904 lines
42 KiB
YAML
904 lines
42 KiB
YAML
name: Nightly Test (AMD ROCm 7.2)
|
|
|
|
on:
|
|
schedule:
|
|
- cron: '0 2 * * *'
|
|
push:
|
|
branches:
|
|
- main
|
|
paths:
|
|
- "python/sglang/version.py"
|
|
workflow_dispatch:
|
|
inputs:
|
|
job_filter:
|
|
description: 'Select which job to run (leave empty or "all" to run all jobs)'
|
|
required: false
|
|
type: choice
|
|
default: 'all'
|
|
options:
|
|
- 'all'
|
|
# MI30x ROCm 7.2 Unit Tests
|
|
- 'nightly-test-1-gpu-unit-rocm720'
|
|
# MI30x ROCm 7.2 Accuracy Tests (GSM8K / MMMU)
|
|
- 'nightly-accuracy-2-gpu-rocm720'
|
|
- 'nightly-accuracy-2-gpu-vlm-rocm720'
|
|
- 'nightly-perf-2-gpu-text-rocm720'
|
|
- 'nightly-perf-2-gpu-vlm-rocm720'
|
|
- 'nightly-accuracy-8-gpu-rocm720'
|
|
# MI30x ROCm 7.2 Accuracy + Performance Tests (combined)
|
|
- 'nightly-8-gpu-grok1-int4-rocm720'
|
|
- 'nightly-8-gpu-grok2-rocm720'
|
|
- 'nightly-8-gpu-deepseek-v31-rocm720'
|
|
- 'nightly-8-gpu-deepseek-v32-rocm720'
|
|
- 'nightly-8-gpu-deepseek-v32-mtp-rocm720'
|
|
- 'nightly-8-gpu-kimi-k25-rocm720'
|
|
# MI35x ROCm 7.2 jobs
|
|
- 'nightly-test-1-gpu-mi35x-rocm720'
|
|
- 'nightly-accuracy-8-gpu-mi35x-rocm720'
|
|
- 'nightly-8-gpu-mi35x-grok1-int4-rocm720'
|
|
- 'nightly-8-gpu-mi35x-grok2-rocm720'
|
|
- 'nightly-8-gpu-mi35x-deepseek-r1-mxfp4-rocm720'
|
|
- 'nightly-accuracy-8-gpu-mi35x-deepseek-v32-rocm720'
|
|
- 'nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp-rocm720'
|
|
- 'nightly-perf-8-gpu-mi35x-deepseek-v32-basic-rocm720'
|
|
- 'nightly-perf-8-gpu-mi35x-deepseek-v32-mtp-rocm720'
|
|
- 'nightly-8-gpu-mi35x-kimi-k25-rocm720'
|
|
workflow_call:
|
|
inputs:
|
|
ref:
|
|
description: 'Git ref (branch, tag, or SHA) to test. If not provided, uses the default branch.'
|
|
required: false
|
|
type: string
|
|
default: ''
|
|
job_filter:
|
|
description: 'Select which job to run (leave empty or "all" to run all jobs)'
|
|
required: false
|
|
type: string
|
|
default: 'all'
|
|
|
|
concurrency:
|
|
group: nightly-test-amd-rocm720-${{ inputs.ref || github.ref }}
|
|
cancel-in-progress: ${{ github.event_name != 'workflow_call' }}
|
|
|
|
jobs:
|
|
# ============================================== MI30x ROCm 7.2 Unit Tests ==============================================
|
|
# 1-GPU Unit Tests - LoRA, debug utils, scheduler, etc. (MI30x ROCm 7.2)
|
|
nightly-test-1-gpu-unit-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-test-1-gpu-unit-rocm720')
|
|
runs-on: linux-mi325-gpu-1
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build
|
|
|
|
- name: Nightly Unit Test ROCm 7.2 (1-GPU)
|
|
timeout-minutes: 90
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-1-gpu --nightly --timeout-per-file 900 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== MI30x ROCm 7.2 Accuracy Tests ==============================================
|
|
# 2-GPU Accuracy Tests - GSM8K eval (MI30x ROCm 7.2)
|
|
nightly-accuracy-2-gpu-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-2-gpu-rocm720')
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build
|
|
|
|
- name: Nightly Test ROCm 7.2 (2-GPU)
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 2-GPU VLM Accuracy Tests - Vision-Language Models MMMU evaluation (ROCm 7.2)
|
|
nightly-accuracy-2-gpu-vlm-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-2-gpu-vlm-rocm720')
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build
|
|
|
|
- name: Nightly Accuracy Test ROCm 7.2 (2-GPU VLM MMMU)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-2-gpu-vlm --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 2-GPU Text Models Performance Tests (ROCm 7.2)
|
|
nightly-perf-2-gpu-text-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-2-gpu-text-rocm720')
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build
|
|
|
|
- name: Performance Test ROCm 7.2 (2-GPU Text Models)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_AITER=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-perf-text-2-gpu --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 2-GPU VLM Performance Tests (ROCm 7.2)
|
|
nightly-perf-2-gpu-vlm-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-2-gpu-vlm-rocm720')
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build
|
|
|
|
- name: Performance Test ROCm 7.2 (2-GPU VLM Models)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_AITER=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-perf-vlm-2-gpu --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 8-GPU Accuracy Tests - GPT-OSS, Grok1-FP8 (ROCm 7.2)
|
|
nightly-accuracy-8-gpu-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU GPT-OSS)
|
|
timeout-minutes: 180
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-gpt-oss --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU Grok1-FP8)
|
|
timeout-minutes: 60
