fix #19
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: lightgbm | |
| # https://lightgbm.readthedocs.io/en/latest/Installation-Guide.html#id18 | |
| # https://github.com/microsoft/LightGBM/tree/master/python-package#build-gpu-version | |
| on: | |
| push: | |
| branches: | |
| - lightgbm | |
| workflow_dispatch: | |
| jobs: | |
| build: | |
| runs-on: windows-2022 | |
| steps: | |
| - uses: actions/setup-python@v5 | |
| with: | |
| python-version: '3.10' | |
| - name: boost | |
| uses: MarkusJx/install-boost@main | |
| id: install-boost | |
| with: | |
| platform_version: 2022 | |
| boost_version: 1.86.0 | |
| toolset: msvc | |
| cache: false | |
| - name: cuda-toolkit | |
| uses: Jimver/cuda-toolkit@master | |
| id: cuda-toolkit | |
| # https://docs.nvidia.com/cuda/archive/12.6.2/cuda-installation-guide-microsoft-windows/index.html#id3 | |
| with: | |
| # cuda: 12.1.1 | |
| method: local | |
| # sub-packages: '["nvcc", "visual_studio_integration"]' | |
| sub-packages: '["nvcc", "cudart"]' | |
| # Nice spot. I did not realize msvc worked without the VS integration of CUDA. | |
| # https://github.com/bitsandbytes-foundation/bitsandbytes/pull/1055 | |
| # https://github.com/Jimver/cuda-toolkit/issues/253 | |
| # Disable caching of the CUDA binaries, since it does not give us any significant performance improvement | |
| use-github-cache: false | |
| use-local-cache: false | |
| # 在GITHUB_ENV中输出CUDA_PATH | |
| - name: echo version | |
| run: | | |
| nvcc -V | |
| echo ${{ steps.cuda-toolkit.outputs.cuda }} | |
| echo "${{ env.CUDA_PATH }}" | |
| echo ${{ steps.install-boost.outputs.BOOST_VER }} | |
| - name: Configure build | |
| run: | | |
| # cmake -B build -S . -A x64 -DUSE_GPU=1 -DBOOST_ROOT=$BOOST_ROOT | |
| # cmake -B build -S . -A x64 -DUSE_GPU=1 -DBOOST_ROOT=BOOST_ROOT -DOpenCL_LIBRARY="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v10.0/lib/x64/OpenCL.lib" -DOpenCL_INCLUDE_DIR="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v10.0/include" | |
| python -V | |
| # pip wheel lightgbm --config-settings=cmake.define.Boost_ROOT=$BOOST_ROOT | |
| # pip wheel lightgbm --config-settings=cmake.define.USE_GPU=ON --config-settings=cmake.define.OpenCL_INCLUDE_DIR="$CUDA_PATH/include"--config-settings=cmake.define.Boost_ROOT=$BOOST_ROOT | |
| git clone --recursive https://github.com/Microsoft/LightGBM | |
| cd LightGBM | |
| sh ./build-python.sh bdist_wheel --gpu --boost-include-dir=$BOOST_ROOT ----opencl-include-dir="$CUDA_PATH/include" | |
| env: | |
| BOOST_ROOT: ${{ steps.install-boost.outputs.BOOST_ROOT }} | |
| - run: tree LightGBM/dist | |
| # - name: Upload all wheel files | |
| # uses: actions/upload-artifact@v4 | |
| # with: | |
| # name: lightgbm-wheels | |
| # path: ' d:\a\win-whl\win-whl\*lightgbm*.whl' | |
| # if-no-files-found: error |