Pygbm, a fast pure-Python implementation of Gradient Boosting Decision Trees
I have recently been working on Pygbm in collaboration with Olivier Grisel. Pygbm is a fast, pure-Python implementation of Gradient Boosting Decision Trees. We’re using Numba to perform Just In Time compilation.
Our implementation is heavily inspired from that of LightGBM, written in C++.
Check out Pygbm GitHub page for more info!
Nicolas Hug
ML Engineer