1355 lines
63 KiB
YAML
1355 lines
63 KiB
YAML
name: Test AITER Latest (AMD Nightly)
|
|
|
|
on:
|
|
workflow_dispatch:
|
|
inputs:
|
|
aiter_ref:
|
|
description: 'AITER git ref to test (branch, tag, or SHA). Default: main (latest commit)'
|
|
required: false
|
|
type: string
|
|
default: 'main'
|
|
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 Unit Tests
|
|
- 'nightly-test-1-gpu-unit'
|
|
# MI30x Accuracy Tests (GSM8K / MMMU)
|
|
- 'nightly-accuracy-2-gpu'
|
|
- 'nightly-accuracy-2-gpu-vlm'
|
|
- 'nightly-perf-2-gpu-text'
|
|
- 'nightly-perf-2-gpu-vlm'
|
|
- 'nightly-accuracy-8-gpu'
|
|
# MI30x Accuracy + Performance Tests (combined)
|
|
- 'nightly-8-gpu-grok1-int4'
|
|
- 'nightly-8-gpu-grok2'
|
|
- 'nightly-8-gpu-deepseek-v31'
|
|
- 'nightly-8-gpu-deepseek-v32'
|
|
- 'nightly-8-gpu-deepseek-v32-mtp'
|
|
- 'nightly-8-gpu-kimi-k25'
|
|
- 'nightly-8-gpu-qwen3-235b'
|
|
- 'nightly-8-gpu-glm5'
|
|
# MI35x jobs
|
|
- 'nightly-test-1-gpu-mi35x'
|
|
- 'nightly-8-gpu-mi35x-qwen3-235b-mxfp4'
|
|
- 'nightly-8-gpu-mi35x-kimi-k25'
|
|
- 'nightly-8-gpu-mi35x-glm5'
|
|
- 'nightly-accuracy-8-gpu-mi35x'
|
|
- 'nightly-8-gpu-mi35x-grok1-int4'
|
|
- 'nightly-8-gpu-mi35x-grok2'
|
|
- 'nightly-8-gpu-mi35x-deepseek-r1-mxfp4'
|
|
- 'nightly-accuracy-8-gpu-mi35x-deepseek-v32'
|
|
- 'nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp'
|
|
- 'nightly-perf-8-gpu-mi35x-deepseek-v32-basic'
|
|
- 'nightly-perf-8-gpu-mi35x-deepseek-v32-mtp'
|
|
|
|
concurrency:
|
|
group: nightly-test-amd-aiter-latest-${{ github.run_id }}
|
|
cancel-in-progress: true
|
|
|
|
jobs:
|
|
# ============================================== Resolve AITER Ref ==============================================
|
|
resolve-aiter:
|
|
runs-on: ubuntu-latest
|
|
outputs:
|
|
aiter_sha: ${{ steps.resolve.outputs.sha }}
|
|
aiter_short_sha: ${{ steps.resolve.outputs.short_sha }}
|
|
aiter_ref: ${{ steps.resolve.outputs.ref }}
|
|
steps:
|
|
- name: Resolve AITER commit
|
|
id: resolve
|
|
run: |
|
|
REF="${{ inputs.aiter_ref || 'main' }}"
|
|
echo "Resolving AITER ref: ${REF}"
|
|
|
|
SHA=$(git ls-remote https://github.com/ROCm/aiter.git "${REF}" | head -1 | cut -f1)
|
|
if [ -z "$SHA" ]; then
|
|
SHA=$(git ls-remote https://github.com/ROCm/aiter.git "refs/heads/${REF}" | head -1 | cut -f1)
|
|
fi
|
|
if [ -z "$SHA" ]; then
|
|
SHA=$(git ls-remote https://github.com/ROCm/aiter.git "refs/tags/${REF}" | head -1 | cut -f1)
|
|
fi
|
|
if [ -z "$SHA" ]; then
|
|
SHA="${REF}"
|
|
fi
|
|
|
|
SHORT_SHA="${SHA:0:12}"
|
|
echo "sha=${SHA}" >> $GITHUB_OUTPUT
|
|
echo "short_sha=${SHORT_SHA}" >> $GITHUB_OUTPUT
|
|
echo "ref=${REF}" >> $GITHUB_OUTPUT
|
|
|
|
echo "### AITER Ref Resolution" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Requested ref:** \`${REF}\`" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Resolved SHA:** \`${SHA}\`" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **AITER repo:** https://github.com/ROCm/aiter/commit/${SHA}" >> $GITHUB_STEP_SUMMARY
|
|
|
|
# ============================================== MI30x Unit Tests ==============================================
|
|
nightly-test-1-gpu-unit:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-test-1-gpu-unit'
|
|
runs-on: linux-mi325-gpu-1
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
echo "Overriding AITER_COMMIT to: ${AITER_SHA}"
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
echo '```' >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang pip show amd-aiter 2>/dev/null >> $GITHUB_STEP_SUMMARY || echo "amd-aiter not installed" >> $GITHUB_STEP_SUMMARY
|
|
echo '```' >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --format='%H %s'" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Nightly Unit Test (1-GPU)
|
|
timeout-minutes: 90
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-1-gpu --nightly --timeout-per-file 900 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== MI30x Accuracy Tests ==============================================
|
|
nightly-accuracy-2-gpu:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-2-gpu'
