chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 11 Packages Projects Releases Wiki Activity

Migrate performance, accuracy, and quantization tests to CI registry (#17177)

Browse Source 
Co-authored-by: Kangyan-Zhou <zky314343421@gmail.com>
This commit is contained in:
Alison Shao
2026-01-18 23:25:24 -08:00
committed by GitHub
parent a3d9a21882
commit 8916b9d080
20 changed files with 1050 additions and 1020 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
.github/workflows/pr-test-amd.yml vendored
View File

@@ -713,23 +713,23 @@ jobs:
- name: Benchmark single latency
timeout-minutes: 20
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_bs1_small
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_bs1_default
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_one_batch_1gpu.TestBenchOneBatch1GPU.test_bs1_small
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_one_batch_1gpu.TestBenchOneBatch1GPU.test_bs1_default
- name: Benchmark online latency
timeout-minutes: 15
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_online_latency_default
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_1gpu_part1.TestBenchServing1GPUPart1.test_online_latency_default
- name: Benchmark offline throughput
timeout-minutes: 15
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_default
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_1gpu_part1.TestBenchServing1GPUPart1.test_offline_throughput_default
- name: Benchmark offline throughput (Non-streaming, small batch size)
timeout-minutes: 15
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_non_stream_small_batch_size
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_1gpu_part1.TestBenchServing1GPUPart1.test_offline_throughput_non_stream_small_batch_size
performance-test-1-gpu-part-2-amd:
needs: [check-changes, stage-a-test-1-amd]
@@ -768,17 +768,17 @@ jobs:
- name: Benchmark offline throughput (w/o RadixAttention)
timeout-minutes: 15
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_without_radix_cache
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_1gpu_part1.TestBenchServing1GPUPart1.test_offline_throughput_without_radix_cache
- name: Benchmark offline throughput (w/ Triton)
timeout-minutes: 15
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_with_triton_attention_backend
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_1gpu_part1.TestBenchServing1GPUPart1.test_offline_throughput_with_triton_attention_backend
- name: Benchmark offline throughput (w/ FP8)
timeout-minutes: 15
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_default_fp8
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_1gpu_large.TestBenchServing1GPULarge.test_offline_throughput_default_fp8
performance-test-2-gpu-amd:
needs: [check-changes, stage-a-test-1-amd]
@@ -822,32 +822,32 @@ jobs:
- name: Benchmark single latency (TP=2)
timeout-minutes: 25
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_moe_tp2_bs1
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_one_batch_2gpu.TestBenchOneBatch2GPU.test_moe_tp2_bs1
- name: Benchmark single latency + torch.compile (TP=2)
timeout-minutes: 25
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_torch_compile_tp2_bs1
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_one_batch_2gpu.TestBenchOneBatch2GPU.test_torch_compile_tp2_bs1
- name: Benchmark offline throughput (TP=2)
timeout-minutes: 25
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_moe_offline_throughput_default
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_2gpu.TestBenchServing2GPU.test_moe_offline_throughput_default
- name: Benchmark offline throughput (w/o RadixAttention) (TP=2)
timeout-minutes: 25
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_moe_offline_throughput_without_radix_cache
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_2gpu.TestBenchServing2GPU.test_moe_offline_throughput_without_radix_cache
- name: Benchmark offline PP decode throughput (PP=2)
timeout-minutes: 10
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_pp_offline_throughput_default_decode
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_2gpu.TestBenchServing2GPU.test_pp_offline_throughput_default_decode
- name: Benchmark offline PP prefill throughput (PP=2)
timeout-minutes: 10
run: |
bash scripts/ci/amd_ci_exec.sh python3 -m unittest test_bench_serving.TestBenchServing.test_pp_long_context_prefill
