The repository contains a list of resources to use a homomorphically encrypted inference from any logistic regression whose weights size is smaller than 214 floating point elements.
From the wheelhouse directory, take the version according to your version of Python. Currently it only works in Linux environments.
git clone <repo_address>
cd wheelhouse
pip3 install <installer>.whl
from HELR import HELR
...
my_private_inferece = HELR(weights, bias)
result = my_private_inference.predict(sample)
...