ci-testing.yml

name: CI testing

# see: https://help.github.com/en/actions/reference/events-that-trigger-workflows
on: # Trigger the workflow on push or pull request, but only for the main branch
  push:
    branches: [main]
  pull_request: {}

concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}-${{ github.head_ref }}
  cancel-in-progress: ${{ github.ref != 'refs/heads/main' }}

defaults:
  run:
    shell: bash

env:
  #CI: "true"
  TORCH_URL_RC: "https://download.pytorch.org/whl/test/cpu/torch_test.html"
  TORCH_URL_NIGHTLY: "https://download.pytorch.org/whl/nightly/cpu"
  TORCH_URL_STABLE: "https://download.pytorch.org/whl/cpu"

jobs:
  pytester:
    runs-on: ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: ["ubuntu-22.04", "macOS-15", "windows-latest"]
        python-version: ["3.10"]
        requires: ["latest", "nightly"] # , 'oldest'
        suite: ["core", "ops", "grads"]
        include:
          - { os: "ubuntu-22.04", python-version: "3.10", requires: "oldest", suite: "core" }
          - { os: "ubuntu-22.04", python-version: "3.10", requires: "oldest", suite: "ops" }
          - { os: "ubuntu-22.04", python-version: "3.10", requires: "oldest", suite: "grads" }
          - { os: "ubuntu-24.04", python-version: "3.11", requires: "latest" }
          - { os: "ubuntu-24.04", python-version: "3.12", requires: "latest" }
          - { os: "ubuntu-24.04", python-version: "3.13", requires: "latest" }
        exclude:
          - { os: "windows-latest", suite: "ops" }
          - { os: "windows-latest", suite: "grads" }

    # Timeout: https://stackoverflow.com/a/59076067/4521646
    timeout-minutes: 35

    steps:
      - uses: actions/checkout@v5
      - name: Set up Python ${{ matrix.python-version }}
        uses: actions/setup-python@v6
        with:
          python-version: ${{ matrix.python-version }}
          cache: "pip"

      - name: Setup Ubuntu
        if: runner.os == 'Linux'
        run: |
          sudo apt-get update -qq --fix-missing
          sudo apt-get install -y libblas3 liblapack3 libopenblas-dev
          pip install --force-reinstall numpy

      - name: Set min. dependencies
        if: matrix.requires == 'oldest'
        run: |
          for fpath in ('requirements/base.txt', 'requirements/test.txt'):
              req = open(fpath).read().replace('>=', '==')
              open(fpath, 'w').write(req)
        shell: python

      - name: set default Torch source
        run: echo "TORCH_URL=$TORCH_URL_STABLE" >> $GITHUB_ENV
      - name: switch Torch source to nightly
        if: matrix.requires == 'nightly'
        run: |
          echo "TORCH_URL=$TORCH_URL_NIGHTLY" >> $GITHUB_ENV
          echo "PIP_EXTRA_FLAG=--pre" >> $GITHUB_ENV

      - name: Install Mac specific dependencies
        if: runner.os == 'macOS'
        run: brew install libomp

      - name: Install package & dependencies
        run: |
          pip install -e . -U \
            ${PIP_EXTRA_FLAG} -r requirements/test.txt \
            --extra-index-url=${TORCH_URL}
          pip list

      - name: re-Install dependencies
        if: matrix.requires != 'nightly'
        run: |
          pip install -q wget
          # download script and adjust torch versions in requirements
          python -m wget https://raw.githubusercontent.com/Lightning-AI/utilities/main/scripts/adjust-torch-versions.py
          for fpath in `ls requirements/*.txt`; do \
            python ./adjust-torch-versions.py $fpath ${{ matrix.pytorch-version }}; \
          done
          # re-install with adjusted torch versions
          pip install ${PIP_EXTRA_FLAG} -r requirements/devel.txt \
            --extra-index-url=${TORCH_URL}
          pip list

      - name: Testing Local
        if: matrix.suite == 'core'
        run: |
          coverage run --source thunder -m \
            pytest thunder/tests/ \
              --ignore=thunder/tests/distributed \
              --ignore=thunder/tests/test_ops.py \
              --ignore=thunder/tests/test_grad.py \
              -v --datefmt="%Y%m%d-%H:%M:%S.%f" \
              --random-order-seed=$GITHUB_RUN_ID \
              -n 4 --durations=250

      - name: Testing Distributed
        # run all found tests in given past as standalone
        if: runner.os == 'Linux' && matrix.suite == 'core'
        run: |
          pytest thunder/tests/distributed/ \
            -v --datefmt="%Y%m%d-%H:%M:%S.%f" \
            --random-order-seed=$GITHUB_RUN_ID \
            --durations=250

      - name: Testing Ops
        if: matrix.suite == 'ops'
        run: |
          coverage run --source thunder -m \
            pytest thunder/tests/test_ops.py \
              -v --datefmt="%Y%m%d-%H:%M:%S.%f" \
              --random-order-seed=$GITHUB_RUN_ID \
              -n 4 --durations=250

      - name: Testing Grads
        if: matrix.suite == 'grads'
        run: |
          coverage run --source thunder -m \
            pytest thunder/tests/test_grad.py \
              -v --datefmt="%Y%m%d-%H:%M:%S.%f" \
              --random-order-seed=$GITHUB_RUN_ID \
              -n 4 --durations=250

      - name: Testing interpreter
        if: matrix.python-version != '3.10'
        run: |
          python -m pytest \
            thunder/tests/test_interpreter.py \
            thunder/tests/test_jit_general.py \
            -v -n 4 --durations=50 --cov=thunder
      - name: Statistics
        run: |
          coverage report
          coverage xml

      - name: Upload coverage to Codecov
        uses: codecov/codecov-action@v5
        with:
          token: ${{ secrets.CODECOV_TOKEN }}
          file: ./coverage.xml
          flags: unittests
          env_vars: OS,PYTHON
          name: codecov-umbrella
          fail_ci_if_error: false

  testing-guardian:
    runs-on: ubuntu-latest
    needs: pytester
    if: always()
    steps:
      - run: echo "${{ needs.pytester.result }}"
      - name: failing...
        if: needs.pytester.result == 'failure'
        run: exit 1
      - name: cancelled or skipped...
        if: contains(fromJSON('["cancelled", "skipped"]'), needs.pytester.result)
        timeout-minutes: 1
        run: sleep 90