feat: KEP for inject PET envs into init-container #3448
Workflow file for this run
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: E2E Tests | |
| on: | |
| pull_request: | |
| permissions: | |
| contents: read | |
| pull-requests: read | |
| jobs: | |
| cpu-e2e-test: | |
| name: CPU E2E Test | |
| runs-on: oracle-vm-16cpu-64gb-x86-64 | |
| env: | |
| GOPATH: ${{ github.workspace }}/go | |
| defaults: | |
| run: | |
| working-directory: ${{ env.GOPATH }}/src/github.com/kubeflow/trainer | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| # Kubernetes versions for e2e tests on Kind cluster. | |
| kubernetes-version: ["1.32.3", "1.33.1", "1.34.0", "1.35.0"] | |
| steps: | |
| - name: Checkout repository for local actions | |
| uses: actions/checkout@v6 | |
| - name: Setup CPU Cluster | |
| uses: ./.github/workflows/template-setup-clusters | |
| with: | |
| cluster_type: 'cpu' | |
| kubernetes_version: ${{ matrix.kubernetes-version }} | |
| - name: Run e2e with Go | |
| run: | | |
| make test-e2e || (kubectl logs -n kubeflow-system -l app.kubernetes.io/name=trainer && exit 1) | |
| - name: Run e2e test for example Notebooks | |
| run: | | |
| mkdir -p artifacts/notebooks | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/pytorch/image-classification/mnist.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_mnist.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/pytorch/question-answering/fine-tune-distilbert.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_fine-tune-distilbert.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/pytorch/speech-recognition/speech-recognition.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_speech-recognition.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/pytorch/audio-classification/audio-classification.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_audio-classification.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/local/local-training-mnist.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_local-training-mnist.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/local/local-container-mnist.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_local-container-mnist.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/xgboost/distributed-training/xgboost-distributed.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_xgboost-distributed.ipynb PAPERMILL_TIMEOUT=1800 | |
| # TODO (andreyvelich): Discuss how we can upload artifacts for multiple Notebooks. | |
| - name: Upload Artifacts to GitHub | |
| uses: actions/upload-artifact@v7 | |
| if: always() | |
| with: | |
| name: cpu-${{ matrix.kubernetes-version }} | |
| path: ${{ env.GOPATH }}/src/github.com/kubeflow/trainer/artifacts/* | |
| retention-days: 1 | |
| gpu-e2e-test: | |
| name: GPU E2E Test | |
| timeout-minutes: 120 | |
| runs-on: | |
| labels: oracle-vm-gpu-a10-2 | |
| group: GPUs | |
| env: | |
| GOPATH: ${{ github.workspace }}/go | |
| defaults: | |
| run: | |
| working-directory: ${{ env.GOPATH }}/src/github.com/kubeflow/trainer | |
| strategy: | |
| fail-fast: false | |
| matrix: | |
| kubernetes-version: ["1.33.1"] | |
| steps: | |
| - name: Checkout repository for local actions | |
| uses: actions/checkout@v6 | |
| - name: Setup GPU Cluster | |
| uses: ./.github/workflows/template-setup-clusters | |
| with: | |
| cluster_type: 'gpu' | |
| kubernetes_version: ${{ matrix.kubernetes-version }} | |
| - name: Run e2e test on GPU cluster | |
| run: | | |
| mkdir -p artifacts/notebooks | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/torchtune/qwen2_5/qwen2.5-1.5B-with-alpaca.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_qwen2_5_with_alpaca-trainjob-yaml.ipynb PAPERMILL_TIMEOUT=1800 | |
| make test-e2e-notebook NOTEBOOK_INPUT=./examples/jax/image-classification/mnist.ipynb NOTEBOOK_OUTPUT=./artifacts/notebooks/${{ matrix.kubernetes-version }}_jax_mnist.ipynb PAPERMILL_PARAMS="-p num_cpu 8 -p num_gpu 2 -p num_nodes 1" PAPERMILL_TIMEOUT=1800 | |
| - name: Upload Artifacts to GitHub | |
| if: always() | |
| uses: actions/upload-artifact@v7 | |
| with: | |
| name: gpu-${{ matrix.kubernetes-version }} | |
| path: ${{ env.GOPATH }}/src/github.com/kubeflow/trainer/artifacts/* | |
| retention-days: 1 |