Update qa_l1_test.yml #2
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: Plugin Tests | |
| on: | |
| push: | |
| branches: | |
| - test-ye | |
| pull_request: | |
| branches: | |
| - main | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}-${{ github.actor }} | |
| cancel-in-progress: true | |
| jobs: | |
| plugin-test: | |
| runs-on: [ self-hosted, TE_FL ] | |
| defaults: | |
| run: | |
| shell: bash | |
| container: | |
| image: localhost:5000/flagscale-cuda12.8.1-torch2.7.1-python3.10-te2.9:20260209 | |
| ports: | |
| - 80:80 | |
| options: >- | |
| --gpus all | |
| --shm-size=500g | |
| --privileged | |
| --ipc=host | |
| --ulimit memlock=-1 | |
| --ulimit stack=67108864 | |
| --ulimit nofile=65535:65535 | |
| --user root | |
| --pull always | |
| steps: | |
| - name: Checkout Code | |
| uses: actions/checkout@v6.0.1 | |
| with: | |
| repository: ${{ github.event.pull_request.head.repo.full_name }} | |
| ref: ${{ github.event.pull_request.head.ref }} | |
| ssh-strict: true | |
| ssh-user: git | |
| persist-credentials: true | |
| clean: true | |
| sparse-checkout-cone-mode: true | |
| fetch-tags: false | |
| show-progress: true | |
| lfs: false | |
| submodules: recursive | |
| set-safe-directory: true | |
| - name: Install dependencies and build transformer_engine | |
| timeout-minutes: 60 | |
| env: | |
| NVTE_FRAMEWORK: pytorch | |
| TE_WITH_NCCL: 1 | |
| run: | | |
| echo "Install transformer_engine" | |
| source /opt/miniconda3/etc/profile.d/conda.sh | |
| conda activate flagscale-train | |
| pip install transformers expecttest | |
| pip install --no-build-isolation -vvv . --no-deps | |
| # Verify installation | |
| python3 tests/pytorch/test_sanity_import.py | |
| - name: GPU Usage Check / Verification | |
| run: | | |
| source .github/workflows/scripts/gpu_check.sh | |
| wait_for_gpu | |
| - name: Plugin Test | |
| timeout-minutes: 30 | |
| run: | | |
| source /opt/miniconda3/etc/profile.d/conda.sh | |
| conda activate flagscale-train | |
| pip install pytest | |
| torchrun --nproc_per_node=8 -m pytest -q -x -p no:warnings transformer_engine/plugin/tests |