diff --git a/.github/workflows/pr-gpu.yaml b/.github/workflows/pr-gpu.yaml index 07e811e244..1d8323ba9f 100644 --- a/.github/workflows/pr-gpu.yaml +++ b/.github/workflows/pr-gpu.yaml @@ -13,18 +13,18 @@ concurrency: group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }} cancel-in-progress: ${{ github.ref != 'refs/heads/main' }} jobs: - pytest-gpu: - uses: mosaicml/ci-testing/.github/workflows/pytest-gpu.yaml@v0.0.5 + pytest-gpu-1: + uses: mosaicml/ci-testing/.github/workflows/pytest-gpu.yaml@v0.0.8 strategy: fail-fast: false matrix: include: - - name: "gpu-2.3.0" + - name: "gpu-2.3.0-1" container: mosaicml/pytorch:2.3.0_cu121-python3.11-ubuntu20.04 markers: "gpu" pytest_command: "coverage run -m pytest" pip_deps: "[all]" - - name: "gpu-2.3.1" + - name: "gpu-2.3.1-1" container: mosaicml/llm-foundry:2.3.1_cu121-latest markers: "gpu" pytest_command: "coverage run -m pytest" @@ -40,5 +40,66 @@ jobs: pytest-command: ${{ matrix.pytest_command }} pytest-markers: ${{ matrix.markers }} python-version: 3.9 + gpu_num: 1 + secrets: + mcloud-api-key: ${{ secrets.MCLOUD_API_KEY }} + pytest-gpu-2: + uses: mosaicml/ci-testing/.github/workflows/pytest-gpu.yaml@v0.0.8 + strategy: + fail-fast: false + matrix: + include: + - name: "gpu-2.3.0-2" + container: mosaicml/pytorch:2.3.0_cu121-python3.11-ubuntu20.04 + markers: "gpu" + pytest_command: "coverage run -m pytest" + pip_deps: "[all]" + - name: "gpu-2.3.1-2" + container: mosaicml/llm-foundry:2.3.1_cu121-latest + markers: "gpu" + pytest_command: "coverage run -m pytest" + pip_deps: "[all]" + name: ${{ matrix.name }} + if: github.repository_owner == 'mosaicml' + with: + container: ${{ matrix.container }} + git_repo: mosaicml/llm-foundry + mcloud-timeout: 1800 + name: ${{ matrix.name }} + pip_deps: ${{ matrix.pip_deps }} + pytest-command: ${{ matrix.pytest_command }} + pytest-markers: ${{ matrix.markers }} + python-version: 3.9 + gpu_num: 2 + secrets: + mcloud-api-key: ${{ secrets.MCLOUD_API_KEY }} + pytest-gpu-4: + uses: mosaicml/ci-testing/.github/workflows/pytest-gpu.yaml@v0.0.8 + strategy: + fail-fast: false + matrix: + include: + - name: "gpu-2.3.0-4" + container: mosaicml/pytorch:2.3.0_cu121-python3.11-ubuntu20.04 + markers: "gpu" + pytest_command: "coverage run -m pytest" + pip_deps: "[all]" + - name: "gpu-2.3.1-4" + container: mosaicml/llm-foundry:2.3.1_cu121-latest + markers: "gpu" + pytest_command: "coverage run -m pytest" + pip_deps: "[all]" + name: ${{ matrix.name }} + if: github.repository_owner == 'mosaicml' + with: + container: ${{ matrix.container }} + git_repo: mosaicml/llm-foundry + mcloud-timeout: 1800 + name: ${{ matrix.name }} + pip_deps: ${{ matrix.pip_deps }} + pytest-command: ${{ matrix.pytest_command }} + pytest-markers: ${{ matrix.markers }} + python-version: 3.9 + gpu_num: 4 secrets: mcloud-api-key: ${{ secrets.MCLOUD_API_KEY }}