convert_notebooks.yml

# Execute notebooks and convert them to Markdown and HTML

name: Convert Notebooks
on:
  workflow_dispatch:
  push:
    branches:
      - 'main'
      - 'latest'
    paths:
      - 'notebooks/**.ipynb'
      - 'notebooks/**.py'
      - 'requirements.txt'
      - 'README.md'
      - '.ci/*'
      - '.github/workflows/convert_notebooks.yml'
  pull_request:
    paths:
      - '.github/workflows/convert_notebooks.yml'
  schedule:
    - cron: '0 0 * * *'

concurrency:
  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
  cancel-in-progress: true

permissions:
  contents: read

jobs:
  build:
    strategy:
      fail-fast: false

      # Matrix is unnecessary here, but this allows easy copying of steps from treon.yml
      matrix:
        os: [ubuntu-22.04]
        python: [3.8]

    runs-on: ${{ matrix.os }}

    steps:
      - name: Maximize build space
        run: |
          sudo rm -rf /usr/local/lib/android # will release about 10 GB if you don't need Android
          sudo rm -rf /usr/share/dotnet # will release about 20GB if you don't need .NET
          sudo rm -rf /opt/ghc
          echo "Available storage:"
          df -h
      - name: Set Swap Space
        uses: pierotofy/set-swap-space@49819abfb41bd9b44fb781159c033dba90353a7c # master
        with:
          swap-size-gb: 10
      - name: Install required packages for rst converstion
        run: |
          sudo apt-get update && sudo apt-get install texlive texlive-latex-extra pandoc -y
        shell: bash

      #### Installation/preparation ####
      #
      # These steps are copied from convert_notebooks.yml
      # This should ideally be a reusable workflow

      - name: Checkout repository
        uses: actions/checkout@a5ac7e51b41094c92402da3b24376905380afc29 # v4.1.6
      - name: Dotenv Action
        id: dotenv
        uses: xom9ikk/dotenv@ac290ca23a42155a0cba1031d23afa46240116a9 # v2.3.0
        with:
          path: ./.github/workflows
      - name: Set up Python
        uses: actions/setup-python@82c7e631bb3cdc910f68e0081d67478d79c6982d # v5.1.0
        with:
          python-version: ${{ matrix.python }}
      - name: Install required packages
        run: |
          if [ "$RUNNER_OS" == "Linux" ]; then
          sudo apt-get install libsndfile1 -y
          fi
        shell: bash

      - name: Cache OpenVINO Pip Packages
        id: cachepip
        uses: actions/cache@0c45773b623bea8c8e75f6c82b208c3cf94ea4f9 # v4.0.2
        with:
          path: |
            pipcache
          key: ${{ env.PIP_CACHE_KEY }}-${{ matrix.os }}-${{ matrix.python }}

      # Cache specific files to reduce downloads or prevent network issues
      - name: Cache Files
        id: cachefiles
        uses: actions/cache@0c45773b623bea8c8e75f6c82b208c3cf94ea4f9 # v4.0.2
        with:
          path: |
            # NOTE: when modifying cache paths, update FILES_CACHE_KEY in .env
            # and change cache paths in both treon.yml and convert_notebooks.yml
            case_00030.zip
            notebooks/ct-segmentation-quantize/kits19_frames_1
            notebooks/pytorch-post-training-quantization-nncf/output/tiny-imagenet-200.zip
            # omz cache location is set to this with test_replace
            notebooks/optical-character-recognition/open_model_zoo_cache
            notebooks/ct-scan-live-inference/kits19_frames_1
            notebooks/pytorch-quantization-aware-training/data/tiny-imagenet-200.zip
          key: ${{ env.FILES_CACHE_KEY }}

      # PaddleGAN stores cache in ppgan directory in CACHE_DIR
      - name: Set CACHE_DIR
        shell: bash
        run: |
          python -c 'import os;print("CACHE_DIR={0}".format(os.path.expanduser(os.path.join("~", ".cache"))))'
          # replace backslashes with forward slashes for Windows paths
          python -c 'import os;print("CACHE_DIR={0}".format(os.path.expanduser(os.path.join("~", ".cache"))))' | sed -e 's/\\/\//g' >> $GITHUB_ENV

