Run Tests workflow (mir-group/allegro)
The Run Tests workflow from mir-group/allegro, explained and optimized by Latchkey.
CI health: C - fair
Point runs-on at Latchkey and get caching, job timeouts, SHA-pinned actions, self-healing for flaky steps, and up to 58% lower cost, applied automatically.
What it does
This is the Run Tests workflow from the mir-group/allegro repository, a real project running GitHub Actions. It is shown here with attribution under its MIT license.
Below, Latchkey shows a faster, safer version produced by its optimization engine.
The workflow
name: Run Tests
on:
push:
branches:
- main
- develop
pull_request:
branches:
- main
- develop
concurrency:
# github.workflow will be the name of the workflow, e.g. "Run Tests"
# github.ref will be the branch or tag ref, e.g. "refs/heads/main" or, for a PR, "refs/pull/123/merge"
group: ${{ github.workflow }}-${{ github.ref }}
cancel-in-progress: true
# yamllint disable rule:line-length
jobs:
lint:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: "3.12"
- name: Install pre-commit
run: pip install pre-commit
- name: Run pre-commit
run: pre-commit run --all-files --show-diff-on-failure
tests:
needs: lint
runs-on: ${{ github.event.repository.private && 'self-hosted' || 'ubuntu-latest' }}
strategy:
matrix:
torch-version: [2.2.0, 2.10.0]
container:
image: nvidia/cuda:12.6.1-cudnn-devel-ubuntu24.04
options: ${{ github.event.repository.private && '--gpus all' || ' ' }}
steps:
- uses: actions/checkout@v4
- name: Install Python and create venv
run: |
apt update && apt install -y python3 python3-pip python3-venv git
python3 -m venv /opt/venv
echo "/opt/venv/bin" >> $GITHUB_PATH
- name: Check Python version
run: |
python3 --version
which python3
- name: Disable pip cache
run: |
echo "PIP_NO_CACHE_DIR=1" >> $GITHUB_ENV
echo "PIP_NO_CACHE=1" >> $GITHUB_ENV
- name: Configure torch cache directories
run: |
echo "TORCH_EXTENSIONS_DIR=/tmp/torch_extensions_${GITHUB_RUN_ID}_${GITHUB_RUN_ATTEMPT}" >> $GITHUB_ENV
echo "TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_${GITHUB_RUN_ID}_${GITHUB_RUN_ATTEMPT}" >> $GITHUB_ENV
echo "TRITON_CACHE_DIR=/tmp/triton_${GITHUB_RUN_ID}_${GITHUB_RUN_ATTEMPT}" >> $GITHUB_ENV
- name: Clone NequIP packages
uses: mir-group/nequip-gh-actions/clone@main
with:
package_names: "nequip"
ssh_keys: "${{ secrets.NEQUIP }}"
private_flag: ${{ github.event.repository.private}}
branch: ${{ github.event_name == 'pull_request' && github.base_ref || github.ref_name}}
- name: Install PyTorch
env:
TORCH: "${{ matrix.torch-version }}"
run: |
# use CPU only on GH runner
if [ "${{ github.event.repository.private }}" = "true" ]; then
pip install torch==${TORCH} --upgrade
else
pip install torch==${TORCH} --index-url https://download.pytorch.org/whl/cpu --upgrade
fi
- name: Install Test Dependencies
run: |
pip install setuptools wheel pytest pytest-xdist[psutil]
# Pin numpy<2 and matscipy 1.1.1 for torch 2.2 to avoid compatibility issues
if [ "${{ matrix.torch-version }}" = "2.2.0" ]; then
pip install "numpy<2" "matscipy==1.1.1"
fi
# install nequip packages
pip install --upgrade-strategy only-if-needed ./nequip
pip install --upgrade-strategy only-if-needed .
if python3 -c "import torch; exit(0 if torch.cuda.is_available() else 1)"; then
echo "Installing CuEquivariance (CUDA available)"
pip install cuequivariance-torch cuequivariance-ops-torch-cu12
else
echo "Skipping CuEquivariance installation (CUDA not available)"
fi
- name: Run tests
run: |
# See https://github.com/pytest-dev/pytest/issues/1075 for why we need to set PYTHONHASHSEED
PYTHONHASHSEED=0 pytest -n auto tests/
- name: Clean up temporary files
if: always()
run: |
rm -rf ./allegro
# yamllint enable
The same workflow, on Latchkey
Estimated ~20% faster on cache hits, plus fewer wasted runs and a safer supply chain. Added and changed lines are highlighted.
