Metadata-Version: 2.1
Name: exlib
Version: 0.2.0
Summary: Toolkit for explainability by BrachioLab
Author-email: Eric Wong <exwong@cis.upenn.edu>, Helen Jin <helenjin@seas.upenn.edu>
Project-URL: Homepage, https://github.com/BrachioLab/exlib
Project-URL: Bug Tracker, https://github.com/BrachioLab/exlib/issues
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.11
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: lime >=0.2
Requires-Dist: shap >=0.42
Requires-Dist: torch >=2.0
Requires-Dist: torchvision >=0.15
Requires-Dist: grad-cam >=1.4.8
Requires-Dist: jsonlines >=3.1.0
Requires-Dist: scikit-image >=0.22
Requires-Dist: datasets >=2.19
Requires-Dist: transformers >=4.38
Requires-Dist: sentence-transformers >=3.0.0
Requires-Dist: sentencepiece >=0.2.0
Requires-Dist: torchxrayvision >=1.2.3
Requires-Dist: bertopic >=0.16.4
Requires-Dist: segment-anything >=1.0
Requires-Dist: timm >=1.0.10
Requires-Dist: notebook
Requires-Dist: ipywidgets
Provides-Extra: dev
Requires-Dist: notebook ; extra == 'dev'

# exlib
[![PyPI](https://img.shields.io/pypi/v/exlib)](https://pypi.org/project/exlib/)
[![License](https://img.shields.io/badge/license-MIT-blue.svg)](https://github.com/BrachioLab/exlib/blob/master/LICENSE)

`exlib` is a comprehensive package showcasing our lab's work on explanation methods, featuring user-friendly modules for easy application of various techniques. 

## Installation
```
pip install exlib
```

<!--
If you have exlib already installed, please check that you have the latest version:
```
python -c "import exlib; print(exlib.__version__)"
# This should print "0.1.0". If it does not, update the package by running:
pip install -U exlib
```
-->

To use `pytorch-gradcam`, install our customized and expanded version at
```
pip install grad-cam@git+https://github.com/brachiolab/pytorch-grad-cam
```

## Projects
We list below some relevant projects that use exlib heavily.

### The FIX Benchmark: Extracting Features Interpretable to eXperts
* Documentation available [here](https://github.com/BrachioLab/exlib/tree/main/fix).
* Quick-start tutorial notebook at [`fix_demo.py`](https://github.com/BrachioLab/exlib/blob/dev/fix_demo.ipynb)
* [<a href="https://arxiv.org/abs/2409.13684">Paper</a>] [<a href="https://brachiolab.github.io/fix/">Website</a>] [<a href="https://debugml.github.io/fix/">Blog Post</a>] 


