LLM inference in C/C++
C++ ggml
ci: add Windows ARM64 CUDA support to the manual workflow (#27300)
shivamkumard-ctrlc0296022f361
diff --git a/.github/actions/windows-setup-cuda/action.yml b/.github/actions/windows-setup-cuda/action.ymlindex 31250eda1b..917513b85e 100644--- a/.github/actions/windows-setup-cuda/action.yml+++ b/.github/actions/windows-setup-cuda/action.yml@@ -6,8 +6,7 @@ inputs:required: truecuda_arch:description: "CUDA target architecture"- required: false- default: "x64"+ required: trueruns:using: "composite"diff --git a/.github/workflows/build-cuda-windows.yml b/.github/workflows/build-cuda-windows.ymlindex 8b59f3975c..95843946f5 100644--- a/.github/workflows/build-cuda-windows.yml+++ b/.github/workflows/build-cuda-windows.yml@@ -22,6 +22,7 @@ env:jobs:cuda:+ name: windows-cuda (${{ matrix.cuda }}, ${{ matrix.arch }})runs-on: windows-2022permissions:@@ -29,7 +30,16 @@ jobs:strategy:matrix:- cuda: ['12.4', '13.3']+ include:+ - cuda: '12.4'+ arch: x64+ defines: '-DGGML_CUDA_CUB_3DOT2=ON'+ - cuda: '13.3'+ arch: x64+ defines: ''+ - cuda: '13.4'+ arch: arm64+ defines: '-DCMAKE_TOOLCHAIN_FILE=cmake/arm64-windows-msvc-cuda.cmake'steps:- name: Clone@@ -39,12 +49,13 @@ jobs:- name: ccacheuses: ggml-org/[email protected]with:- key: release-windows-2022-x64-cuda-${{ matrix.cuda }}+ key: release-windows-2022-${{ matrix.arch }}-cuda-${{ matrix.cuda }}- name: Install Cuda Toolkituses: ./.github/actions/windows-setup-cudawith:cuda_version: ${{ matrix.cuda }}+ cuda_arch: ${{ matrix.arch }}- name: Install Ninjaid: install_ninja@@ -54,26 +65,21 @@ jobs:- name: Buildid: cmake_buildshell: cmd- # TODO: Remove GGML_CUDA_CUB_3DOT2 flag once CCCL 3.2 is bundled within CTK and that CTK version is used in this projectrun: |- call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvarsall.bat" x64+ call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvarsall.bat" ${{ matrix.arch == 'x64' && 'x64' || 'amd64_arm64' }}cmake -S . -B build -G "Ninja Multi-Config" ^- -DLLAMA_BUILD_SERVER=ON ^- -DLLAMA_BUILD_BORINGSSL=ON ^- -DGGML_NATIVE=OFF ^-DGGML_BACKEND_DL=ON ^- -DGGML_CPU_ALL_VARIANTS=ON ^+ -DGGML_NATIVE=OFF ^+ -DGGML_CPU=OFF ^-DGGML_CUDA=ON ^- -DGGML_RPC=ON ^- -DGGML_CUDA_CUB_3DOT2=ON+ -DLLAMA_BUILD_BORINGSSL=ON ${{ matrix.defines }}set /A NINJA_JOBS=%NUMBER_OF_PROCESSORS%-1- cmake --build build --config Release -j %NINJA_JOBS% -t ggml- cmake --build build --config Release+ cmake --build build --config Release -j %NINJA_JOBS% --target ggml-cuda- name: ccache-clearuses: ./.github/actions/ccache-clearwith:- key: release-windows-2022-x64-cuda-${{ matrix.cuda }}+ key: release-windows-2022-${{ matrix.arch }}-cuda-${{ matrix.cuda }}hip:runs-on: windows-2022