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

[CI] Rename basic CI stage-a/b/c -> base-a/b/c for symmetry with extra CI (#25420)

Browse Source
This commit is contained in:
Liangsheng Yin
2026-05-15 18:26:55 -07:00
committed by GitHub
parent d0c38329b2
commit b7d62bd724
473 changed files with 745 additions and 746 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/pr-test.yml
+79 -79
View File
@@ -1,4 +1,4 @@
name: PR Test
name: PR Test Base
on:
schedule:
@@ -34,8 +34,8 @@ on:
required: false
type: boolean
default: false
skip_stage_health_check:
description: "Skip stage health check fast-fail (e.g. for release branch cuts)"
skip_pr_test_health_check:
description: "Skip PR test health check fast-fail (e.g. for release branch cuts)"
required: false
type: boolean
default: false
@@ -52,7 +52,7 @@ env:
SGLANG_IS_IN_CI: true
SGLANG_CUDA_COREDUMP: "1"
SGLANG_JIT_DEEPGEMM_FAST_WARMUP: true
SKIP_STAGE_HEALTH_CHECK: ${{ (inputs.skip_stage_health_check == true || inputs.test_parallel_dispatch == true || inputs.run_all_tests == true) && 'true' || 'false' }}
SKIP_PR_TEST_HEALTH_CHECK: ${{ (inputs.skip_pr_test_health_check == true || inputs.test_parallel_dispatch == true || inputs.run_all_tests == true) && 'true' || 'false' }}
# TEMP: rebuild deepep against the new torch for torch-211-merge PR only — revert before merging to main.
FORCE_REBUILD_DEEPEP: '1'
# Schedule / main-branch dispatch / workflow_call from main use refs/heads/main; PR events use refs/pull/*/merge
@@ -89,7 +89,7 @@ jobs:
# For PRs with the `bypass-fastfail` label: wait jobs run but return success immediately
# (handled inside the wait-for-jobs action), so downstream stages dispatch in parallel.
wait-for-stage-a:
wait-for-base-a:
needs: [check-changes, call-gate]
if: |
always() &&
@@ -100,7 +100,7 @@ jobs:
(needs.call-gate.result == 'success' || needs.call-gate.result == 'skipped')
runs-on: ubuntu-latest
outputs:
stage_a_result: ${{ steps.wait.outputs.result }}
base_a_result: ${{ steps.wait.outputs.result }}
steps:
- uses: actions/checkout@v4
@@ -109,27 +109,27 @@ jobs:
- uses: ./.github/actions/wait-for-jobs
id: wait
with:
stage-name: stage-a
stage-name: base-a
jobs: |
[
{"prefix": "stage-a-test-1-gpu-small", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['stage-a-test-1-gpu-small'].size }}},
{"prefix": "stage-a-test-cpu", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['stage-a-test-cpu'].size }}}
{"prefix": "base-a-test-1-gpu-small", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['base-a-test-1-gpu-small'].size }}},
{"prefix": "base-a-test-cpu", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['base-a-test-cpu'].size }}}
]
max-wait-minutes: '240'
wait-for-stage-b:
needs: [check-changes, call-gate, wait-for-stage-a]
wait-for-base-b:
needs: [check-changes, call-gate, wait-for-base-a]
if: |
always() &&
!cancelled() &&
github.event_name == 'pull_request' &&
inputs.test_parallel_dispatch != true &&
(needs.check-changes.outputs.main_package == 'true' || needs.check-changes.outputs.sgl_kernel == 'true') &&
(needs.wait-for-stage-a.result == 'success' || needs.wait-for-stage-a.result == 'skipped') &&
(needs.wait-for-base-a.result == 'success' || needs.wait-for-base-a.result == 'skipped') &&
(needs.call-gate.result == 'success' || needs.call-gate.result == 'skipped')
runs-on: ubuntu-latest
outputs:
stage_b_result: ${{ steps.wait.outputs.result }}
base_b_result: ${{ steps.wait.outputs.result }}
steps:
- uses: actions/checkout@v4
@@ -138,13 +138,13 @@ jobs:
