Add vllm to cross provider engine #964
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Test Inference Engines | |
| on: | |
| pull_request: | |
| paths: | |
| - 'src/unitxt/inference.py' | |
| - 'tests/inference/*.py' | |
| - '.github/workflows/inference_tests.yml' | |
| release: | |
| types: [published] | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.event_name == 'pull_request' && github.event.pull_request.number || github.ref_name }} | |
| cancel-in-progress: true | |
| jobs: | |
| inference: | |
| runs-on: ubuntu-latest | |
| timeout-minutes: 30 | |
| env: | |
| OS: ubuntu-latest | |
| UNITXT_DEFAULT_VERBOSITY: error | |
| DATASETS_VERBOSITY: error | |
| HF_HUB_VERBOSITY: error | |
| HF_DATASETS_DISABLE_PROGRESS_BARS: "True" | |
| TQDM_DISABLE: "True" | |
| WML_URL: ${{ secrets.WML_URL }} | |
| WML_PROJECT_ID: ${{ secrets.WML_PROJECT_ID }} | |
| WML_APIKEY: ${{ secrets.WML_APIKEY }} | |
| WX_URL: ${{ secrets.WML_URL }} # Similar to WML_URL | |
| WX_PROJECT_ID: ${{ secrets.WML_PROJECT_ID }} # Similar to WML_PROJECT_ID | |
| WX_API_KEY: ${{ secrets.WML_APIKEY }} # Similar to WML_APIKEY | |
| GENAI_KEY: ${{ secrets.GENAI_KEY }} | |
| steps: | |
| - uses: actions/checkout@v4 | |
| - uses: actions/setup-python@v5 | |
| with: | |
| python-version: '3.10' | |
| - run: curl -LsSf https://astral.sh/uv/install.sh | sh | |
| - run: uv pip install --system ".[tests,watsonx,inference-tests]" | |
| - name: Hugging Face Login | |
| run: | | |
| for i in {1..5}; do | |
| huggingface-cli login --token ${{ secrets.UNITXT_READ_HUGGINGFACE_HUB_FOR_TESTS }} && break || sleep $((2 ** i)); | |
| done | |
| - name: Install Ollama | |
| run: curl -fsSL https://ollama.com/install.sh | sh | |
| - name: Pull Llama 3.2:1b model | |
| run: ollama pull llama3.2:1b | |
| - name: Start serving the model | |
| run: nohup ollama serve --model llama3.2:1b --port 5000 & | |
| - name: Run Tests | |
| run: python -m unittest discover -s tests/inference -p "test_*.py" |