minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

Disable unit-test-deepep-8-gpu (#17176)

Browse Source
This commit is contained in:
Alison Shao
2026-01-15 15:12:45 -08:00
committed by GitHub
parent 8b99af9af8
commit 69822c7271
2 changed files with 55 additions and 51 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/pr-test.yml
+49 -48
View File
@@ -1614,53 +1614,54 @@ jobs:
fi fi
python3 run_suite.py --suite per-commit-4-gpu-deepep $RETRY_FLAG $CONTINUE_ON_ERROR_FLAG python3 run_suite.py --suite per-commit-4-gpu-deepep $RETRY_FLAG $CONTINUE_ON_ERROR_FLAG
unit-test-deepep-8-gpu: # Disabled: IBGDA/cudaHostRegister environment issues on 8-GPU runner, see #17175
needs: [check-changes, call-gate, stage-b-test-small-1-gpu, stage-b-test-large-1-gpu, stage-b-test-large-2-gpu, stage-b-test-4-gpu-b200] # unit-test-deepep-8-gpu:
if: | # needs: [check-changes, call-gate, stage-b-test-small-1-gpu, stage-b-test-large-1-gpu, stage-b-test-large-2-gpu, stage-b-test-4-gpu-b200]
always() && # if: |
( # always() &&
(inputs.target_stage == 'unit-test-deepep-8-gpu') || # (
( # (inputs.target_stage == 'unit-test-deepep-8-gpu') ||
!inputs.target_stage && # (
(github.event_name == 'schedule' || (!failure() && !cancelled())) && # !inputs.target_stage &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true')) # (github.event_name == 'schedule' || (!failure() && !cancelled())) &&
) # ((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
) # )
runs-on: 8-gpu-h200 # )
env: # runs-on: 8-gpu-h200
RUNNER_LABELS: 8-gpu-h200 # env:
steps: # RUNNER_LABELS: 8-gpu-h200
- name: Checkout code # steps:
uses: actions/checkout@v4 # - name: Checkout code
with: # uses: actions/checkout@v4
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }} # with:
# ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts #
if: needs.check-changes.outputs.sgl_kernel == 'true' # - name: Download artifacts
uses: actions/download-artifact@v4 # if: needs.check-changes.outputs.sgl_kernel == 'true'
with: # uses: actions/download-artifact@v4
path: sgl-kernel/dist/ # with:
merge-multiple: true # path: sgl-kernel/dist/
pattern: wheel-python3.10-cuda12.9 # merge-multiple: true
# pattern: wheel-python3.10-cuda12.9
- name: Install dependencies #
timeout-minutes: 10 # - name: Install dependencies
run: | # timeout-minutes: 10
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_deepep.sh # run: |
# CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_deepep.sh
- name: Run test #
timeout-minutes: 20 # - name: Run test
run: | # timeout-minutes: 20
cd test/srt # run: |
RETRY_FLAG="" # cd test/srt
if [[ "${{ needs.check-changes.outputs.enable_retry }}" == "true" ]]; then # RETRY_FLAG=""
RETRY_FLAG="--enable-retry" # if [[ "${{ needs.check-changes.outputs.enable_retry }}" == "true" ]]; then
fi # RETRY_FLAG="--enable-retry"
CONTINUE_ON_ERROR_FLAG="" # fi
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then # CONTINUE_ON_ERROR_FLAG=""
CONTINUE_ON_ERROR_FLAG="--continue-on-error" # if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
fi # CONTINUE_ON_ERROR_FLAG="--continue-on-error"
python3 run_suite.py --suite per-commit-8-gpu-h200-deepep $RETRY_FLAG $CONTINUE_ON_ERROR_FLAG # fi
# python3 run_suite.py --suite per-commit-8-gpu-h200-deepep $RETRY_FLAG $CONTINUE_ON_ERROR_FLAG
unit-test-backend-4-gpu-b200: unit-test-backend-4-gpu-b200:
needs: [check-changes, call-gate, stage-b-test-small-1-gpu, stage-b-test-large-1-gpu, stage-b-test-large-2-gpu, stage-b-test-4-gpu-b200] needs: [check-changes, call-gate, stage-b-test-small-1-gpu, stage-b-test-large-1-gpu, stage-b-test-large-2-gpu, stage-b-test-4-gpu-b200]
@@ -1800,7 +1801,7 @@ jobs:
accuracy-test-1-gpu, accuracy-test-1-gpu,
accuracy-test-2-gpu, accuracy-test-2-gpu,
unit-test-deepep-4-gpu, unit-test-deepep-4-gpu,
unit-test-deepep-8-gpu, # unit-test-deepep-8-gpu, # Disabled, see #17175
unit-test-backend-4-gpu-b200, unit-test-backend-4-gpu-b200,
unit-test-backend-4-gpu-gb200, unit-test-backend-4-gpu-gb200,
] ]
test/srt/run_suite.py
+6 -3
View File
@@ -48,9 +48,11 @@ suites = {
TestFile("ep/test_deepep_small.py", 531), TestFile("ep/test_deepep_small.py", 531),
TestFile("ep/test_mooncake_ep_small.py", 660), TestFile("ep/test_mooncake_ep_small.py", 660),
], ],
"per-commit-8-gpu-h200-deepep": [ # Disabled: IBGDA/cudaHostRegister environment issues on 8-GPU runner, see #17175
TestFile("ep/test_deepep_large.py", 563), # 4-GPU DeepEP tests provide sufficient coverage
], # "per-commit-8-gpu-h200-deepep": [
# TestFile("ep/test_deepep_large.py", 563),
# ],
"quantization_test": [ "quantization_test": [
TestFile("quant/test_awq.py", 163), TestFile("quant/test_awq.py", 163),
TestFile("quant/test_marlin_moe.py", 200), TestFile("quant/test_marlin_moe.py", 200),
@@ -74,6 +76,7 @@ suites = {
TestFile( TestFile(
"models/test_qwen3_next_models_pcg.py" "models/test_qwen3_next_models_pcg.py"
), # Disabled: intermittent failures, see #17039 ), # Disabled: intermittent failures, see #17039
TestFile("ep/test_deepep_large.py", 563), # Disabled: see #17175
], ],
} }