- uses: ./.github/actions/wait-for-jobs
id: wait
with:
stage-name: stage-b
stage-name: base-b
jobs: |
[
{"prefix": "stage-b-test-1-gpu-small", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['stage-b-test-1-gpu-small'].size }}},
{"prefix": "stage-b-test-1-gpu-large", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['stage-b-test-1-gpu-large'].size }}},
{"prefix": "stage-b-test-2-gpu-large", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['stage-b-test-2-gpu-large'].size }}},
{"prefix": "stage-b-test-4-gpu-b200", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['stage-b-test-4-gpu-b200'].size }}}
{"prefix": "base-b-test-1-gpu-small", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['base-b-test-1-gpu-small'].size }}},
{"prefix": "base-b-test-1-gpu-large", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['base-b-test-1-gpu-large'].size }}},
{"prefix": "base-b-test-2-gpu-large", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['base-b-test-2-gpu-large'].size }}},
{"prefix": "base-b-test-4-gpu-b200", "expected_count": ${{ fromJson(needs.check-changes.outputs.partitions)['base-b-test-4-gpu-b200'].size }}}
]
max-wait-minutes: '480'
@@ -180,7 +180,7 @@ jobs:
runs_on: x64-kernel-build-node
job_display_name: Build Wheel
git_ref: ${{ inputs.git_ref || '' }}
skip_stage_health_check: ${{ inputs.skip_stage_health_check == true }}
skip_pr_test_health_check: ${{ inputs.skip_pr_test_health_check == true }}
secrets: inherit
sgl-kernel-build-wheels-arm:
@@ -198,7 +198,7 @@ jobs:
job_display_name: Build Wheel Arm
arch_suffix: '-aarch64'
git_ref: ${{ inputs.git_ref || '' }}
skip_stage_health_check: ${{ inputs.skip_stage_health_check == true }}
skip_pr_test_health_check: ${{ inputs.skip_pr_test_health_check == true }}
secrets: inherit
call-sgl-kernel-tests:
@@ -213,7 +213,7 @@ jobs:
runs_on_map: ${{ needs.check-changes.outputs.runs_on_map }}
sgl_kernel: ${{ needs.check-changes.outputs.sgl_kernel }}
git_ref: ${{ inputs.git_ref || '' }}
skip_stage_health_check: ${{ inputs.skip_stage_health_check == true }}
skip_pr_test_health_check: ${{ inputs.skip_pr_test_health_check == true }}
secrets: inherit
# =============================================== jit-kernel ====================================================
@@ -234,18 +234,18 @@ jobs:
sgl_kernel: ${{ needs.check-changes.outputs.sgl_kernel }}
git_ref: ${{ inputs.git_ref || '' }}
test_parallel_dispatch: ${{ inputs.test_parallel_dispatch == true && 'true' || 'false' }}
skip_stage_health_check: ${{ inputs.skip_stage_health_check == true }}
skip_pr_test_health_check: ${{ inputs.skip_pr_test_health_check == true }}
secrets: inherit
# =============================================== primary ====================================================
# Runs on 5090 (32GB, SM120)
stage-a-test-1-gpu-small:
base-a-test-1-gpu-small:
needs: [check-changes, call-gate, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-a-test-1-gpu-small
self_name: base-a-test-1-gpu-small
runner_config: 1-gpu-small
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -253,7 +253,7 @@ jobs:
run_timeout_minutes: '10'
secrets: inherit
stage-a-test-cpu:
base-a-test-cpu:
needs: [check-changes, call-gate]
if: |
always() &&
@@ -263,9 +263,9 @@ jobs:
timeout-minutes: 240
strategy:
fail-fast: false
max-parallel: ${{ fromJson(needs.check-changes.outputs.partitions)['stage-a-test-cpu'].max_parallel }}
max-parallel: ${{ fromJson(needs.check-changes.outputs.partitions)['base-a-test-cpu'].max_parallel }}
