Enable PagedAttention through Pallas #7764
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: Build and test | |
| on: | |
| pull_request: | |
| branches: | |
| - master | |
| - r[0-9]+.[0-9]+ | |
| paths-ignore: | |
| - 'experimental/torch_xla2/**' | |
| push: | |
| branches: | |
| - master | |
| - r[0-9]+.[0-9]+ | |
| paths-ignore: | |
| - 'experimental/torch_xla2/**' | |
| workflow_dispatch: | |
| concurrency: | |
| group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.sha }}-${{ github.event_name == 'workflow_dispatch' }}-${{ github.event_name == 'schedule' }} | |
| cancel-in-progress: true | |
| jobs: | |
| # Old CI workflow | |
| build: | |
| name: "Build PyTorch/XLA (GPU)" | |
| uses: ./.github/workflows/_build.yml | |
| with: | |
| ecr-docker-image-base: 308535385114.dkr.ecr.us-east-1.amazonaws.com/pytorch/xla_base | |
| gcr-docker-image: gcr.io/tpu-pytorch/xla_base:dev-3.8_cuda_12.1 | |
| cuda: 1 | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| test-cpp-cpu: | |
| name: "CPU C++ tests" | |
| uses: ./.github/workflows/_test_cpp.yml | |
| needs: build | |
| with: | |
| docker-image: ${{ needs.build.outputs.docker-image }} | |
| timeout-minutes: 120 | |
| collect-coverage: false # TODO(yeounoh) separate from CPU coverage metrics | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| test-cpp-cuda: | |
| name: "GPU C++ tests" | |
| uses: ./.github/workflows/_test_cpp.yml | |
| needs: build | |
| with: | |
| docker-image: ${{ needs.build.outputs.docker-image }} | |
| runner: linux.8xlarge.nvidia.gpu | |
| timeout-minutes: 300 | |
| collect-coverage: false # TODO(yeounoh) separate from CPU coverage metrics | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| push-docs: | |
| name: "Build & publish docs" | |
| if: github.event_name == 'push' && (github.event.ref == 'refs/heads/master' || startsWith(github.event.ref, 'refs/tags/r')) | |
| uses: ./.github/workflows/_docs.yml | |
| needs: build | |
| with: | |
| docker-image: ${{ needs.build.outputs.docker-image }} | |
| secrets: | |
| torchxla-bot-token: ${{ secrets.TORCH_XLA_BOT_TOKEN }} | |
| # New CI workflow | |
| build-torch-xla: | |
| name: "Build PyTorch/XLA (TPU)" | |
| uses: ./.github/workflows/_build_torch_xla.yml | |
| with: | |
| dev-image: us-central1-docker.pkg.dev/tpu-pytorch-releases/docker/development:3.10_tpuvm | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| build-cuda-plugin: | |
| name: "Build XLA CUDA plugin" | |
| uses: ./.github/workflows/_build_plugin.yml | |
| with: | |
| dev-image: us-central1-docker.pkg.dev/tpu-pytorch-releases/docker/development:3.10_cuda_12.1 | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| test-python-cpu: | |
| name: "CPU Python tests" | |
| uses: ./.github/workflows/_test_python.yml | |
| needs: build-torch-xla | |
| with: | |
| dev-image: us-central1-docker.pkg.dev/tpu-pytorch-releases/docker/development:3.10_tpuvm | |
| timeout-minutes: 120 | |
| collect-coverage: false | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| test-python-cuda: | |
| name: "GPU Python tests" | |
| uses: ./.github/workflows/_test_python.yml | |
| needs: [build-torch-xla, build-cuda-plugin] | |
| with: | |
| dev-image: us-central1-docker.pkg.dev/tpu-pytorch-releases/docker/development:3.10_cuda_12.1 | |
| runner: linux.8xlarge.nvidia.gpu | |
| timeout-minutes: 300 | |
| collect-coverage: false | |
| install-cuda-plugin: true | |
| secrets: | |
| gcloud-service-key: ${{ secrets.GCLOUD_SERVICE_KEY }} | |
| test-tpu: | |
| name: "TPU tests" | |
| uses: ./.github/workflows/_tpu_ci.yml | |
| needs: build-torch-xla | |
| # Only run this for HEAD and releases | |
| if: github.event_name == 'push' | |