Skip to content

Smoke test

Smoke test #132

name: Smoke test
on:
# regular runs
schedule:
# Every day at 08:00 UTC
- cron: '0 8 * * *'
# allow button click
workflow_dispatch:
permissions:
contents: read
issues: write
jobs:
test:
runs-on: ubuntu-latest
environment: Scheduled Testing
steps:
- uses: actions/checkout@v3
- name: Store commit
run: |
echo "GIT_COMMIT=$(git rev-parse HEAD)" >> $GITHUB_ENV
- name: Set up Python 3.10
uses: actions/setup-python@v3
with:
python-version: "3.10"
- name: cache poetry install
uses: actions/cache@v3
id: cache-poetry
with:
path: ~/.local
key: poetry-1.7.1-0
- name: Install and configure Poetry
uses: snok/install-poetry@v1
with:
version: 1.7.1
virtualenvs-create: true
virtualenvs-in-project: true
- name: cache deps
id: cache-deps
uses: actions/cache@v3
with:
path: .venv
key: pydeps-${{ hashFiles('**/poetry.lock') }}
- name: Install dependencies with caching
run: poetry install --no-interaction --no-root
if: steps.cache-deps.outputs.cache-hit != 'true'
- name: Install with plugins
run: poetry install --no-interaction --sync --extras all_plugins
- name: Write secrets
env:
SECRETS_CONFIG: |
[together]
api_key = "${{ secrets.TOGETHER_API_KEY }}"
[openai]
api_key = "${{ secrets.OPENAI_API_KEY }}"
[demo]
api_key="12345"
run: |
mkdir -p config
echo "$SECRETS_CONFIG" > config/secrets.toml
- name: Test object creation with plugins
run: |
source .venv/bin/activate
pytest --expensive-tests
- name: Ensure the artifact published on Pypi still works as expected
run: |
rm -rf .venv
mkdir -p ../installation/config
cat ./tests/data/install_pyproject.toml > ../installation/pyproject.toml
cd ../installation
touch ./config/secrets.toml
poetry lock
poetry install --no-root
poetry run modelgauge list-tests
- name: Test standard run
run: |
source .venv/bin/activate
modelbench benchmark --debug -m 1
- uses: JasonEtco/create-an-issue@v2
if: failure()
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
RUN_ID: ${{ github.run_id }}
with:
filename: .github/failed-scheduled-issue.md
- name: Test v1 run
run: |
source .venv/bin/activate
modelbench benchmark --debug -m 1 --benchmark GeneralPurposeAiChatBenchmarkV1
- uses: JasonEtco/create-an-issue@v2
if: failure()
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
RUN_ID: ${{ github.run_id }}
with:
filename: .github/failed-scheduled-issue.md