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Nightly Test (2-GPU)
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-accuracy-2-gpu-vlm:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-2-gpu-vlm'
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Nightly Accuracy Test (2-GPU VLM MMMU)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-2-gpu-vlm --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-perf-2-gpu-text:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-2-gpu-text'
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Performance Test (2-GPU Text Models)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_AITER=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-perf-text-2-gpu --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-perf-2-gpu-vlm:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-2-gpu-vlm'
|
|
runs-on: linux-mi325-gpu-2
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Performance Test (2-GPU VLM Models)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_AITER=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-perf-vlm-2-gpu --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-accuracy-8-gpu:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU GPT-OSS)
|
|
timeout-minutes: 180
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-gpt-oss --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Accuracy Test (8-GPU Grok1-FP8)
|
|
timeout-minutes: 60
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-grok1-fp8 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== MI30x Combined Accuracy + Performance Tests ==============================================
|
|
nightly-8-gpu-grok1-int4:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-grok1-int4'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-grok1-int4 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-grok1-int4 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-grok2:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-grok2'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-grok2 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-grok2 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-deepseek-v31:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-deepseek-v31'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU DeepSeek-V3.1)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_AITER=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-deepseek-v31 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test (8-GPU DeepSeek-V3.1)
|
|
timeout-minutes: 300
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e SGLANG_USE_ROCM700A=1 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-deepseek-v31 --nightly --timeout-per-file 18000 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-deepseek-v32:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-deepseek-v32'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU DeepSeek-V3.2 Basic)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-deepseek-v32 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test (8-GPU DeepSeek-V3.2 Basic)
|
|
timeout-minutes: 150
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-deepseek-v32-basic --nightly --timeout-per-file 5400 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-deepseek-v32-mtp:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-deepseek-v32-mtp'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU DeepSeek-V3.2 MTP)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-deepseek-v32-mtp --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test (8-GPU DeepSeek-V3.2 MTP)
|
|
timeout-minutes: 180
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-deepseek-v32-mtp --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-kimi-k25:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-kimi-k25'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU Kimi-K2.5)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-kimi-k25 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-qwen3-235b:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-qwen3-235b'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test + Performance Test (8-GPU Qwen3)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-8-gpu-qwen3-235b --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-glm5:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-glm5'