name: Run Tests on: push: branches: - main - develop pull_request: branches: - main - develop concurrency: # github.workflow will be the name of the workflow, e.g. "Run Tests" # github.ref will be the branch or tag ref, e.g. "refs/heads/main" or, for a PR, "refs/pull/123/merge" group: ${{ github.workflow }}-${{ github.ref }} cancel-in-progress: true # yamllint disable rule:line-length jobs: lint: timeout-minutes: 30 runs-on: latchkey-small steps: - uses: actions/checkout@v4 - name: Set up Python uses: actions/setup-python@v5 with: cache: 'pip' python-version: "3.12" - name: Install pre-commit run: pip install pre-commit - name: Run pre-commit run: pre-commit run --all-files --show-diff-on-failure tests: timeout-minutes: 30 needs: lint runs-on: ${{ github.event.repository.private && 'self-hosted' || 'ubuntu-latest' }} strategy: matrix: torch-version: [2.2.0, 2.10.0] container: image: nvidia/cuda:12.6.1-cudnn-devel-ubuntu24.04 options: ${{ github.event.repository.private && '--gpus all' || ' ' }} steps: - uses: actions/checkout@v4 - name: Install Python and create venv run: | apt update && apt install -y python3 python3-pip python3-venv git python3 -m venv /opt/venv echo "/opt/venv/bin" >> $GITHUB_PATH - name: Check Python version run: | python3 --version which python3 - name: Disable pip cache run: | echo "PIP_NO_CACHE_DIR=1" >> $GITHUB_ENV echo "PIP_NO_CACHE=1" >> $GITHUB_ENV - name: Configure torch cache directories run: | echo "TORCH_EXTENSIONS_DIR=/tmp/torch_extensions_${GITHUB_RUN_ID}_${GITHUB_RUN_ATTEMPT}" >> $GITHUB_ENV echo "TORCHINDUCTOR_CACHE_DIR=/tmp/torchinductor_${GITHUB_RUN_ID}_${GITHUB_RUN_ATTEMPT}" >> $GITHUB_ENV echo "TRITON_CACHE_DIR=/tmp/triton_${GITHUB_RUN_ID}_${GITHUB_RUN_ATTEMPT}" >> $GITHUB_ENV - name: Clone NequIP packages uses: mir-group/nequip-gh-actions/clone@main with: package_names: "nequip" ssh_keys: "${{ secrets.NEQUIP }}" private_flag: ${{ github.event.repository.private}} branch: ${{ github.event_name == 'pull_request' && github.base_ref || github.ref_name}} - name: Install PyTorch env: TORCH: "${{ matrix.torch-version }}" run: | # use CPU only on GH runner if [ "${{ github.event.repository.private }}" = "true" ]; then pip install torch==${TORCH} --upgrade else pip install torch==${TORCH} --index-url https://download.pytorch.org/whl/cpu --upgrade fi - name: Install Test Dependencies run: | pip install setuptools wheel pytest pytest-xdist[psutil] # Pin numpy<2 and matscipy 1.1.1 for torch 2.2 to avoid compatibility issues if [ "${{ matrix.torch-version }}" = "2.2.0" ]; then pip install "numpy<2" "matscipy==1.1.1" fi # install nequip packages pip install --upgrade-strategy only-if-needed ./nequip pip install --upgrade-strategy only-if-needed . if python3 -c "import torch; exit(0 if torch.cuda.is_available() else 1)"; then echo "Installing CuEquivariance (CUDA available)" pip install cuequivariance-torch cuequivariance-ops-torch-cu12 else echo "Skipping CuEquivariance installation (CUDA not available)" fi - name: Run tests run: | # See https://github.com/pytest-dev/pytest/issues/1075 for why we need to set PYTHONHASHSEED PYTHONHASHSEED=0 pytest -n auto tests/ - name: Clean up temporary files if: always() run: | rm -rf ./allegro # yamllint enable
What changed
- Run on Latchkey managed runners with one line (
runs-on), which apply the fixes below automatically and self-heal transient failures. This example useslatchkey-small; pick the runner size that fits the job. - Cache dependency installs on the setup step so they are served from cache.
- Add a job timeout so a hung step cannot burn hours of runner time.
1 third-party action is referenced by a movable tag. Pin it to the commit SHA (Latchkey resolves and applies this automatically) so a repointed tag cannot change what runs.
What Latchkey heals here
This workflow has steps that commonly fail on transient issues (network, registries, flaky browsers). On Latchkey managed runners they are detected, retried, and self-healed instead of failing your build:
- Dependency installs
This workflow runs 2 jobs (3 with the matrix expanded) per trigger. On Latchkey the same minutes cost up to 58% less than GitHub-hosted, with zero queue time.