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/perf python3 -m unittest test_bench_serving_2gpu.TestBenchServing2GPU.test_pp_long_context_prefill
accuracy-test-1-gpu-amd:
needs: [check-changes, stage-a-test-1-amd]
@@ -886,7 +886,7 @@ jobs:
- name: Evaluate Accuracy
timeout-minutes: 30
run: |
bash scripts/ci/amd_ci_exec.sh -e SGLANG_USE_AITER=0 python3 test_eval_accuracy_large.py
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/eval -e SGLANG_USE_AITER=0 python3 test_eval_accuracy_large.py
accuracy-test-2-gpu-amd:
needs: [check-changes, accuracy-test-1-gpu-amd]
@@ -926,7 +926,7 @@ jobs:
- name: Evaluate accuracy (TP=2)
timeout-minutes: 30
run: |
bash scripts/ci/amd_ci_exec.sh -e SGLANG_USE_AITER_AR=0 -e SGLANG_USE_AITER=0 -e HF_HUB_ENABLE_HF_TRANSFER=0 python3 test_moe_eval_accuracy_large.py
bash scripts/ci/amd_ci_exec.sh -w /sglang-checkout/test/registered/eval -e SGLANG_USE_AITER_AR=0 -e SGLANG_USE_AITER=0 -e HF_HUB_ENABLE_HF_TRANSFER=0 python3 test_moe_eval_accuracy_large.py
pr-test-amd-finish:
needs:

672
.github/workflows/pr-test.yml vendored
View File

@@ -186,6 +186,7 @@ jobs:
needs: [check-changes, call-gate]
if: needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: x64-kernel-build-node
timeout-minutes: 60
strategy:
matrix:
include:
@@ -233,6 +234,7 @@ jobs:
needs: [check-changes, call-gate]
if: needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: arm-kernel-build-node
timeout-minutes: 60
strategy:
matrix:
include:
@@ -283,6 +285,7 @@ jobs:
!inputs.target_stage &&
needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-runner
steps:
@@ -319,6 +322,7 @@ jobs:
!inputs.target_stage &&
needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-runner
steps:
@@ -355,6 +359,7 @@ jobs:
!inputs.target_stage &&
needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
CI: true
RUNNER_LABELS: 1-gpu-runner
@@ -404,6 +409,7 @@ jobs:
!inputs.target_stage &&
needs.check-changes.outputs.sgl_kernel == 'true'
runs-on: ${{ needs.check-changes.outputs.b200_runner }}
timeout-minutes: 60
env:
RUNNER_LABELS: ${{ needs.check-changes.outputs.b200_runner }}
steps:
@@ -473,6 +479,7 @@ jobs:
!inputs.target_stage &&
needs.check-changes.outputs.jit_kernel == 'true'
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-runner
steps:
@@ -506,6 +513,7 @@ jobs:
)
)
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-runner
steps:
@@ -552,6 +560,7 @@ jobs:
)
)
runs-on: ubuntu-latest
timeout-minutes: 60
steps:
- name: Free disk space
run: |
@@ -597,6 +606,7 @@ jobs:
)
)
runs-on: 1-gpu-5090
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-5090
IS_BLACKWELL: "1"
@@ -650,6 +660,7 @@ jobs:
)
)
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-runner
strategy:
@@ -699,6 +710,7 @@ jobs:
)
)
runs-on: 2-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 2-gpu-runner
strategy:
@@ -734,6 +746,247 @@ jobs:
fi
python3 run_suite.py --hw cuda --suite stage-b-test-large-2-gpu --auto-partition-id ${{ matrix.partition }} --auto-partition-size 2 $CONTINUE_ON_ERROR_FLAG
stage-b-test-small-1-gpu-performance:
needs: [check-changes, call-gate, stage-a-test-1, sgl-kernel-build-wheels]
if: |
always() &&
(
(inputs.target_stage == 'stage-b-test-small-1-gpu-performance') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-5090
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-5090
IS_BLACKWELL: "1"
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
source /etc/profile.d/sglang-ci.sh
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Run test
timeout-minutes: 30
run: |
source /etc/profile.d/sglang-ci.sh
cd test/
CONTINUE_ON_ERROR_FLAG=""
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
CONTINUE_ON_ERROR_FLAG="--continue-on-error"
fi
python3 run_suite.py --hw cuda --suite stage-b-test-small-1-gpu-performance $CONTINUE_ON_ERROR_FLAG
stage-b-test-large-1-gpu-performance:
needs: [check-changes, call-gate, stage-a-test-1, sgl-kernel-build-wheels]
if: |
always() &&
(
(inputs.target_stage == 'stage-b-test-large-1-gpu-performance') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-runner
strategy:
fail-fast: false
matrix:
partition: [0, 1]
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Run test
timeout-minutes: 40
run: |
cd test/
CONTINUE_ON_ERROR_FLAG=""
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
CONTINUE_ON_ERROR_FLAG="--continue-on-error"
fi
python3 run_suite.py --hw cuda --suite stage-b-test-large-1-gpu-performance --auto-partition-id ${{ matrix.partition }} --auto-partition-size 2 --timeout-per-file 1800 $CONTINUE_ON_ERROR_FLAG
