fix: Check device type in _initialize_llm before moving to cuda. (#…
#581
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
| # This workflow will install Python dependencies and run all tests marked as `slow` on a single Python version. | |
| # The tests will run on a high-memory AWS compute instance to accommodate memory-intensive workloads. | |
| name: pytest (slow) | |
| on: | |
| push: | |
| branches: ["master", "release-*"] | |
| tags: ["v*.*.*"] | |
| jobs: | |
| start-runner: | |
| name: Start self-hosted EC2 runner | |
| if: > | |
| github.event_name == 'schedule' && github.repository == 'ludwig-ai/ludwig' || | |
| github.event_name == 'push' && github.repository == 'ludwig-ai/ludwig' || | |
| github.event_name == 'pull_request' && github.event.pull_request.base.repo.full_name == 'ludwig-ai/ludwig' && !github.event.pull_request.head.repo.fork | |
| runs-on: ubuntu-latest | |
| outputs: | |
| label: ${{ steps.start-ec2-runner.outputs.label }} | |
| ec2-instance-id: ${{ steps.start-ec2-runner.outputs.ec2-instance-id }} | |
| steps: | |
| - name: Configure AWS credentials | |
| uses: aws-actions/configure-aws-credentials@v1 | |
| with: | |
| aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }} | |
| aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }} | |
| aws-region: ${{ secrets.AWS_REGION }} | |
| - name: Start EC2 runner | |
| id: start-ec2-runner | |
| uses: machulav/[email protected] | |
| with: | |
| mode: start | |
| github-token: ${{ secrets.GH_PERSONAL_ACCESS_TOKEN }} | |
| ec2-image-id: ami-0759580dedc953d1f | |
| ec2-instance-type: r5.large | |
| subnet-id: subnet-0983be43 | |
| security-group-id: sg-4cba0d08 | |
| aws-resource-tags: > | |
| [ | |
| {"Key": "Name", "Value": "ludwig-github-${{ github.head_ref || github.sha }}"}, | |
| {"Key": "GitHubRepository", "Value": "${{ github.repository }}"}, | |
| {"Key": "GitHubHeadRef", "Value": "${{ github.head_ref }}"}, | |
| {"Key": "GitHubSHA", "Value": "${{ github.sha }}"} | |
| ] | |
| slow-pytest: | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| os: [ubuntu-latest] | |
| python-version: [3.8] | |
| test-markers: ["slow"] | |
| include: | |
| - python-version: 3.8 | |
| pytorch-version: 2.0.0 | |
| torchscript-version: 1.10.2 | |
| ray-version: 2.3.1 | |
| env: | |
| PYTORCH: ${{ matrix.pytorch-version }} | |
| MARKERS: ${{ matrix.test-markers }} | |
| NEUROPOD_BASE_DIR: "/usr/local/lib/neuropod" | |
| NEUROPOD_VERSION: "0.3.0-rc6" | |
| TORCHSCRIPT_VERSION: ${{ matrix.torchscript-version }} | |
| RAY_VERSION: ${{ matrix.ray-version }} | |
| AWS_ACCESS_KEY_ID: ${{ secrets.LUDWIG_TESTS_AWS_ACCESS_KEY_ID }} | |
| AWS_SECRET_ACCESS_KEY: ${{ secrets.LUDWIG_TESTS_AWS_SECRET_ACCESS_KEY }} | |
| EXCLUDED_MARKERS: "benchmark" | |
| TOKENIZERS_PARALLELISM: false | |
| name: py${{ matrix.python-version }}, torch-${{ matrix.pytorch-version }}, ${{ matrix.test-markers }}, ${{ matrix.os }}, ray ${{ matrix.ray-version }} | |
| if: needs.start-runner.result != 'skipped' | |
| needs: start-runner # required to start the main job when the runner is ready | |
| runs-on: ${{ needs.start-runner.outputs.label }} # run the job on the newly created runners | |
| timeout-minutes: 60 | |
| steps: | |
| - uses: actions/checkout@v2 | |
| - name: Set up Python ${{ matrix.python-version }} | |
| uses: actions/setup-python@v2 | |
| with: | |
| python-version: ${{ matrix.python-version }} | |
| - name: Setup Linux | |
| if: runner.os == 'linux' | |
| run: | | |
| sudo apt-get update && sudo apt-get install -y libsndfile1 cmake ccache build-essential g++-8 gcc-8 | |
| cmake --version | |
| - name: Setup macOS | |
| if: runner.os == 'macOS' | |
| run: | | |
| brew install libuv | |
| - name: pip cache | |
| if: ${{ !env.ACT }} | |
| uses: actions/cache@v2 | |
