Skip to content

Refactor Github Action per b/485167538 #3179

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
google-admin wants to merge 1 commit into from
+51 −25
Closed

Refactor Github Action per b/485167538 #3179

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 14 additions & 6 deletions .github/workflows/build_and_push_docker_image.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -66,13 +66,18 @@ jobs:
id: check
shell: bash
run: |
if [[ "${{ github.event_name }}" == "workflow_dispatch" && "${{ github.event.inputs.target_device }}" != "all" && "${{ github.event.inputs.target_device }}" != "${{ inputs.device }}" ]]; then
if [[ "${{ github.event_name }}" == "workflow_dispatch" && "${GITHUB_EVENT_INPUTS_TARGET_DEVICE}" != "all" && "${GITHUB_EVENT_INPUTS_TARGET_DEVICE}" != "${INPUTS_DEVICE}" ]]; then
echo "should_run=false" >> $GITHUB_OUTPUT
echo "Skipping ${{ inputs.image_name }} build for device: ${{ inputs.device }} in ${{ inputs.build_mode }} mode."
echo "Skipping ${INPUTS_IMAGE_NAME} build for device: ${INPUTS_DEVICE} in ${INPUTS_BUILD_MODE} mode."
else
echo "should_run=true" >> $GITHUB_OUTPUT
echo "Building ${{ inputs.image_name }} for device: ${{ inputs.device }} in ${{ inputs.build_mode }} mode."
echo "Building ${INPUTS_IMAGE_NAME} for device: ${INPUTS_DEVICE} in ${INPUTS_BUILD_MODE} mode."
fi
env:
GITHUB_EVENT_INPUTS_TARGET_DEVICE: ${{ github.event.inputs.target_device }}
INPUTS_DEVICE: ${{ inputs.device }}
INPUTS_IMAGE_NAME: ${{ inputs.image_name }}
INPUTS_BUILD_MODE: ${{ inputs.build_mode }}

- name: Checkout MaxText
uses: actions/checkout@v5
Expand Down Expand Up @@ -125,13 +130,13 @@ jobs:
if: steps.check.outputs.should_run == 'true'
shell: bash
run: |
SOURCE_IMAGE="gcr.io/tpu-prod-env-multipod/${{ inputs.image_name }}"
SOURCE_IMAGE="gcr.io/tpu-prod-env-multipod/${INPUTS_IMAGE_NAME}"

# Add date tag
gcloud container images add-tag "$SOURCE_IMAGE:latest" "$SOURCE_IMAGE:${{ inputs.image_date }}" --quiet
gcloud container images add-tag "$SOURCE_IMAGE:latest" "$SOURCE_IMAGE:${INPUTS_IMAGE_DATE}" --quiet

# Convert date to YYYYMMDD format
clean_date=$(echo "${{ inputs.image_date }}" | sed 's/[-:]//g' | cut -c1-8)
clean_date=$(echo "${INPUTS_IMAGE_DATE}" | sed 's/[-:]//g' | cut -c1-8)

# Add MaxText tag
maxtext_hash=$(git rev-parse --short HEAD)
Expand All @@ -146,3 +151,6 @@ jobs:
fi
done
fi
env:
INPUTS_IMAGE_NAME: ${{ inputs.image_name }}
INPUTS_IMAGE_DATE: ${{ inputs.image_date }}
42 changes: 28 additions & 14 deletions .github/workflows/build_and_test_maxtext.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -57,8 +57,8 @@ jobs:
exit 0
fi

git fetch origin ${{ github.base_ref }}
CHANGED_FILES=$(git diff --name-only origin/${{ github.base_ref }}...HEAD)
git fetch origin ${GITHUB_BASE_REF}
CHANGED_FILES=$(git diff --name-only origin/${GITHUB_BASE_REF}...HEAD)

echo "Changed files:"
echo "$CHANGED_FILES"
Expand Down Expand Up @@ -273,20 +273,20 @@ jobs:
- name: Check test results
run: |
# If doc-only, all tests should be skipped
if [ "${{ needs.doc_only_check.outputs.run_tests }}" == "false" ]; then
if [ "${NEEDS_DOC_ONLY_CHECK_OUTPUTS_RUN_TESTS}" == "false" ]; then
echo "Documentation-only changes detected, tests were skipped"
exit 0
fi

