training.yml

name: Full training validation
on:
  workflow_dispatch:
  schedule:
    - cron: '23 6 * * 1'
permissions:
  contents: read
concurrency:
  group: full-training-${{ github.ref }}
  cancel-in-progress: false
jobs:
  notebooks:
    name: Train ${{ matrix.notebook }}
    runs-on: ubuntu-24.04
    timeout-minutes: 210
    strategy:
      fail-fast: false
      max-parallel: 3
      matrix:
        notebook:
          - Appendix/Appendix.ipynb
          - Chapter 1/Ch1_Introduction.ipynb
          - Chapter 2/Ch2_book.ipynb
          - Chapter 3/Ch3_book.ipynb
          - Chapter 4/Ch4_book.ipynb
          - Chapter 5/Ch5_book.ipynb
          - Chapter 6/Ch6_book.ipynb
          - Chapter 6/MNIST Genetic Algorithm.ipynb
          - Chapter 7/Ch7_book.ipynb
          - Chapter 8/Ch8_book.ipynb
          - Chapter 9/Ch9_book.ipynb
          - Chapter 10/Ch10_book.ipynb
          - old_but_more_detailed/Ch4_PolicyGradients_.ipynb
    steps:
      - uses: actions/checkout@v4
        with:
          persist-credentials: false
      - uses: actions/setup-python@v5
        with:
          python-version: '3.13'
          cache: pip
      - name: Install dependencies
        run: |
          python -m pip install --upgrade pip
          python -m pip install torch==2.13.0 torchvision==0.28.0 --index-url https://download.pytorch.org/whl/cpu
          python -m pip install -r requirements.txt
          python -m pip check
      - name: Execute all cells with full training budgets
        env:
          NOTEBOOK: ${{ matrix.notebook }}
        run: python scripts/check_notebooks.py --mode full --notebook "$NOTEBOOK"
      - name: Train standalone genetic CartPole with default budget
        if: matrix.notebook == 'Chapter 6/Ch6_book.ipynb'
        env:
          OMP_NUM_THREADS: '1'
          MKL_NUM_THREADS: '1'
        run: |
          mkdir -p artifacts/standalone
          python "Chapter 6/main.py" --save_file artifacts/standalone/weights.pth > artifacts/standalone/training.log 2>&1
      - name: Retain notebook logs and metrics
        if: always()
        uses: actions/upload-artifact@v4
        with:
          name: training-${{ strategy.job-index }}
          path: artifacts/
          retention-days: 30