stage-b-test-large-2-gpu-performance:
needs: [check-changes, call-gate, stage-a-test-1, sgl-kernel-build-wheels]
if: |
always() &&
(
(inputs.target_stage == 'stage-b-test-large-2-gpu-performance') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 2-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 2-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Run test
timeout-minutes: 30
run: |
cd test/
CONTINUE_ON_ERROR_FLAG=""
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
CONTINUE_ON_ERROR_FLAG="--continue-on-error"
fi
python3 run_suite.py --hw cuda --suite stage-b-test-large-2-gpu-performance $CONTINUE_ON_ERROR_FLAG
stage-b-test-small-1-gpu-accuracy:
needs: [check-changes, call-gate, stage-a-test-1, sgl-kernel-build-wheels]
if: |
always() &&
(
(inputs.target_stage == 'stage-b-test-small-1-gpu-accuracy') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-5090
timeout-minutes: 60
env:
RUNNER_LABELS: 1-gpu-5090
IS_BLACKWELL: "1"
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
source /etc/profile.d/sglang-ci.sh
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
git clone https://github.com/merrymercy/human-eval.git
cd human-eval
pip install -e .
- name: Run test
timeout-minutes: 25
run: |
source /etc/profile.d/sglang-ci.sh
cd test/
CONTINUE_ON_ERROR_FLAG=""
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
CONTINUE_ON_ERROR_FLAG="--continue-on-error"
fi
python3 run_suite.py --hw cuda --suite stage-b-test-small-1-gpu-accuracy $CONTINUE_ON_ERROR_FLAG
stage-b-test-large-2-gpu-accuracy:
needs: [check-changes, call-gate, stage-a-test-1, sgl-kernel-build-wheels]
if: |
always() &&
(
(inputs.target_stage == 'stage-b-test-large-2-gpu-accuracy') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 2-gpu-runner
timeout-minutes: 60
env:
RUNNER_LABELS: 2-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
git clone https://github.com/merrymercy/human-eval.git
cd human-eval
pip install -e .
- name: Run test
timeout-minutes: 25
run: |
cd test/
CONTINUE_ON_ERROR_FLAG=""
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
CONTINUE_ON_ERROR_FLAG="--continue-on-error"
fi
python3 run_suite.py --hw cuda --suite stage-b-test-large-2-gpu-accuracy $CONTINUE_ON_ERROR_FLAG
stage-b-test-4-gpu-b200:
needs: [check-changes, call-gate, stage-a-test-1, sgl-kernel-build-wheels]
if: |
@@ -747,6 +1000,7 @@ jobs:
)
)
runs-on: ${{ needs.check-changes.outputs.b200_runner }}
timeout-minutes: 60
env:
RUNNER_LABELS: ${{ needs.check-changes.outputs.b200_runner }}
strategy:
@@ -794,6 +1048,7 @@ jobs:
)
)
runs-on: 4-gpu-h100
timeout-minutes: 60
env:
RUNNER_LABELS: 4-gpu-h100
steps:
@@ -838,6 +1093,7 @@ jobs:
)
)
runs-on: ${{ needs.check-changes.outputs.b200_runner }}
timeout-minutes: 60
env:
RUNNER_LABELS: ${{ needs.check-changes.outputs.b200_runner }}
steps:
@@ -878,6 +1134,7 @@ jobs:
)
)
runs-on: 1-gpu-runner
timeout-minutes: 60
strategy:
fail-fast: false
matrix:
@@ -928,6 +1185,7 @@ jobs:
)
)
runs-on: 2-gpu-runner
timeout-minutes: 60
strategy:
fail-fast: false
matrix:
@@ -965,52 +1223,6 @@ jobs:
--total-partitions 2 \
$CONTINUE_ON_ERROR_FLAG
quantization-test:
needs: [check-changes, call-gate, stage-a-test-1]
if: |
always() &&
(
(inputs.target_stage == 'quantization-test') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Run test
timeout-minutes: 30
run: |
cd test/srt
RETRY_FLAG=""
if [[ "${{ needs.check-changes.outputs.enable_retry }}" == "true" ]]; then
RETRY_FLAG="--enable-retry"
fi
CONTINUE_ON_ERROR_FLAG=""
if [[ "${{ needs.check-changes.outputs.continue_on_error }}" == "true" ]]; then
CONTINUE_ON_ERROR_FLAG="--continue-on-error"
fi
python3 run_suite.py --suite quantization_test $RETRY_FLAG $CONTINUE_ON_ERROR_FLAG
unit-test-backend-4-gpu:
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]
if: |
@@ -1024,6 +1236,7 @@ jobs:
)
)
runs-on: 4-gpu-h100
timeout-minutes: 60
env:
RUNNER_LABELS: 4-gpu-h100
strategy:
@@ -1076,6 +1289,7 @@ jobs:
)
)
runs-on: 8-gpu-h200
timeout-minutes: 60
env:
RUNNER_LABELS: 8-gpu-h200
strategy:
@@ -1134,6 +1348,7 @@ jobs:
)
)
runs-on: 8-gpu-h20
timeout-minutes: 60
env:
SGLANG_CI_RDMA_ALL_DEVICES: "mlx5_1,mlx5_2,mlx5_3,mlx5_4"
RUNNER_LABELS: 8-gpu-h20
@@ -1174,356 +1389,6 @@ jobs:
fi
python3 run_suite.py --suite per-commit-8-gpu-h20 --auto-partition-id ${{ matrix.part }} --auto-partition-size 2 $RETRY_FLAG $CONTINUE_ON_ERROR_FLAG
performance-test-1-gpu-part-1:
needs: [check-changes, call-gate, stage-a-test-1]