      # PaddleHub stores cache in directory pointed to by HUB_HOME environment variable
      - name: Set HUB_HOME
        shell: bash
        run: |
          echo HUB_HOME=${{ env.CACHE_DIR }}/.paddlehub >> $GITHUB_ENV

      # Cache PaddlePaddle cache directories to prevent CI failing due to network/download issues
      - name: Cache PaddlePaddle cache directories (per OS)
        id: cacheusercache
        uses: actions/cache@0c45773b623bea8c8e75f6c82b208c3cf94ea4f9 # v4.0.2
        with:
          path: |
            ${{ env.HUB_HOME }}
            ${{ env.CACHE_DIR }}/paddle
            ${{ env.CACHE_DIR }}/ppgan
          key: ${{ env.USER_CACHE_KEY }}-${{ runner.os }}
      - name: Cache openvino packages
        if: steps.cachepip.outputs.cache-hit != 'true'
        run: |
          python -m pip install --upgrade pip==21.3.*
          mkdir pipcache
          python -m pip install --cache-dir pipcache --no-deps openvino openvino-dev nncf
          cp -r pipcache pipcache_openvino
          python -m pip uninstall -y openvino openvino-dev nncf

      # Download a small dataset to use for testing purposes in monai-kidney training notebook
      - name: Download CT files
        if: steps.cachefiles.outputs.cache-hit != 'true'
        run: |
          curl -O https://storage.openvinotoolkit.org/data/test_data/openvino_notebooks/kits19/case_00030.zip
      - name: Copy CT files
        run: |
          mkdir notebooks/ct-segmentation-quantize/kits19
          mkdir notebooks/ct-segmentation-quantize/kits19/kits19_frames
          unzip case_00030.zip
          cp -r case_00030 case_00001
          mv case_00030 notebooks/ct-segmentation-quantize/kits19/kits19_frames
          mv case_00001 notebooks/ct-segmentation-quantize/kits19/kits19_frames
      - name: Download shared datasets
        run: |
          mkdir -p notebooks/data/librispeech
          wget  -O notebooks/data/librispeech/test-clean.tar.gz http://openslr.elda.org/resources/12/test-clean.tar.gz
          tar -xvf notebooks/data/librispeech/test-clean.tar.gz
          mkdir -p notebooks/data/cifar10
          wget -O notebooks/data/cifar10/cifar-10-python.tar.gz https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz
      - name: Install dependencies
        run: |
          python -m pip install --upgrade pip==21.3.*
          python -m pip install -r .ci/dev-requirements.txt --cache-dir pipcache
          python -m ipykernel install --user --name openvino_env

        # Cache OpenVINO packages. mv works cross-platform
      - name: Make pipcache directory with OpenVINO packages
        if: steps.cachepip.outputs.cache-hit != 'true'
        run: |
          mv pipcache pipcache_full
          mv pipcache_openvino pipcache

      # Create list of installed pip packages that can be downloaded as artifacts
      # to verify the exact environment of a specific test run
      - name: Pip freeze
        run: |
          python -m pip freeze
          python -m pip freeze > pip-freeze-${{ github.sha }}-${{matrix.os}}-${{ matrix.python }}.txt
      - name: Archive pip freeze
        uses: actions/upload-artifact@65462800fd760344b1a7b4382951275a0abb4808 # v4.3.3
        with:
          name: pip-freeze-${{matrix.os}}-${{ matrix.python }}
          path: pip-freeze-${{ github.sha }}-${{matrix.os}}-${{ matrix.python }}.txt

      #### End installation/preparation

      - name: convert_notebooks
        shell: bash
        run: .ci/convert_notebooks.sh
      - name: Save reStructuredText files
        uses: actions/upload-artifact@65462800fd760344b1a7b4382951275a0abb4808 # v4.3.3
        with:
          name: rst_files
          path: rst_files