QData/TextAttack
TextAttack 🐙 is a Python framework for adversarial attacks, data augmentation, and model training in NLP https://textattack.readthedocs.io/en/master/
This helps NLP researchers and practitioners understand and improve the robustness of their language models. It takes an existing NLP model (like for sentiment analysis or paraphrase detection) and generates slightly altered text inputs that can trick the model, revealing its weaknesses. The output helps users either uncover vulnerabilities in their models or generate new, diverse training data to make models more resilient.
3,377 stars. No commits in the last 6 months. Available on PyPI.
Use this if you need to test the limits of your NLP model's understanding, explore its failure modes, or create a more robust model through data augmentation.
Not ideal if you are looking for a general-purpose NLP library for common tasks like basic text classification or entity recognition without a focus on adversarial testing or robustness.
Stars
3,377
Forks
439
Language
Python
License
MIT
Category
Last pushed
Jul 10, 2025
Commits (30d)
0
Dependencies
22
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/QData/TextAttack"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
ebagdasa/backdoors101
Backdoors Framework for Deep Learning and Federated Learning. A light-weight tool to conduct...
THUYimingLi/backdoor-learning-resources
A list of backdoor learning resources
zhangzp9970/MIA
Unofficial pytorch implementation of paper: Model Inversion Attacks that Exploit Confidence...
LukasStruppek/Plug-and-Play-Attacks
[ICML 2022 / ICLR 2024] Source code for our papers "Plug & Play Attacks: Towards Robust and...
VinAIResearch/Warping-based_Backdoor_Attack-release
WaNet - Imperceptible Warping-based Backdoor Attack (ICLR 2021)