This is a Python implementation of the Dirichlet Calibration presented in
Beyond temperature scaling: Obtaining well-calibrated multi-class probabilities
with Dirichlet calibration at NeurIPS 2019. The original version used Python
3.8 and reached version 0.4.2. The code started using Python 3.12 from
version 0.5.0, you can see the other version in the GitHub history, tags, or
in Pypi.
# Clone the repository
git clone git@github.com:dirichletcal/dirichlet_python.git
# Go into the folder
cd dirichlet_python
# Create a new virtual environment with Python3
python3.12 -m venv venv
# Load the generated virtual environment
source venv/bin/activate
# Upgrade pip
pip install --upgrade pip
# Install all the dependencies
pip install -r requirements.txt
pip install --upgrade jaxlib
python -m unittest discover dirichletcal
If you use this code in a publication please cite the following paper
@inproceedings{kull2019dircal,
title={Beyond temperature scaling: Obtaining well-calibrated multi-class probabilities with Dirichlet calibration},
author={Kull, Meelis and Nieto, Miquel Perello and K{\"a}ngsepp, Markus and Silva Filho, Telmo and Song, Hao and Flach, Peter},
booktitle={Advances in Neural Information Processing Systems},
pages={12295--12305},
year={2019}
}
You can find some examples on how to use this package in the folder examples
To push a new version to Pypi first build the package
python3.12 setup.py sdist
And then upload to Pypi with twine
twine upload dist/*
It may require user and password if these are not set in your home directory a file .pypirc
[pypi]
username = __token__
password = pypi-yourtoken