if: |
always() &&
(
(inputs.target_stage == 'performance-test-1-gpu-part-1') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-runner
env:
RUNNER_LABELS: 1-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Benchmark single latency
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_bs1_small
python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_bs1_default
- name: Benchmark online latency
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_online_latency_default
- name: Benchmark offline throughput
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_default
- name: Benchmark offline throughput (Non-streaming, small batch size)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_non_stream_small_batch_size
- name: Benchmark online latency (EAGLE)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_online_latency_eagle
- name: Benchmark online latency (LoRA)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_lora_online_latency
python3 -m unittest test_bench_serving.TestBenchServing.test_lora_online_latency_with_concurrent_adapter_updates
performance-test-1-gpu-part-2:
needs: [check-changes, call-gate, stage-a-test-1]
if: |
always() &&
(
(inputs.target_stage == 'performance-test-1-gpu-part-2') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-runner
env:
RUNNER_LABELS: 1-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Benchmark offline throughput (w/o RadixAttention)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_without_radix_cache
- name: Benchmark offline throughput (w/ Triton)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_with_triton_attention_backend
- name: Benchmark offline throughput (w/ FP8)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_offline_throughput_default_fp8
- name: Benchmark VLM offline throughput
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_vlm_offline_throughput
- name: Benchmark VLM online latency
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_vlm_online_latency
performance-test-1-gpu-part-3:
needs: [check-changes, call-gate, stage-a-test-1]
if: |
always() &&
(
(inputs.target_stage == 'performance-test-1-gpu-part-3') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-runner
env:
RUNNER_LABELS: 1-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Benchmark Scores online latency and throughput
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_score_api_latency_throughput
- name: Benchmark Scores online latency and throughput (batch size scaling)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_score_api_batch_scaling
- name: Benchmark Embeddings online latency and throughput
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_embeddings_api_latency_throughput
- name: Benchmark Embeddings online latency and throughput (batch size scaling)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_embeddings_api_batch_scaling
performance-test-2-gpu:
needs: [check-changes, call-gate, stage-b-test-4-gpu-b200]
if: |
always() &&
(
(inputs.target_stage == 'performance-test-2-gpu') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 2-gpu-runner
env:
RUNNER_LABELS: 2-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
- name: Benchmark single latency (TP=2)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_moe_tp2_bs1
- name: Benchmark single latency + torch.compile (TP=2)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_one_batch.TestBenchOneBatch.test_torch_compile_tp2_bs1
- name: Benchmark offline throughput (TP=2)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_moe_offline_throughput_default
- name: Benchmark offline throughput (w/o RadixAttention) (TP=2)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_moe_offline_throughput_without_radix_cache
- name: Benchmark offline PP decode throughput (PP=2)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_pp_offline_throughput_default_decode
- name: Benchmark offline PP prefill throughput (PP=2)
timeout-minutes: 10
run: |
cd test/srt
python3 -m unittest test_bench_serving.TestBenchServing.test_pp_long_context_prefill
accuracy-test-1-gpu:
needs: [check-changes, call-gate, stage-a-test-1]
if: |
always() &&
(
(inputs.target_stage == 'accuracy-test-1-gpu') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 1-gpu-runner
env:
RUNNER_LABELS: 1-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
git clone https://github.com/merrymercy/human-eval.git
cd human-eval
pip install -e .