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-grok1-fp8 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== MI30x ROCm 7.2 Combined Accuracy + Performance Tests ==============================================
|
|
# 8-GPU Grok1-INT4 (Accuracy + Performance) ROCm 7.2
|
|
nightly-8-gpu-grok1-int4-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-grok1-int4-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-grok1-int4 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test ROCm 7.2 (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-grok1-int4 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 8-GPU Grok2 (Accuracy + Performance) ROCm 7.2
|
|
nightly-8-gpu-grok2-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-grok2-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-grok2 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test ROCm 7.2 (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-grok2 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 8-GPU DeepSeek-V3.1 (Accuracy + Performance) ROCm 7.2
|
|
nightly-8-gpu-deepseek-v31-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-deepseek-v31-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU DeepSeek-V3.1)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_AITER=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-deepseek-v31 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test ROCm 7.2 (8-GPU DeepSeek-V3.1)
|
|
timeout-minutes: 300
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_ROCM700A=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-deepseek-v31 --nightly --timeout-per-file 18000 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 8-GPU DeepSeek-V3.2 (Basic Accuracy + Perf) ROCm 7.2
|
|
nightly-8-gpu-deepseek-v32-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-deepseek-v32-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU DeepSeek-V3.2 Basic)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-deepseek-v32 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test ROCm 7.2 (8-GPU DeepSeek-V3.2 Basic)
|
|
timeout-minutes: 150
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-deepseek-v32-basic --nightly --timeout-per-file 5400 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 8-GPU DeepSeek-V3.2 MTP (MTP Accuracy + Perf) ROCm 7.2
|
|
nightly-8-gpu-deepseek-v32-mtp-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-deepseek-v32-mtp-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU DeepSeek-V3.2 MTP)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-deepseek-v32-mtp --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test ROCm 7.2 (8-GPU DeepSeek-V3.2 MTP)
|
|
timeout-minutes: 180
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-deepseek-v32-mtp --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# 8-GPU Kimi-K2.5 (Accuracy) ROCm 7.2
|
|
nightly-8-gpu-kimi-k25-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-kimi-k25-rocm720')
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
|
|
- name: Accuracy Test ROCm 7.2 (8-GPU Kimi-K2.5)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-kimi-k25 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== MI35x ROCm 7.2 Tests ==============================================
|
|
# MI35x 1-GPU ROCm 7.2 tests
|
|
nightly-test-1-gpu-mi35x-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-test-1-gpu-mi35x-rocm720')
|
|
runs-on: linux-mi35x-gpu-1
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build
|
|
|
|
- name: Nightly Test MI35x ROCm 7.2 (1-GPU)
|
|
timeout-minutes: 90
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-1-gpu-mi35x --nightly --timeout-per-file 900 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU Accuracy Tests - GPT-OSS (ROCm 7.2)
|
|
nightly-accuracy-8-gpu-mi35x-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-mi35x-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU GPT-OSS)
|
|
timeout-minutes: 180
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU Grok1-INT4 (Accuracy + Performance) ROCm 7.2
|
|
nightly-8-gpu-mi35x-grok1-int4-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-grok1-int4-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-grok1-int4 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test MI35x ROCm 7.2 (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-grok1-int4 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU Grok2 (Accuracy + Performance) ROCm 7.2
|
|
nightly-8-gpu-mi35x-grok2-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-grok2-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-grok2 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test MI35x ROCm 7.2 (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-grok2 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU DeepSeek-R1-MXFP4 (Accuracy + Performance) ROCm 7.2
|
|
