News for hackers. Stuff that matters.
Re:LightGBM – A fast, distributed, gradient boosting framework by TheGuyWhoCodes on Sunday October 16, 2016 @10:28PM (#12722343)
Looks fantastic!

I'd love to have a python interface for this, just drop a pandas frame, maybe scikit-learn interface with fit/predict. Saving/Loading models... This will definitely boost adoption.

Re:LightGBM – A fast, distributed, gradient boosting framework by minimaxir on Sunday October 16, 2016 @11:46PM (#12722568)
That is explicitly in the future plans: https://github.com/Microsoft/LightGBM/wiki/Features