# Otherwise, check that build and all tests passed or were skipped
echo "Build result: ${{ needs.build_and_upload_maxtext_package.result }}"
echo "CPU tests: ${{ needs.maxtext_cpu_unit_tests.result }}"
echo "TPU tests: ${{ needs.maxtext_tpu_unit_tests.result }}"
echo "TPU integration: ${{ needs.maxtext_tpu_integration_tests.result }}"
echo "TPU pathways: ${{ needs.maxtext_tpu_pathways_unit_tests.result }}"
echo "TPU pathways integration: ${{ needs.maxtext_tpu_pathways_integration_tests.result }}"
echo "GPU tests: ${{ needs.maxtext_gpu_unit_tests.result }}"
echo "GPU integration: ${{ needs.maxtext_gpu_integration_tests.result }}"
echo "Build result: ${NEEDS_BUILD_AND_UPLOAD_MAXTEXT_PACKAGE_RESULT}"
echo "CPU tests: ${NEEDS_MAXTEXT_CPU_UNIT_TESTS_RESULT}"
echo "TPU tests: ${NEEDS_MAXTEXT_TPU_UNIT_TESTS_RESULT}"
echo "TPU integration: ${NEEDS_MAXTEXT_TPU_INTEGRATION_TESTS_RESULT}"
echo "TPU pathways: ${NEEDS_MAXTEXT_TPU_PATHWAYS_UNIT_TESTS_RESULT}"
echo "TPU pathways integration: ${NEEDS_MAXTEXT_TPU_PATHWAYS_INTEGRATION_TESTS_RESULT}"
echo "GPU tests: ${NEEDS_MAXTEXT_GPU_UNIT_TESTS_RESULT}"
echo "GPU integration: ${NEEDS_MAXTEXT_GPU_INTEGRATION_TESTS_RESULT}"

# Fail only if any job failed or was cancelled (skipped is OK)
if [ "${{ contains(needs.*.result, 'failure') }}" == "true" ] || [ "${{ contains(needs.*.result, 'cancelled') }}" == "true" ]; then
Expand All @@ -295,6 +295,16 @@ jobs:
fi

echo "All required tests passed successfully"
env:
NEEDS_DOC_ONLY_CHECK_OUTPUTS_RUN_TESTS: ${{ needs.doc_only_check.outputs.run_tests }}
NEEDS_BUILD_AND_UPLOAD_MAXTEXT_PACKAGE_RESULT: ${{ needs.build_and_upload_maxtext_package.result }}
NEEDS_MAXTEXT_CPU_UNIT_TESTS_RESULT: ${{ needs.maxtext_cpu_unit_tests.result }}
NEEDS_MAXTEXT_TPU_UNIT_TESTS_RESULT: ${{ needs.maxtext_tpu_unit_tests.result }}
NEEDS_MAXTEXT_TPU_INTEGRATION_TESTS_RESULT: ${{ needs.maxtext_tpu_integration_tests.result }}
NEEDS_MAXTEXT_TPU_PATHWAYS_UNIT_TESTS_RESULT: ${{ needs.maxtext_tpu_pathways_unit_tests.result }}
NEEDS_MAXTEXT_TPU_PATHWAYS_INTEGRATION_TESTS_RESULT: ${{ needs.maxtext_tpu_pathways_integration_tests.result }}
NEEDS_MAXTEXT_GPU_UNIT_TESTS_RESULT: ${{ needs.maxtext_gpu_unit_tests.result }}
NEEDS_MAXTEXT_GPU_INTEGRATION_TESTS_RESULT: ${{ needs.maxtext_gpu_integration_tests.result }}

all_notebooks_passed:
name: All Notebooks Passed
Expand All @@ -304,14 +314,14 @@ jobs:
steps:
- name: Check notebooks results
run: |
if [ "${{ needs.doc_only_check.outputs.run_notebooks }}" == "false" ]; then
if [ "${NEEDS_DOC_ONLY_CHECK_OUTPUTS_RUN_NOTEBOOKS}" == "false" ]; then
echo "Non-notebook changes detected, runs were skipped"
exit 0
fi