|
|
runs-on: linux-mi325-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install git+https://github.com/huggingface/transformers.git
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test (8-GPU GLM-5 NSA)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-glm5 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== MI35x Tests ==============================================
|
|
nightly-test-1-gpu-mi35x:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-test-1-gpu-mi35x'
|
|
runs-on: linux-mi35x-gpu-1
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Nightly Test MI35x (1-GPU)
|
|
timeout-minutes: 90
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-1-gpu-mi35x --nightly --timeout-per-file 900 --continue-on-error || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-accuracy-8-gpu-mi35x:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-mi35x'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU GPT-OSS)
|
|
timeout-minutes: 180
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-mi35x-grok1-int4:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-grok1-int4'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU Grok1-INT4)
|
|
timeout-minutes: 90
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-grok1-int4 --nightly --timeout-per-file 5400 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test MI35x (8-GPU Grok1-INT4)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-grok1-int4 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-mi35x-grok2:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-grok2'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-grok2 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test MI35x (8-GPU Grok2)
|
|
timeout-minutes: 60
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e RCCL_MSCCL_ENABLE=0 \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-grok2 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-mi35x-deepseek-r1-mxfp4:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-deepseek-r1-mxfp4'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU DeepSeek-R1-MXFP4)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x-deepseek-r1-mxfp4 --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
- name: Performance Test MI35x (8-GPU DeepSeek-R1-MXFP4)
|
|
timeout-minutes: 300
|
|
continue-on-error: true
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 registered/amd/perf/mi35x/test_deepseek_r1_mxfp4_perf_mi35x.py || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-accuracy-8-gpu-mi35x-deepseek-v32:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-mi35x-deepseek-v32'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU DeepSeek-V3.2)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x-deepseek-v32 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU DeepSeek-V3.2 TP+MTP)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-deepseek-v32-mtp --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-perf-8-gpu-mi35x-deepseek-v32-basic:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-8-gpu-mi35x-deepseek-v32-basic'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Performance Test MI35x (8-GPU DeepSeek-V3.2 Basic)
|
|
timeout-minutes: 150
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-deepseek-v32-basic --nightly --timeout-per-file 5400 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-mi35x-kimi-k25:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-kimi-k25'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU Kimi-K2.5)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-accuracy-8-gpu-mi35x-kimi-k25 --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-mi35x-qwen3-235b-mxfp4:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-qwen3-235b-mxfp4'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test + Performance Test MI35x (8-GPU Qwen3-235B-MXFP4)
