Refactor llm perf backend handling #155
This file contains 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: CLI CUDA vLLM Tests | |
| on: | |
| workflow_dispatch: | |
| push: | |
| branches: | |
| - main | |
| pull_request: | |
| branches: | |
| - main | |
| types: | |
| - opened | |
| - reopened | |
| - synchronize | |
| - labeled | |
| - unlabeled | |
| concurrency: | |
| cancel-in-progress: true | |
| group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }} | |
| jobs: | |
| run_cli_cuda_vllm_single_gpu_tests: | |
| if: ${{ | |
| (github.event_name == 'push') || | |
| (github.event_name == 'workflow_dispatch') || | |
| contains( github.event.pull_request.labels.*.name, 'cli') || | |
| contains( github.event.pull_request.labels.*.name, 'cuda') || | |
| contains( github.event.pull_request.labels.*.name, 'vllm') || | |
| contains( github.event.pull_request.labels.*.name, 'single_gpu') || | |
| contains( github.event.pull_request.labels.*.name, 'cli_cuda_vllm_single_gpu') | |
| }} | |
| runs-on: [single-gpu, nvidia-gpu, a10, ci] | |
| container: | |
| image: vllm/vllm-openai:latest | |
| options: --ipc host --gpus all --entrypoint /bin/bash | |
| steps: | |
| - name: Checkout | |
| uses: actions/checkout@v4 | |
| - name: Install dependencies | |
| run: | | |
| pip install -e .[testing] | |
| - name: Run tests | |
| run: | | |
| FORCE_SERIAL=1 pytest -x -s -k "cli and cuda and vllm and not (tp or pp)" | |
| run_cli_cuda_vllm_multi_gpu_tests: | |
| if: ${{ | |
| (github.event_name == 'push') || | |
| (github.event_name == 'workflow_dispatch') || | |
| contains( github.event.pull_request.labels.*.name, 'cli') || | |
| contains( github.event.pull_request.labels.*.name, 'cuda') || | |
| contains( github.event.pull_request.labels.*.name, 'vllm') || | |
| contains( github.event.pull_request.labels.*.name, 'multi_gpu') || | |
| contains( github.event.pull_request.labels.*.name, 'cli_cuda_vllm_multi_gpu') | |
| }} | |
| runs-on: [multi-gpu, nvidia-gpu, 4-a10, ci] | |
| container: | |
| image: vllm/vllm-openai:latest | |
| options: --ipc host --gpus all --entrypoint /bin/bash | |
| steps: | |
| - name: Checkout | |
| uses: actions/checkout@v4 | |
| - name: Install dependencies | |
| run: | | |
| pip install -e .[testing] | |
| - name: Run tests | |
| run: | | |
| FORCE_SERIAL=1 pytest -x -s -k "cli and cuda and vllm and (tp or pp)" |