Optimize the auto num_proc calculation of operators in ray mode #694
This file contains hidden or 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, run tests and lint with a single version of Python | |
| # For more information see: https://docs.github.com/en/actions/automating-builds-and-tests/building-and-testing-python | |
| name: unittest-partial | |
| on: | |
| workflow_dispatch: | |
| pull_request: | |
| permissions: | |
| contents: read | |
| env: | |
| ACTIONS_ALLOW_USE_UNSECURE_NODE_VERSION: true | |
| jobs: | |
| unittest-single: | |
| runs-on: [GPU, unittest] | |
| environment: Testing | |
| steps: | |
| - uses: actions/checkout@v3 | |
| with: | |
| path: dj-${{ github.run_id }} | |
| fetch-depth: 0 | |
| - name: Setup docker compose | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose up -d | |
| - name: Install data-juicer | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head bash -c '/root/.local/bin/uv pip install --system -e .\[all\]' | |
| - name: Print Pip Dependency Tree | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head bash -c '/root/.local/bin/uv pip install --system pipdeptree' | |
| docker compose exec ray-head bash -c 'pipdeptree' | |
| - name: Clean dataset cache | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head rm -rf /data/huggingface/dataset | |
| - name: Run unittest standalone | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| if [ "${{ github.event.pull_request.head.repo.full_name }}" != "${{ github.repository }}" ]; then | |
| docker compose exec -e OPENAI_BASE_URL=${{ secrets.OPENAI_BASE_URL }} -e OPENAI_API_KEY=${{ secrets.OPENAI_API_KEY }} ray-head bash -c 'python tests/run.py --tag standalone --mode partial --from-fork True' | |
| else | |
| docker compose exec -e OPENAI_BASE_URL=${{ secrets.OPENAI_BASE_URL }} -e OPENAI_API_KEY=${{ secrets.OPENAI_API_KEY }} ray-head bash -c 'python tests/run.py --tag standalone --mode partial' | |
| fi | |
| docker compose exec ray-head bash -c 'coverage combine' | |
| docker compose exec ray-head bash -c 'mv .coverage .coverage.standalone' | |
| - name: Upload coverage report of standalone | |
| uses: actions/upload-artifact@v4 | |
| with: | |
| name: coverage_report_standalone | |
| include-hidden-files: true | |
| path: dj-${{ github.run_id }}/.coverage.standalone | |
| - name: Remove docker compose | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| if: always() | |
| run: | | |
| docker compose down --remove-orphans | |
| - name: Cleanup workspace | |
| if: always() | |
| run: | | |
| rm -rf dj-${{ github.run_id }} | |
| unittest-dist: | |
| runs-on: [GPU, unittest] | |
| environment: Testing | |
| steps: | |
| - uses: actions/checkout@v3 | |
| with: | |
| path: dj-${{ github.run_id }} | |
| fetch-depth: 0 | |
| - name: Setup docker compose | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose up -d | |
| - name: Install data-juicer | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head bash -c '/root/.local/bin/uv pip install --system -e .\[all\]' | |
| docker compose exec ray-worker bash -c '/root/.local/bin/uv pip install --system -e .\[all\]' | |
| - name: Clean dataset cache | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head rm -rf /data/huggingface/dataset | |
| - name: Run unittest ray | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head bash -c 'python tests/run.py --tag ray --mode partial' | |
| docker compose exec ray-head bash -c 'coverage combine' | |
| docker compose exec ray-head bash -c 'mv .coverage .coverage.ray' | |
| - name: Upload coverage report of ray | |
| uses: actions/upload-artifact@v4 | |
| with: | |
| name: coverage_report_ray | |
| include-hidden-files: true | |
| path: dj-${{ github.run_id }}/.coverage.ray | |
| - name: Remove docker compose | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| if: always() | |
| run: | | |
| docker compose down --remove-orphans | |
| - name: Cleanup workspace | |
| if: always() | |
| run: | | |
| rm -rf dj-${{ github.run_id }} | |
| combine-coverage: | |
| needs: [unittest-single, unittest-dist] | |
| runs-on: [GPU, unittest] | |
| steps: | |
| - uses: actions/checkout@v3 | |
| with: | |
| path: dj-${{ github.run_id }} | |
| fetch-depth: 0 | |
| - name: Setup docker compose | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose up -d | |
| - name: Install coverage | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head bash -c '/root/.local/bin/uv pip install --system coverage' | |
| - name: Download Coverage Report Standalone | |
| uses: actions/download-artifact@v4 | |
| with: | |
| name: coverage_report_standalone | |
| path: dj-${{ github.run_id }} | |
| - name: Download Coverage Report Ray | |
| uses: actions/download-artifact@v4 | |
| with: | |
| name: coverage_report_ray | |
| path: dj-${{ github.run_id }} | |
| - name: Combine Coverage Reports | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| run: | | |
| docker compose exec ray-head bash -c 'coverage combine' | |
| docker compose exec ray-head bash -c 'coverage report -m' | |
| - name: Upload Overall Coverage Report | |
| uses: actions/upload-artifact@v4 | |
| with: | |
| name: coverage_report_all | |
| include-hidden-files: true | |
| path: dj-${{ github.run_id }}/.coverage | |
| - name: Remove docker compose | |
| working-directory: dj-${{ github.run_id }}/.github/workflows/docker | |
| if: always() | |
| run: | | |
| docker compose down --remove-orphans | |
| - name: Cleanup workspace | |
| if: always() | |
| run: | | |
| rm -rf dj-${{ github.run_id }} |