matrix:
partition: ${{ fromJson(needs.check-changes.outputs.partitions)['stage-a-test-cpu'].arr }}
partition: ${{ fromJson(needs.check-changes.outputs.partitions)['base-a-test-cpu'].arr }}
steps:
- name: Free disk space
run: |
@@ -277,7 +277,7 @@ jobs:
with:
ref: ${{ inputs.git_ref || github.sha }}
- uses: ./.github/actions/check-stage-health
- uses: ./.github/actions/check-pr-test-health
- uses: ./.github/actions/check-maintenance
@@ -316,15 +316,15 @@ jobs:
CONTINUE_ON_ERROR_FLAG: ${{ needs.check-changes.outputs.continue_on_error == 'true' && '--continue-on-error' || '' }}
run: |
cd test/
python3 run_suite.py --hw cpu --suite stage-a-test-cpu --auto-partition-id ${{ matrix.partition }} --auto-partition-size ${{ fromJson(needs.check-changes.outputs.partitions)['stage-a-test-cpu'].size }} $CONTINUE_ON_ERROR_FLAG
python3 run_suite.py --hw cpu --suite base-a-test-cpu --auto-partition-id ${{ matrix.partition }} --auto-partition-size ${{ fromJson(needs.check-changes.outputs.partitions)['base-a-test-cpu'].size }} $CONTINUE_ON_ERROR_FLAG
# Runs on 5090 (32GB, SM120)
stage-b-test-1-gpu-small:
needs: [check-changes, call-gate, wait-for-stage-a, sgl-kernel-build-wheels]
base-b-test-1-gpu-small:
needs: [check-changes, call-gate, wait-for-base-a, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-b-test-1-gpu-small
self_name: base-b-test-1-gpu-small
runner_config: 1-gpu-small
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -333,12 +333,12 @@ jobs:
secrets: inherit
# Runs on H100 (80GB, SM90) - tests that don't pass on 5090 (FA3, FP8, high VRAM, etc.)
stage-b-test-1-gpu-large:
needs: [check-changes, call-gate, wait-for-stage-a, sgl-kernel-build-wheels]
base-b-test-1-gpu-large:
needs: [check-changes, call-gate, wait-for-base-a, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-b-test-1-gpu-large
self_name: base-b-test-1-gpu-large
runner_config: 1-gpu-large
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -347,12 +347,12 @@ jobs:
timeout_per_file: '1800'
secrets: inherit
stage-b-test-2-gpu-large:
needs: [check-changes, call-gate, wait-for-stage-a, sgl-kernel-build-wheels]
base-b-test-2-gpu-large:
needs: [check-changes, call-gate, wait-for-base-a, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-b-test-2-gpu-large
self_name: base-b-test-2-gpu-large
runner_config: 2-gpu-large
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -360,12 +360,12 @@ jobs:
run_timeout_minutes: '30'
secrets: inherit
stage-b-test-4-gpu-b200:
needs: [check-changes, call-gate, wait-for-stage-a, sgl-kernel-build-wheels]
base-b-test-4-gpu-b200:
needs: [check-changes, call-gate, wait-for-base-a, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-b-test-4-gpu-b200
self_name: base-b-test-4-gpu-b200
runner_config: 4-gpu-b200
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -391,15 +391,15 @@ jobs:
git_ref: ${{ inputs.git_ref || '' }}
test_parallel_dispatch: ${{ inputs.test_parallel_dispatch == true && 'true' || 'false' }}
caller_needs_failure: ${{ (needs.call-gate.result == 'failure' || needs.sgl-kernel-build-wheels.result == 'failure' || needs.check-changes.result == 'failure') && 'true' || 'false' }}
skip_stage_health_check: ${{ inputs.skip_stage_health_check == true && 'true' || 'false' }}
skip_pr_test_health_check: ${{ inputs.skip_pr_test_health_check == true && 'true' || 'false' }}
secrets: inherit
stage-c-test-4-gpu-h100:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-4-gpu-h100:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-4-gpu-h100