- name: Evaluate accuracy
timeout-minutes: 25
run: |
cd test/srt
python3 -m sglang.test.ci.run_with_retry test_eval_accuracy_large.py
accuracy-test-2-gpu:
needs: [check-changes, call-gate, accuracy-test-1-gpu]
if: |
always() &&
(
(inputs.target_stage == 'accuracy-test-2-gpu') ||
(
!inputs.target_stage &&
(github.event_name == 'schedule' || (!failure() && !cancelled())) &&
((needs.check-changes.outputs.main_package == 'true') || (needs.check-changes.outputs.sgl_kernel == 'true'))
)
)
runs-on: 2-gpu-runner
env:
RUNNER_LABELS: 2-gpu-runner
steps:
- name: Checkout code
uses: actions/checkout@v4
with:
ref: ${{ inputs.pr_head_sha || inputs.ref || github.sha }}
- name: Download artifacts
if: needs.check-changes.outputs.sgl_kernel == 'true'
uses: actions/download-artifact@v4
with:
path: sgl-kernel/dist/
merge-multiple: true
pattern: wheel-python3.10-cuda12.9
- name: Install dependencies
timeout-minutes: 10
run: |
CUSTOM_BUILD_SGL_KERNEL=${{needs.check-changes.outputs.sgl_kernel}} bash scripts/ci/ci_install_dependency.sh
git clone https://github.com/merrymercy/human-eval.git
cd human-eval
pip install -e .
- name: Evaluate accuracy (TP=2)
timeout-minutes: 25
run: |
cd test/srt
python3 -m sglang.test.ci.run_with_retry test_moe_eval_accuracy_large.py
unit-test-deepep-4-gpu:
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]
if: |
@@ -1537,6 +1402,7 @@ jobs:
)
)
runs-on: 4-gpu-h100
timeout-minutes: 60
env:
RUNNER_LABELS: 4-gpu-h100
steps:
@@ -1634,6 +1500,7 @@ jobs:
)
)
runs-on: ${{ needs.check-changes.outputs.b200_runner }}
timeout-minutes: 60
env:
RUNNER_LABELS: ${{ needs.check-changes.outputs.b200_runner }}
strategy:
@@ -1687,6 +1554,7 @@ jobs:
)
)
runs-on: 4-gpu-gb200
timeout-minutes: 60
env:
RUNNER_LABELS: 4-gpu-gb200
strategy:
@@ -1746,18 +1614,16 @@ jobs:
stage-b-test-small-1-gpu,
stage-b-test-large-1-gpu,
stage-b-test-large-2-gpu,
stage-b-test-small-1-gpu-performance,
stage-b-test-large-1-gpu-performance,
stage-b-test-large-2-gpu-performance,
stage-b-test-small-1-gpu-accuracy,
stage-b-test-large-2-gpu-accuracy,
stage-c-test-large-4-gpu,
quantization-test,
stage-b-test-4-gpu-b200,
unit-test-backend-4-gpu,
unit-test-backend-8-gpu-h20,
unit-test-backend-8-gpu-h200,
performance-test-1-gpu-part-1,
performance-test-1-gpu-part-2,
performance-test-1-gpu-part-3,
performance-test-2-gpu,
accuracy-test-1-gpu,
accuracy-test-2-gpu,
unit-test-deepep-4-gpu,
# unit-test-deepep-8-gpu, # Disabled, see #17175
unit-test-backend-4-gpu-b200,