Add tensor parallelism for attention QKVO #3211
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: PR GPU tests | |
| on: | |
| push: | |
| branches: | |
| - main | |
| - release/* | |
| pull_request_target: | |
| branches: | |
| - main | |
| - release/** | |
| workflow_dispatch: | |
| # Cancel old runs when a new commit is pushed to the same branch if not on main or dev | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }} | |
| cancel-in-progress: ${{ github.ref != 'refs/heads/main' }} | |
| jobs: | |
| pytest-gpu: | |
| uses: ./.github/workflows/pytest-gpu.yaml | |
| strategy: | |
| matrix: | |
| # TODO: After the PR with the flash attention 2 images goes in, add the new unit test suite | |
| include: | |
| - name: 'gpu-latest' | |
| container: mosaicml/pytorch:latest # mosaicml/pytorch:1.13.1_cu117-python3.10-ubuntu20.04 | |
| markers: 'gpu' | |
| pytest_command: 'coverage run -m pytest' | |
| - name: 'gpu-2.0.1' | |
| container: mosaicml/pytorch:2.0.1_cu118-python3.10-ubuntu20.04 | |
| markers: 'gpu' | |
| pytest_command: 'coverage run -m pytest' | |
| - name: 'gpu-2.1.0' | |
| container: mosaicml/pytorch:2.1.0_cu121-python3.10-ubuntu20.04 | |
| markers: 'gpu' | |
| pytest_command: 'coverage run -m pytest' | |
| - name: 'gpu-2.1.0-flash2' | |
| container: mosaicml/llm-foundry:2.1.0_cu121_flash2-latest | |
| markers: 'gpu' | |
| pytest_command: 'coverage run -m pytest' | |
| name: ${{ matrix.name }} | |
| if: github.repository_owner == 'mosaicml' | |
| with: | |
| container: ${{ matrix.container }} | |
| mcloud-timeout: 1800 | |
| name: ${{ matrix.name }} | |
| pytest-command: ${{ matrix.pytest_command }} | |
| pytest-markers: ${{ matrix.markers }} | |
| python-version: 3.9 | |
| secrets: | |
| mcloud-api-key: ${{ secrets.MCLOUD_API_KEY }} |