Files
sglang/.github/workflows/pr-test-multimodal-gen.yml
T

396 lines
13 KiB
YAML

name: PR Test - Multimodal Gen
on:
workflow_call:
inputs:
multimodal_gen:
required: true
type: string
sgl_kernel:
required: true
type: string
b200_runner:
required: true
type: string
continue_on_error:
required: false
type: string
default: 'false'
pr_head_sha:
required: false
type: string
default: ''
git_ref:
required: false
type: string
default: ''
target_stage:
required: false
type: string
default: ''
test_parallel_dispatch:
required: false
type: string
default: 'false'
caller_needs_failure:
required: false
type: string
default: 'false'
skip_stage_health_check:
required: false
type: string
default: 'false'
# Workflow-level env is NOT inherited from the caller in reusable workflows.
# The github context (including github.event_name) IS inherited from the caller.
env:
SGLANG_IS_IN_CI: true
SGLANG_CUDA_COREDUMP: "1"
SGLANG_PR_TEST_BYPASS_MAINTENANCE_ON_MAIN: ${{ github.ref == 'refs/heads/main' && 'true' || 'false' }}
SKIP_STAGE_HEALTH_CHECK: ${{ inputs.skip_stage_health_check == 'true' }}
jobs:
compute-diffusion-partitions:
if: |
(inputs.target_stage == 'multimodal-gen-test-1-gpu') ||
(inputs.target_stage == 'multimodal-gen-test-2-gpu') ||
(
!inputs.target_stage &&
inputs.multimodal_gen == 'true'
)
runs-on: ubuntu-latest
outputs:
matrix-1gpu: ${{ steps.compute.outputs.matrix-1gpu }}
matrix-2gpu: ${{ steps.compute.outputs.matrix-2gpu }}
partition-count-1gpu: ${{ steps.compute.outputs['partition-count-1gpu'] }}
partition-count-2gpu: ${{ steps.compute.outputs['partition-count-2gpu'] }}
plan-1gpu: ${{ steps.compute.outputs.plan-1gpu }}
plan-2gpu: ${{ steps.compute.outputs.plan-2gpu }}
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: '3.10'
- name: Compute partitions
id: compute
run: |
python scripts/ci/utils/diffusion/compute_diffusion_partitions.py --min-time 1200 --target-time 1800 --max-time 2400 --max-partitions 10
multimodal-gen-test-1-gpu:
needs: compute-diffusion-partitions
if: |
always() &&
needs.compute-diffusion-partitions.result == 'success' &&
needs.compute-diffusion-partitions.outputs.matrix-1gpu != '{"include":[]}' &&
(
(inputs.target_stage == 'multimodal-gen-test-1-gpu') ||
(
!inputs.target_stage &&
((github.event_name == 'schedule' || inputs.test_parallel_dispatch == 'true') || (inputs.caller_needs_failure != 'true' && !cancelled())) &&
inputs.multimodal_gen == 'true'
)
)
runs-on: 1-gpu-h100
timeout-minutes: 240
strategy:
fail-fast: false
matrix: ${{ fromJson(needs.compute-diffusion-partitions.outputs.matrix-1gpu) }}
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- uses: ./.github/actions/check-stage-health
- uses: ./.github/actions/check-maintenance
- name: Download artifacts
if: inputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 20
run: |
CUSTOM_BUILD_SGL_KERNEL=${{inputs.sgl_kernel}} bash scripts/ci/cuda/ci_install_dependency.sh diffusion
- name: Run diffusion server tests
timeout-minutes: 240
env:
RUNAI_STREAMER_MEMORY_LIMIT: 0
CONTINUE_ON_ERROR_FLAG: ${{ inputs.continue_on_error == 'true' && '--continue-on-error' || '' }}
PARTITION_PLAN_JSON: ${{ needs.compute-diffusion-partitions.outputs.plan-1gpu }}
run: |
cd python
python3 sglang/multimodal_gen/test/run_suite.py \
--suite 1-gpu \
--partition-id ${{ matrix.part }} \
--total-partitions ${{ needs.compute-diffusion-partitions.outputs['partition-count-1gpu'] }} \
--partition-plan-json "$PARTITION_PLAN_JSON" \
$CONTINUE_ON_ERROR_FLAG
- name: Upload execution report
if: always()
uses: actions/upload-artifact@v4
with:
name: diffusion-report-1gpu-${{ matrix.part }}
path: python/sglang/multimodal_gen/test/execution_report_*.json
retention-days: 1
- uses: ./.github/actions/upload-cuda-coredumps
if: failure()
with:
artifact-suffix: ${{ matrix.part }}
multimodal-gen-test-2-gpu:
needs: compute-diffusion-partitions
if: |
always() &&
needs.compute-diffusion-partitions.result == 'success' &&
needs.compute-diffusion-partitions.outputs.matrix-2gpu != '{"include":[]}' &&
(
(inputs.target_stage == 'multimodal-gen-test-2-gpu') ||
(
!inputs.target_stage &&
((github.event_name == 'schedule' || inputs.test_parallel_dispatch == 'true') || (inputs.caller_needs_failure != 'true' && !cancelled())) &&
inputs.multimodal_gen == 'true'
)
)
runs-on: 2-gpu-h100
timeout-minutes: 240
strategy:
fail-fast: false
matrix: ${{ fromJson(needs.compute-diffusion-partitions.outputs.matrix-2gpu) }}
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- uses: ./.github/actions/check-stage-health
- uses: ./.github/actions/check-maintenance
- name: Download artifacts
if: inputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 20
run: |
CUSTOM_BUILD_SGL_KERNEL=${{inputs.sgl_kernel}} bash scripts/ci/cuda/ci_install_dependency.sh diffusion
- name: Run diffusion server tests