nightly-8-gpu-mi35x-deepseek-r1-mxfp4-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-deepseek-r1-mxfp4-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU DeepSeek-R1-MXFP4)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x-deepseek-r1-mxfp4 --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test MI35x ROCm 7.2 (8-GPU DeepSeek-R1-MXFP4)
|
|
timeout-minutes: 300
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 registered/amd/perf/mi35x/test_deepseek_r1_mxfp4_perf_mi35x.py || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU DeepSeek-V3.2 Accuracy Test (ROCm 7.2)
|
|
nightly-accuracy-8-gpu-mi35x-deepseek-v32-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-mi35x-deepseek-v32-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU DeepSeek-V3.2)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x-deepseek-v32 --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU DeepSeek-V3.2 TP+MTP Accuracy Test (ROCm 7.2)
|
|
nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU DeepSeek-V3.2 TP+MTP)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-deepseek-v32-mtp --nightly --timeout-per-file 3600 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU DeepSeek-V3.2 Performance Test (Basic) ROCm 7.2
|
|
nightly-perf-8-gpu-mi35x-deepseek-v32-basic-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-8-gpu-mi35x-deepseek-v32-basic-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Performance Test MI35x ROCm 7.2 (8-GPU DeepSeek-V3.2 Basic)
|
|
timeout-minutes: 150
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-deepseek-v32-basic --nightly --timeout-per-file 5400 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU Kimi-K2.5 (Accuracy) ROCm 7.2
|
|
nightly-8-gpu-mi35x-kimi-k25-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-kimi-k25-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Accuracy Test MI35x ROCm 7.2 (8-GPU Kimi-K2.5)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-kimi-k25 --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# MI35x 8-GPU DeepSeek-V3.2 Performance Test (MTP) ROCm 7.2
|
|
nightly-perf-8-gpu-mi35x-deepseek-v32-mtp-rocm720:
|
|
if: (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request') && (inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-8-gpu-mi35x-deepseek-v32-mtp-rocm720')
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
with:
|
|
ref: ${{ inputs.ref || github.ref }}
|
|
|
|
- name: Setup docker (ROCm 7.2)
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh --rocm-version rocm720
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh --skip-aiter-build --skip-test-time-deps
|
|
# Install tabulate for run_suite.py (missing in MI35x container)
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Performance Test MI35x ROCm 7.2 (8-GPU DeepSeek-V3.2 MTP)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md # Clear summary file
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-deepseek-v32-mtp --nightly --timeout-per-file 7200 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
check-all-jobs:
|
|
if: always() && (github.repository == 'sgl-project/sglang' || github.event_name == 'pull_request' || github.event_name == 'workflow_dispatch')
|
|
needs:
|
|
# MI30x ROCm 7.2 Unit Tests
|
|
- nightly-test-1-gpu-unit-rocm720
|
|
# MI30x ROCm 7.2 Accuracy Tests
|
|
- nightly-accuracy-2-gpu-rocm720
|
|
- nightly-accuracy-2-gpu-vlm-rocm720
|
|
# MI30x ROCm 7.2 Performance Tests
|
|
- nightly-perf-2-gpu-text-rocm720
|
|
- nightly-perf-2-gpu-vlm-rocm720
|
|
- nightly-accuracy-8-gpu-rocm720
|
|
# MI30x ROCm 7.2 Combined Accuracy + Performance Tests
|
|
- nightly-8-gpu-grok1-int4-rocm720
|
|
- nightly-8-gpu-grok2-rocm720
|
|
- nightly-8-gpu-deepseek-v31-rocm720
|
|
- nightly-8-gpu-deepseek-v32-rocm720
|
|
- nightly-8-gpu-deepseek-v32-mtp-rocm720
|
|
- nightly-8-gpu-kimi-k25-rocm720
|
|
# MI35x ROCm 7.2 jobs
|
|
- nightly-test-1-gpu-mi35x-rocm720
|
|
- nightly-accuracy-8-gpu-mi35x-rocm720
|
|
- nightly-8-gpu-mi35x-grok1-int4-rocm720
|
|
- nightly-8-gpu-mi35x-grok2-rocm720
|
|
- nightly-8-gpu-mi35x-deepseek-r1-mxfp4-rocm720
|
|
- nightly-accuracy-8-gpu-mi35x-deepseek-v32-rocm720
|
|
- nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp-rocm720
|
|
- nightly-perf-8-gpu-mi35x-deepseek-v32-basic-rocm720
|
|
- nightly-perf-8-gpu-mi35x-deepseek-v32-mtp-rocm720
|
|
- nightly-8-gpu-mi35x-kimi-k25-rocm720
|
|
runs-on: ubuntu-latest
|
|
steps:
|
|
- name: Check if any job failed
|
|
run: |
|
|
if [[ "${{ contains(needs.*.result, 'failure') }}" == "true" ]]; then
|
|
echo "One or more ROCm 7.2 nightly test jobs failed"
|
|
exit 1
|
|
fi
|
|
if [[ "${{ contains(needs.*.result, 'cancelled') }}" == "true" ]]; then
|
|
echo "One or more ROCm 7.2 nightly test jobs were cancelled"
|
|
exit 1
|
|
fi
|
|
echo "All ROCm 7.2 nightly test jobs passed"
|