[Triton] FA v3 API #451
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: Aiter Test | |
| on: | |
| push: | |
| branches: [main] | |
| pull_request: | |
| branches: [main] # Triggers on PRs targeting `main` | |
| workflow_dispatch: | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.ref }} | |
| cancel-in-progress: true | |
| env: | |
| DOCKER_IMAGE: "rocm/pytorch:latest" | |
| jobs: | |
| define-runners: | |
| runs-on: ubuntu-latest | |
| outputs: | |
| standard_runners: ${{ steps.machines.outputs.standard_runners }} | |
| multigpu_runners: ${{ steps.machines.outputs.multigpu_runners }} | |
| steps: | |
| - name: Define whether runs on MI35X | |
| id: machines | |
| run: | | |
| set -euo pipefail | |
| pr_title="${{ github.event.pull_request.title }}" | |
| if [[ "${{ github.ref }}" == "refs/heads/main" ]]; then | |
| echo "It's main branch, running tests on MI300 and MI35X..." | |
| echo 'standard_runners=["aiter-mi355-1gpu"]' >> "$GITHUB_OUTPUT" | |
| echo 'multigpu_runners=["aiter-mi355-8gpu"]' >> "$GITHUB_OUTPUT" | |
| elif echo "$pr_title" | grep -qi "mi35x"; then | |
| echo "PR title contains 'MI35X', running tests on MI300 and MI35X..." | |
| echo 'standard_runners=["aiter-mi355-1gpu"]' >> "$GITHUB_OUTPUT" | |
| echo 'multigpu_runners=["aiter-mi355-8gpu"]' >> "$GITHUB_OUTPUT" | |
| else | |
| echo "Not main branch and PR title does not contain mi35x, only running on MI300..." | |
| echo 'standard_runners=["aiter-mi355-1gpu"]' >> "$GITHUB_OUTPUT" | |
| echo 'multigpu_runners=["aiter-mi355-8gpu"]' >> "$GITHUB_OUTPUT" | |
| fi | |
| echo "$GITHUB_OUTPUT" | |
| - name: Show output variable | |
| run: | | |
| echo "Standard: ${{ steps.machines.outputs.standard_runners }}" | |
| echo "Multipe: ${{ steps.machines.outputs.multigpu_runners }}" | |
| standard: | |
| needs: define-runners | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| runner: ${{ fromJSON(needs.define-runners.outputs.standard_runners) }} | |
| runs-on: ${{ matrix.runner }} | |
| steps: | |
| - name: Checkout code | |
| uses: actions/checkout@v4 | |
| - name: Sync submodules | |
| run: | | |
| set -euo pipefail | |
| if [[ ${{ github.ref }} == "refs/heads/main" ]]; then | |
| echo "It's main branch, syncing latest CK..." | |
| git submodule sync | |
| git submodule update --init --recursive --remote --depth 1 --jobs 4 | |
| else | |
| echo "It's a PR branch, syncing specific CK..." | |
| git submodule sync | |
| git submodule update --init --recursive --depth 1 --jobs 4 | |
| fi | |
| - name: Clean up Rocm processes | |
| run: | | |
| ./.github/scripts/clean_up_rocm.sh | |
| - name: Run the container | |
| run: | | |
| set -ex | |
| echo "Starting container: aiter_test" | |
| docker run -dt \ | |
| --device=/dev/dri \ | |
| --device=/dev/kfd \ | |
| --shm-size=16G \ | |
| --group-add $(getent group render | cut -d: -f3) \ | |
| --group-add $(getent group video | cut -d: -f3) \ | |
| -v "${{ github.workspace }}:/workspace" \ | |
| -w /workspace \ | |
| --name aiter_test \ | |
| ${{ env.DOCKER_IMAGE }} | |
| - name: Setup-Triton | |
| run: | | |
| set -ex | |
| echo "Setting up Triton..." | |
| docker exec \ | |
| -w /workspace \ | |
| aiter_test \ | |
| ./.github/scripts/build_triton.sh | |
| - name: Tests | |
| run: | | |
| set -ex | |
| docker exec \ | |
| -w /workspace \ | |
| aiter_test \ | |
| ./.github/scripts/aiter_test.sh | |
| - name: Upload test logs | |
| uses: actions/upload-artifact@v4 | |
| if: always() | |
| with: | |
| name: standard-test-log-${{ matrix.runner }} | |
| path: latest_test.log | |
| - name: Cleanup container | |
| if: always() | |
| run: | | |
| docker rm -f aiter_test || true | |
| - name: Clean up Rocm processes | |
| if: always() | |
| run: | | |
| ./.github/scripts/clean_up_rocm.sh | |
| multi-gpu: | |
| needs: define-runners | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| runner: ${{ fromJSON(needs.define-runners.outputs.multigpu_runners) }} | |
| runs-on: ${{ matrix.runner }} | |
| steps: | |
| - name: Checkout code | |
| uses: actions/checkout@v4 | |
| - name: Sync submodules | |
| run: | | |
| set -euo pipefail | |
| if [[ ${{ github.ref }} == "refs/heads/main" ]]; then | |
| echo "It's main branch, syncing latest CK..." | |
| git submodule sync | |
| git submodule update --init --recursive --remote --depth 1 --jobs 4 | |
| else | |
| echo "It's a PR branch, syncing specific CK..." | |
| git submodule sync | |
| git submodule update --init --recursive --depth 1 --jobs 4 | |
| fi | |
| - name: Clean up Rocm processes | |
| run: ./.github/scripts/clean_up_rocm.sh | |
| - name: Run the container | |
| run: | | |
| set -ex | |
| echo "Starting container: aiter_test" | |
| docker run -dt \ | |
| --device=/dev/dri \ | |
| --device=/dev/kfd \ | |
| --shm-size=16G \ | |
| --group-add $(getent group render | cut -d: -f3) \ | |
| --group-add $(getent group video | cut -d: -f3) \ | |
| -v "${{ github.workspace }}:/workspace" \ | |
| -w /workspace \ | |
| --name aiter_test \ | |
| ${{ env.DOCKER_IMAGE }} | |
| - name: Setup-Triton | |
| run: | | |
| set -ex | |
| echo "Setting up Triton..." | |
| docker exec \ | |
| -w /workspace \ | |
| aiter_test \ | |
| ./.github/scripts/build_triton.sh | |
| - name: Tests | |
| run: | | |
| set -ex | |
| docker exec \ | |
| -e MULTIGPU=TRUE \ | |
| -w /workspace \ | |
| aiter_test \ | |
| ./.github/scripts/aiter_test.sh | |
| - name: Upload test logs | |
| uses: actions/upload-artifact@v4 | |
| if: always() | |
| with: | |
| name: multigpu-test-${{ matrix.runner }} | |
| path: latest_test.log | |
| - name: Cleanup container | |
| if: always() | |
| run: | | |
| docker rm -f aiter_test || true | |
| - name: Clean up Rocm processes | |
| if: always() | |
| run: | | |
| ./.github/scripts/clean_up_rocm.sh |