|
|
timeout-minutes: 120
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-8-gpu-mi35x-qwen3-235b-mxfp4 --nightly --timeout-per-file 3600 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-8-gpu-mi35x-glm5:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-8-gpu-mi35x-glm5'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install git+https://github.com/huggingface/transformers.git
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Accuracy Test MI35x (8-GPU GLM-5 NSA)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-amd-8-gpu-mi35x-glm5 --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
nightly-perf-8-gpu-mi35x-deepseek-v32-mtp:
|
|
needs: resolve-aiter
|
|
if: inputs.job_filter == '' || inputs.job_filter == 'all' || inputs.job_filter == 'nightly-perf-8-gpu-mi35x-deepseek-v32-mtp'
|
|
runs-on: linux-mi35x-gpu-8
|
|
steps:
|
|
- name: Checkout code
|
|
uses: actions/checkout@v4
|
|
|
|
- name: Override AITER commit in Dockerfile
|
|
run: |
|
|
AITER_SHA="${{ needs.resolve-aiter.outputs.aiter_sha }}"
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm.Dockerfile
|
|
sed -i "s/AITER_COMMIT=\"[^\"]*\"/AITER_COMMIT=\"${AITER_SHA}\"/g" docker/rocm720.Dockerfile
|
|
|
|
- name: Setup docker
|
|
run: |
|
|
touch github_summary.md
|
|
bash scripts/ci/amd/amd_ci_start_container.sh
|
|
env:
|
|
GITHUB_WORKSPACE: ${{ github.workspace }}
|
|
|
|
- name: Install dependencies
|
|
run: |
|
|
bash scripts/ci/amd/amd_ci_install_dependency.sh
|
|
bash scripts/ci/amd/amd_ci_exec.sh pip install tabulate
|
|
|
|
- name: Verify AITER version
|
|
run: |
|
|
echo "### AITER Version Info" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **Target:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\` (\`${{ needs.resolve-aiter.outputs.aiter_short_sha }}\`)" >> $GITHUB_STEP_SUMMARY
|
|
docker exec ci_sglang bash -c "cd /sgl-workspace/aiter 2>/dev/null && git log -1 --oneline" >> $GITHUB_STEP_SUMMARY 2>/dev/null || true
|
|
|
|
- name: Performance Test MI35x (8-GPU DeepSeek-V3.2 MTP)
|
|
timeout-minutes: 180
|
|
run: |
|
|
> github_summary.md
|
|
bash scripts/ci/amd/amd_ci_exec.sh -w /sglang-checkout/test \
|
|
-e GITHUB_STEP_SUMMARY="/sglang-checkout/github_summary.md" \
|
|
python3 run_suite.py --hw amd --suite nightly-perf-8-gpu-mi35x-deepseek-v32-mtp --nightly --timeout-per-file 7200 || TEST_EXIT_CODE=$?
|
|
echo "$(<github_summary.md )" >> $GITHUB_STEP_SUMMARY || true
|
|
exit ${TEST_EXIT_CODE:-0}
|
|
|
|
# ============================================== Summary ==============================================
|
|
check-all-jobs:
|
|
if: always()
|
|
needs:
|
|
- resolve-aiter
|
|
# MI30x Unit Tests
|
|
- nightly-test-1-gpu-unit
|
|
# MI30x Accuracy Tests
|
|
- nightly-accuracy-2-gpu
|
|
- nightly-accuracy-2-gpu-vlm
|
|
- nightly-accuracy-8-gpu
|
|
# MI30x Combined Accuracy + Performance Tests
|
|
- nightly-8-gpu-grok1-int4
|
|
- nightly-8-gpu-grok2
|
|
- nightly-8-gpu-deepseek-v31
|
|
- nightly-8-gpu-deepseek-v32
|
|
- nightly-8-gpu-deepseek-v32-mtp
|
|
- nightly-8-gpu-kimi-k25
|
|
- nightly-8-gpu-qwen3-235b
|
|
- nightly-8-gpu-glm5
|
|
# MI35x jobs
|
|
- nightly-test-1-gpu-mi35x
|
|
- nightly-accuracy-8-gpu-mi35x
|
|
- nightly-8-gpu-mi35x-grok1-int4
|
|
- nightly-8-gpu-mi35x-grok2
|
|
- nightly-8-gpu-mi35x-deepseek-r1-mxfp4
|
|
- nightly-accuracy-8-gpu-mi35x-deepseek-v32
|
|
- nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp
|
|
- nightly-8-gpu-mi35x-kimi-k25
|
|
- nightly-8-gpu-mi35x-qwen3-235b-mxfp4
|
|
- nightly-8-gpu-mi35x-glm5
|
|
runs-on: ubuntu-latest
|
|
steps:
|
|
- name: Summary
|
|
run: |
|
|
echo "## AITER Latest Test Results" >> $GITHUB_STEP_SUMMARY
|
|
echo "" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **AITER ref:** \`${{ needs.resolve-aiter.outputs.aiter_ref }}\`" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **AITER SHA:** \`${{ needs.resolve-aiter.outputs.aiter_sha }}\`" >> $GITHUB_STEP_SUMMARY