timeout-minutes: 240
env:
RUNAI_STREAMER_MEMORY_LIMIT: 0
CONTINUE_ON_ERROR_FLAG: ${{ inputs.continue_on_error == 'true' && '--continue-on-error' || '' }}
PARTITION_PLAN_JSON: ${{ needs.compute-diffusion-partitions.outputs.plan-2gpu }}
run: |
cd python
python3 sglang/multimodal_gen/test/run_suite.py \
--suite 2-gpu \
--partition-id ${{ matrix.part }} \
--total-partitions ${{ needs.compute-diffusion-partitions.outputs['partition-count-2gpu'] }} \
--partition-plan-json "$PARTITION_PLAN_JSON" \
$CONTINUE_ON_ERROR_FLAG
- name: Upload execution report
if: always()
uses: actions/upload-artifact@v4
with:
name: diffusion-report-2gpu-${{ matrix.part }}
path: python/sglang/multimodal_gen/test/execution_report_*.json
retention-days: 1
- uses: ./.github/actions/upload-cuda-coredumps
if: failure()
with:
artifact-suffix: ${{ matrix.part }}
multimodal-gen-component-accuracy:
if: |
(
inputs.target_stage == 'multimodal-gen-component-accuracy' ||
inputs.target_stage == 'multimodal-gen-component-accuracy-1-gpu' ||
inputs.target_stage == 'multimodal-gen-component-accuracy-2-gpu'
) ||
(
!inputs.target_stage &&
((github.event_name == 'schedule' || inputs.test_parallel_dispatch == 'true') || (inputs.caller_needs_failure != 'true' && !cancelled())) &&
inputs.multimodal_gen == 'true'
)
runs-on: 2-gpu-h100
timeout-minutes: 240
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- uses: ./.github/actions/check-stage-health
- uses: ./.github/actions/check-maintenance
- name: Download artifacts
if: inputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 20
run: |
CUSTOM_BUILD_SGL_KERNEL=${{inputs.sgl_kernel}} bash scripts/ci/cuda/ci_install_dependency.sh diffusion
- name: Run diffusion component accuracy tests
timeout-minutes: 240
env:
RUNAI_STREAMER_MEMORY_LIMIT: 0
CONTINUE_ON_ERROR_FLAG: ${{ inputs.continue_on_error == 'true' && '--continue-on-error' || '' }}
run: |
cd python
python3 sglang/multimodal_gen/test/run_suite.py \
--suite component-accuracy \
$CONTINUE_ON_ERROR_FLAG
- uses: ./.github/actions/upload-cuda-coredumps
if: always()
with:
artifact-suffix: component-accuracy
multimodal-gen-test-1-b200:
if: |
(inputs.target_stage == 'multimodal-gen-test-1-b200') ||
(
!inputs.target_stage &&
((github.event_name == 'schedule' || inputs.test_parallel_dispatch == 'true') || (inputs.caller_needs_failure != 'true' && !cancelled())) &&
inputs.multimodal_gen == 'true'
)
runs-on: ${{ inputs.b200_runner }}
timeout-minutes: 240
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- uses: ./.github/actions/check-stage-health
- uses: ./.github/actions/check-maintenance
- name: Download artifacts
if: inputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 20
run: |
CUSTOM_BUILD_SGL_KERNEL=${{inputs.sgl_kernel}} bash scripts/ci/cuda/ci_install_dependency.sh diffusion
- name: Run diffusion server tests
timeout-minutes: 240
env:
RUNAI_STREAMER_MEMORY_LIMIT: 0
CONTINUE_ON_ERROR_FLAG: ${{ inputs.continue_on_error == 'true' && '--continue-on-error' || '' }}
run: |
cd python
python3 sglang/multimodal_gen/test/run_suite.py \
--suite 1-gpu-b200 \
$CONTINUE_ON_ERROR_FLAG
- uses: ./.github/actions/upload-cuda-coredumps
if: failure()
multimodal-gen-unit-test:
if: |
(inputs.target_stage == 'multimodal-gen-unit-test') ||
(
!inputs.target_stage &&
((github.event_name == 'schedule' || inputs.test_parallel_dispatch == 'true') || (inputs.caller_needs_failure != 'true' && !cancelled())) &&
inputs.multimodal_gen == 'true'
)
runs-on: 1-gpu-h100
timeout-minutes: 120
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- uses: ./.github/actions/check-stage-health
- uses: ./.github/actions/check-maintenance
- name: Download artifacts
if: inputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 20
run: |
CUSTOM_BUILD_SGL_KERNEL=${{inputs.sgl_kernel}} bash scripts/ci/cuda/ci_install_dependency.sh diffusion
- name: Run diffusion unit tests
timeout-minutes: 60
run: |
cd python
python3 sglang/multimodal_gen/test/run_suite.py --suite unit
diffusion-coverage-check:
needs: [multimodal-gen-test-1-gpu, multimodal-gen-test-2-gpu]
if: |
always() &&
inputs.multimodal_gen == 'true' &&
(
needs.multimodal-gen-test-1-gpu.result == 'success' ||
needs.multimodal-gen-test-1-gpu.result == 'failure' ||
needs.multimodal-gen-test-2-gpu.result == 'success' ||
needs.multimodal-gen-test-2-gpu.result == 'failure'
)
runs-on: ubuntu-latest
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.git_ref || github.sha }}
- name: Set up Python
uses: actions/setup-python@v5
with:
python-version: '3.10'
- name: Download all execution reports
uses: actions/download-artifact@v4
with:
path: reports/
pattern: diffusion-report-*
merge-multiple: true
- name: Verify coverage
run: |
python scripts/ci/utils/diffusion/verify_diffusion_coverage.py --reports-dir reports/