a tree or receive the swags! What's more, you can also...
@gchhablani Is it possible to not have the config file with the algorithms for the zoo_attack? If you need it for testing can the config folder be included with just...
Metadata
Owner
Metadata
This is a collection of algorithms and approaches used in the book adversarial deep learning