Fair Language Processing (FairLangProc)

The Fair Language Processing package is a extensible open-source Python library containing techniques developed by the research community to help detect and mitigate bias in Natural Language Processing throughout the AI application lifecycle.

The FairLangProc package includes:

1. Data sets to test for biases in NLP models.
2. Metrics based on different philosophies to quantified said biases.
3. Algorithms to mitigate biases.

It has been created with the intention of encouraging the use of bias mitigation strategies in the NLP community, and with the hope of democratizing these tools for the ever-increasing set of NLP practitioners. We invite you to use it and improve it.

• Companion paper: https://arxiv.org/abs/2508.03677.
• Source code: https://github.com/arturo-perez-peralta/FairLangProc/tree/main/FairLangProc.
• Notebooks with examples: https://github.com/arturo-perez-peralta/FairLangProc/tree/main/notebooks.
• Documentation: https://fairlangproc.readthedocs.io/en/latest/.

We have developed the package with extensibility in mind. This library is still in development. We encourage your contributions.

Supported fairness datasets

Supported fairness metrics

• Generalized association tests (WEAT) (Caliskan et al., 2016)
• Log Probability Bias Score (LPBS) (Kurita et al., 2019)
• Categorical Bias Score (CBS) (Ahn et al., 2021)
• CrowS-Pairs Score (CPS) (Nangia et al., 2020)
• All Unmasked Score (AUL) (Kaneko et al., 2021)
• Demographic Representation (DR) (Liang et al., 2022)
• Stereotypical Association (SA) (Liang et al., 2022)
• HONEST (Nozza et al., 2021)

Supported bias mitigation algorithms

• Counterfactual Data Augmentation (CDA) (Webster et al. 2020)
• Projection based debiasing (Bolukbasi et al., 2023)
• Bias removaL wIth No Demographics (BLIND) (Orgad et al., 2023)
• Adapter-based DEbiasing of LanguagE models (Lauscher et al., 2021)
• Modular Debiasing with Diff Subnetworks (Hauzenberger et al., 2023)
• Entropy Attention Temperature (EAT) scaling (Zayed et al., 2023)
• Entropy Attention Regularizer (EAR) (Attanasio et al., 2022)
• Embedding based regularizer (Liu et al., 2020)
• Selective unfreezing (Gira et al., 2024)

Setup

Python

To install the latest stable version from PyPI, run:

pip install FairLangProc

Has been tested and ran with both Python 3.13 and Python 3.10. Compatibility with older versions is possible and expected, although we are still testing older configurations. The minimum tested versions of the requirements are:

• pandas>=2.2.3
• scikit-learn>=1.6.1
• torch>=2.6.0
• transformers>=4.47.1
• datasets>=3.4.1
• adapter-transformers>=1.1.0
• accelerate>=0.26.0
• pytest>=8.4.1

Manual installation

Clone the latest version of this repository:

git clone https://github.com/arturo-perez-peralta/FairLangProc

Using FairLangProc

The notebooks directory contains a diverse collection of jupyter notebooks that showcase how to use the different processors, metrics and data sets. If you'd like to run the examples requiring , download the data sets now and place them in a folder named Fair-LLM-Benchmarks inside the 'FairLangProc/datsets' path or simply clone the repository from Gallegos et al.

Using the BiasDataLoader

In order to run the BiasDataLoader first you need to download the datasets from the repository from I. Gallegos https://github.com/i-gallegos/Fair-LLM-Benchmark. In order to do this you first need to find the path of the package. This can be done with the following Python command:

python -c "import FairLangProc; print(FairLangProc.__file__)" 

Now you only need to download the datasets:

git clone https://github.com/i-gallegos/Fair-LLM-Benchmark [absolute path to your Python packages folder]/FairLangProc/datasets

Credits

For attribution in academic contexts, please use the bibtex entry below:

@misc{pérezperalta2025fairlangprocpythonpackagefairness,
      title={FairLangProc: A Python package for fairness in NLP}, 
      author={Arturo Pérez-Peralta and Sandra Benítez-Peña and Rosa E. Lillo},
      year={2025},
      eprint={2508.03677},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2508.03677}, 
}

We thank Víctor Agulló for his inputs on many different questions that arose during the making of the package as well as for his contributions on different parts of the code, specially those related to the BiasDataLoader method.