Metadata-Version: 2.2
Name: textual-adversarial-defense
Version: 0.1.0
Summary: Unicode-based adversarial defenses for text models
Author-Email: Carol Khor <shunting.khor1@gmail.com>
License: MIT
Project-URL: Homepage, https://github.com/khorcarol/textual-adversarial-defense
Project-URL: Repository, https://github.com/khorcarol/textual-adversarial-defense.git
Requires-Python: >=3.9
Requires-Dist: numpy>=1.19.0
Provides-Extra: dev
Requires-Dist: pytest>=7.0; extra == "dev"
Description-Content-Type: text/markdown


### Textual Adversarial Defenses
Building defenses for imperceptible textual attacks against NLP models. Work is based off https://arxiv.org/pdf/2106.09898.

#### Project Folders
- `core/` top level class definitions
- `modules/` source files for defenses
- `python/` Python bindings
- `tests/` Unit tests