self_name: base-c-test-4-gpu-h100
runner_config: 4-gpu-h100
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -407,12 +407,12 @@ jobs:
run_timeout_minutes: '30'
secrets: inherit
stage-c-test-8-gpu-h200:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-8-gpu-h200:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-8-gpu-h200
self_name: base-c-test-8-gpu-h200
runner_config: 8-gpu-h200
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -427,12 +427,12 @@ jobs:
warmup_timeout_minutes: '60'
secrets: inherit
stage-c-test-8-gpu-h20:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-8-gpu-h20:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-8-gpu-h20
self_name: base-c-test-8-gpu-h20
runner_config: 8-gpu-h20
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -440,12 +440,12 @@ jobs:
run_timeout_minutes: '30'
secrets: inherit
stage-c-test-deepep-4-gpu-h100:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-deepep-4-gpu-h100:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-deepep-4-gpu-h100
self_name: base-c-test-deepep-4-gpu-h100
runner_config: deepep-4-gpu-h100
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -455,12 +455,12 @@ jobs:
warmup_server_models: 'lmsys/sglang-ci-dsv3-test:4'
secrets: inherit
stage-c-test-4-gpu-b200:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-4-gpu-b200:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-4-gpu-b200
self_name: base-c-test-4-gpu-b200
runner_config: 4-gpu-b200
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -469,12 +469,12 @@ jobs:
timeout_per_file: '1800'
secrets: inherit
stage-c-test-dsv4-4-gpu-b200:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-dsv4-4-gpu-b200:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-dsv4-4-gpu-b200
self_name: base-c-test-dsv4-4-gpu-b200
runner_config: dsv4-4-gpu-b200
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -483,12 +483,12 @@ jobs:
timeout_per_file: '1800'
secrets: inherit
stage-c-test-dsv4-8-gpu-h200:
needs: [check-changes, call-gate, wait-for-stage-b, sgl-kernel-build-wheels]
base-c-test-dsv4-8-gpu-h200:
needs: [check-changes, call-gate, wait-for-base-b, sgl-kernel-build-wheels]
if: ${{ !failure() && !cancelled() }}
uses: ./.github/workflows/_pr-test-stage.yml
with:
self_name: stage-c-test-dsv4-8-gpu-h200
self_name: base-c-test-dsv4-8-gpu-h200
runner_config: dsv4-8-gpu-h200
check_changes: ${{ toJson(needs.check-changes.outputs) }}
caller_inputs: ${{ toJson(inputs) }}
@@ -507,27 +507,27 @@ jobs:
sgl-kernel-build-wheels-arm,
call-sgl-kernel-tests,
wait-for-stage-a,
wait-for-stage-b,
wait-for-base-a,
wait-for-base-b,
call-jit-kernel-tests,
call-multimodal-gen-tests,
stage-a-test-1-gpu-small,
stage-a-test-cpu,
stage-b-test-1-gpu-small,
stage-b-test-1-gpu-large,
stage-b-test-2-gpu-large,
stage-b-test-4-gpu-b200,
stage-c-test-4-gpu-h100,
stage-c-test-8-gpu-h20,
stage-c-test-8-gpu-h200,
stage-c-test-deepep-4-gpu-h100,
stage-c-test-4-gpu-b200,
stage-c-test-dsv4-4-gpu-b200,
stage-c-test-dsv4-8-gpu-h200,
# stage-c-test-4-gpu-gb200, # Temporarily disabled — no GB200 runner
base-a-test-1-gpu-small,
base-a-test-cpu,
base-b-test-1-gpu-small,
base-b-test-1-gpu-large,
base-b-test-2-gpu-large,
base-b-test-4-gpu-b200,
base-c-test-4-gpu-h100,
base-c-test-8-gpu-h20,
base-c-test-8-gpu-h200,
base-c-test-deepep-4-gpu-h100,
base-c-test-4-gpu-b200,
base-c-test-dsv4-4-gpu-b200,
base-c-test-dsv4-8-gpu-h200,
# base-c-test-4-gpu-gb200, # Temporarily disabled — no GB200 runner
]
if: always()
runs-on: ubuntu-latest