setup-client

GitHub Action step for adding GPUs to your run via LUPINE.

This lets you run workloads that need a GPU directly on GitHub Actions or any other CPU-only actions runner.

By default, this action uses https://lupine.sh/ hosted free GPUs, however you can point it at your self-hosted GPU with the server argument.

Only Linux runners are supported.

Usage

jobs:
  test:
    runs-on: ubuntu-24.04
    steps:
      - uses: actions/checkout@v6

      - uses: lupinemachines/setup-client@v1
        with:
          cuda-version: 13.1.0
          ubuntu-version: "24.04"

      - run: |
          echo "$LUPINE_LIBCUDA"
          ls -l "$LUPINE_LIB_DIR"

By default, the action uses the latest public LUPINE release and installs the CUDA 13.1.0 / Ubuntu 24.04 / x86_64 client asset.

This action installs the LUPINE libcuda and libnvidia-ml shim libraries. It does not install CUDA runtime libraries, CUDA development tools, or nvidia-smi; install those separately if your job needs them, we recommend this action.

Inputs

Input	Default	Description
version	latest	LUPINE release tag to download, or latest.
repository	lupinemachines/lupine	Repository that publishes the release assets.
cuda-version	13.1.0	CUDA version in the client asset name.
ubuntu-version	24.04	Ubuntu version in the client asset name. ubuntu24.04 is also accepted.
architecture	auto	Release asset architecture. Currently resolves to x86_64.
install-dir	runner tool cache	Directory where the client files should be installed.
github-token	empty	Optional token for private repositories or higher rate limits.
server	empty	Optional LUPINE_SERVER value exported for later steps.
export-env	true	Export LUPINE_* and LD_LIBRARY_PATH variables.

Outputs

Output	Description
install-dir	Directory containing the installed release files.
lib-dir	Directory containing the shim libraries.
libcuda	Path to libcuda.so.1.
libnvidia-ml	Path to libnvidia-ml.so.1.
asset-name	Release asset downloaded by the action.
download-url	URL used to download the release asset.