# Otherwise, check that build and notebooks run passed or were skipped
echo "Build result: ${{ needs.build_and_upload_maxtext_package.result }}"
echo "Jupyter Notebooks result: ${{ needs.maxtext_jupyter_notebooks.result }}"
echo "Build result: ${NEEDS_BUILD_AND_UPLOAD_MAXTEXT_PACKAGE_RESULT}"
echo "Jupyter Notebooks result: ${NEEDS_MAXTEXT_JUPYTER_NOTEBOOKS_RESULT}"

# Fail only if any job failed or was cancelled (skipped is OK)
if [ "${{ contains(needs.*.result, 'failure') }}" == "true" ] || [ "${{ contains(needs.*.result, 'cancelled') }}" == "true" ]; then
Expand All @@ -320,6 +330,10 @@ jobs:
fi

echo "All required notebooks passed successfully"
env:
NEEDS_DOC_ONLY_CHECK_OUTPUTS_RUN_NOTEBOOKS: ${{ needs.doc_only_check.outputs.run_notebooks }}
NEEDS_BUILD_AND_UPLOAD_MAXTEXT_PACKAGE_RESULT: ${{ needs.build_and_upload_maxtext_package.result }}
NEEDS_MAXTEXT_JUPYTER_NOTEBOOKS_RESULT: ${{ needs.maxtext_jupyter_notebooks.result }}

notify_failure:
name: Notify failed build # creates an issue or modifies last open existing issue for failed build
Expand Down
14 changes: 9 additions & 5 deletions .github/workflows/run_tests_against_package.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -101,18 +101,18 @@ jobs:
- name: Run Tests
shell: bash
run: |
if [ "${{ inputs.is_scheduled_run }}" == "true" ]; then
FINAL_PYTEST_MARKER="${{ inputs.pytest_marker }}"
if [ "${INPUTS_IS_SCHEDULED_RUN}" == "true" ]; then
FINAL_PYTEST_MARKER="${INPUTS_PYTEST_MARKER}"
else
FINAL_PYTEST_MARKER="${{ inputs.pytest_marker }} and not scheduled_only"
FINAL_PYTEST_MARKER="${INPUTS_PYTEST_MARKER} and not scheduled_only"
fi
# TODO: Use package data for testing and remove the env vars
export MAXTEXT_REPO_ROOT=$(pwd)
export MAXTEXT_ASSETS_ROOT=$(pwd)/src/maxtext/assets
export MAXTEXT_TEST_ASSETS_ROOT=$(pwd)/tests/assets
export MAXTEXT_PKG_DIR=$(pwd)/src/maxtext
# omit this libtpu init args for gpu tests
if [ "${{ inputs.device_type }}" != "cuda12" ]; then
if [ "${INPUTS_DEVICE_TYPE}" != "cuda12" ]; then
export LIBTPU_INIT_ARGS='--xla_tpu_scoped_vmem_limit_kib=65536'
fi
if [ "${{ inputs.total_workers }}" -gt 1 ]; then
Expand All @@ -122,7 +122,7 @@ jobs:
SPLIT_ARGS=""
fi
# TODO: Fix the skipped tests and remove the deselect flags
.venv/bin/python3 -m pytest ${{ inputs.pytest_addopts }} \
.venv/bin/python3 -m pytest ${INPUTS_PYTEST_ADDOPTS} \
-v \
-m "${FINAL_PYTEST_MARKER}" \
--durations=0 \
Expand All @@ -134,6 +134,10 @@ jobs:
$SPLIT_ARGS
env:
PYTHONPATH: "${{ github.workspace }}/src"
INPUTS_IS_SCHEDULED_RUN: ${{ inputs.is_scheduled_run }}
INPUTS_PYTEST_MARKER: ${{ inputs.pytest_marker }}
INPUTS_DEVICE_TYPE: ${{ inputs.device_type }}
INPUTS_PYTEST_ADDOPTS: ${{ inputs.pytest_addopts }}
- name: Upload results to Codecov
uses: codecov/codecov-action@v5
continue-on-error: true
Expand Down
Loading