Skip to content

Latest commit

 

History

History
19 lines (15 loc) · 571 Bytes

README.md

File metadata and controls

19 lines (15 loc) · 571 Bytes

arboretum - Gradient Boosting on GPU

Dependencies

  • Python 2.7
  • Cuda toolkit 8(there's no cuda 8 sprecifiec features are used so it should also work with cuda 7, but it has never been tested).
  • Cuda thrust - it's a part of cuda toolkit so it should work out of box.
  • Cuda cub https://github.com/NVlabs/cub as submodule.

Linux

  • gcc-5

MacOS

  • clang-omp++ can be install by brew

Installation process is similar to xgboost