| with: | |
| path: ~/.cache/pip | |
| key: ${{ runner.os }}-pip-py${{ matrix.python-version }}-torch${{ matrix.pytorch-version }}-${{ matrix.test-markers }}-${{ hashFiles('requirements*.txt') }} | |
| restore-keys: | | |
| ${{ runner.os }}-pip-py${{ matrix.python-version }}-torch${{ matrix.pytorch-version }}-${{ matrix.test-markers }}- | |
| - name: Install dependencies | |
| run: | | |
| python --version | |
| pip --version | |
| python -m pip install -U pip | |
| cmake --version | |
| echo "MARKERS:" $MARKERS | |
| if [ "$PYTORCH" == "nightly" ]; then | |
| cat requirements.txt | sed '/^torch[>=<]/d' | sed '/^torchtext[>=<]/d' | sed '/^torchvision[>=<]/d' | sed '/^torchaudio[>=<]/d' > requirements-temp && mv requirements-temp requirements.txt | |
| extra_index_url=https://download.pytorch.org/whl/nightly/cpu | |
| pip install --pre torch torchtext torchvision torchaudio --extra-index-url $extra_index_url | |
| else | |
| extra_index_url=https://download.pytorch.org/whl/cpu | |
| pip install torch==$PYTORCH torchtext torchvision torchaudio --extra-index-url $extra_index_url | |
| fi | |
| if [ "$RAY_VERSION" == "nightly" ]; then | |
| # NOTE: hardcoded for python 3.9 on Linux | |
| pip install https://s3-us-west-2.amazonaws.com/ray-wheels/latest/ray-3.0.0.dev0-cp39-cp39-manylinux2014_x86_64.whl | |
| else | |
| # installing `six` early resolves ModuleNotFound error in ray==2.1.0 | |
| pip install six | |
| pip install ray==$RAY_VERSION | |
| fi | |
| ray_expected=$(python -c "import ray; print(ray.__version__)") | |
| torch_expected=$(python -c "import torch; print(torch.__version__)") | |
| pip install '.[test]' --extra-index-url $extra_index_url | |
| pip list | |
| python -c "import torch; assert torch.__version__ == \"$torch_expected\", f\"torch {torch.__version__} != $torch_expected\"" | |
| python -c "import ray; assert ray.__version__ == \"$ray_expected\", f\"ray {ray.__version__} != $ray_expected\"" | |
| shell: bash | |
| - name: Install Neuropod backend | |
| run: | | |
| sudo mkdir -p "$NEUROPOD_BASE_DIR" | |
| curl -L https://github.com/uber/neuropod/releases/download/v${{ env.NEUROPOD_VERSION }}/libneuropod-cpu-linux-v${{ env.NEUROPOD_VERSION }}-torchscript-${{ env.TORCHSCRIPT_VERSION }}-backend.tar.gz | sudo tar -xz -C "$NEUROPOD_BASE_DIR" | |
| shell: bash | |
| - name: Tests | |
| env: | |
| TRANSFORMERS_CACHE: "/root/huggingface_cache" | |
| run: | | |
| RUN_PRIVATE=1 LUDWIG_TEST_SUITE_TIMEOUT_S=6000 pytest -vs --timeout 450 --durations 100 -m "($MARKERS) and (not $EXCLUDED_MARKERS)" --junitxml pytest_slow.xml tests | |
| - name: Upload Unit Test Results | |
| if: ${{ always() && !env.ACT }} | |
| uses: actions/upload-artifact@v2 | |
| with: | |
| name: Unit Test Results (Python ${{ matrix.python-version }} ${{ matrix.test-markers }}) | |
| path: pytest.xml | |
| stop-runner: | |
| name: Stop self-hosted EC2 runner | |
| # required to stop the runner even if the error happened in the previous job | |
| if: always() && needs.start-runner.result != 'skipped' | |
| needs: | |
| - start-runner # required to get output from the start-runner job | |
| - slow-pytest # required to wait when the main job is done | |
| runs-on: ubuntu-latest | |
| steps: | |
| - name: Configure AWS credentials | |
| uses: aws-actions/configure-aws-credentials@v1 | |
| with: | |
| aws-access-key-id: ${{ secrets.AWS_ACCESS_KEY_ID }} | |
| aws-secret-access-key: ${{ secrets.AWS_SECRET_ACCESS_KEY }} | |
| aws-region: ${{ secrets.AWS_REGION }} | |
| - name: Stop EC2 runner | |
| uses: machulav/[email protected] | |
| with: | |
| mode: stop | |
| github-token: ${{ secrets.GH_PERSONAL_ACCESS_TOKEN }} | |
| label: ${{ needs.start-runner.outputs.label }} | |
| ec2-instance-id: ${{ needs.start-runner.outputs.ec2-instance-id }} |