|
|
echo "- **AITER commit:** https://github.com/ROCm/aiter/commit/${{ needs.resolve-aiter.outputs.aiter_sha }}" >> $GITHUB_STEP_SUMMARY
|
|
echo "" >> $GITHUB_STEP_SUMMARY
|
|
echo "### MI30x Results" >> $GITHUB_STEP_SUMMARY
|
|
echo "| Job | Result |" >> $GITHUB_STEP_SUMMARY
|
|
echo "|-----|--------|" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-test-1-gpu-unit | \`${{ needs.nightly-test-1-gpu-unit.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-accuracy-2-gpu | \`${{ needs.nightly-accuracy-2-gpu.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-accuracy-2-gpu-vlm | \`${{ needs.nightly-accuracy-2-gpu-vlm.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-perf-2-gpu-text | \`${{ needs.nightly-perf-2-gpu-text.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-perf-2-gpu-vlm | \`${{ needs.nightly-perf-2-gpu-vlm.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-accuracy-8-gpu | \`${{ needs.nightly-accuracy-8-gpu.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-grok1-int4 | \`${{ needs.nightly-8-gpu-grok1-int4.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-grok2 | \`${{ needs.nightly-8-gpu-grok2.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-deepseek-v31 | \`${{ needs.nightly-8-gpu-deepseek-v31.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-deepseek-v32 | \`${{ needs.nightly-8-gpu-deepseek-v32.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-deepseek-v32-mtp | \`${{ needs.nightly-8-gpu-deepseek-v32-mtp.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-kimi-k25 | \`${{ needs.nightly-8-gpu-kimi-k25.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-qwen3-235b | \`${{ needs.nightly-8-gpu-qwen3-235b.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-glm5 | \`${{ needs.nightly-8-gpu-glm5.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "" >> $GITHUB_STEP_SUMMARY
|
|
echo "### MI35x Results" >> $GITHUB_STEP_SUMMARY
|
|
echo "| Job | Result |" >> $GITHUB_STEP_SUMMARY
|
|
echo "|-----|--------|" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-test-1-gpu-mi35x | \`${{ needs.nightly-test-1-gpu-mi35x.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-accuracy-8-gpu-mi35x | \`${{ needs.nightly-accuracy-8-gpu-mi35x.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-mi35x-grok1-int4 | \`${{ needs.nightly-8-gpu-mi35x-grok1-int4.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-mi35x-grok2 | \`${{ needs.nightly-8-gpu-mi35x-grok2.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-mi35x-deepseek-r1-mxfp4 | \`${{ needs.nightly-8-gpu-mi35x-deepseek-r1-mxfp4.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-accuracy-8-gpu-mi35x-deepseek-v32 | \`${{ needs.nightly-accuracy-8-gpu-mi35x-deepseek-v32.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp | \`${{ needs.nightly-accuracy-8-gpu-mi35x-deepseek-v32-mtp.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-perf-8-gpu-mi35x-deepseek-v32-basic | \`${{ needs.nightly-perf-8-gpu-mi35x-deepseek-v32-basic.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-perf-8-gpu-mi35x-deepseek-v32-mtp | \`${{ needs.nightly-perf-8-gpu-mi35x-deepseek-v32-mtp.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-mi35x-kimi-k25 | \`${{ needs.nightly-8-gpu-mi35x-kimi-k25.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-mi35x-qwen3-235b-mxfp4 | \`${{ needs.nightly-8-gpu-mi35x-qwen3-235b-mxfp4.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
echo "| nightly-8-gpu-mi35x-glm5 | \`${{ needs.nightly-8-gpu-mi35x-glm5.result }}\` |" >> $GITHUB_STEP_SUMMARY
|
|
|
|
- name: Check if any job failed
|
|
run: |
|
|
if [[ "${{ contains(needs.*.result, 'failure') }}" == "true" ]]; then
|
|
echo "One or more test jobs failed with AITER ref: ${{ needs.resolve-aiter.outputs.aiter_ref }}"
|
|
exit 1
|
|
fi
|
|
if [[ "${{ contains(needs.*.result, 'cancelled') }}" == "true" ]]; then
|
|
echo "One or more test jobs were cancelled"
|
|
exit 1
|
|
fi
|
|
echo "All test jobs passed with AITER ${{ needs.resolve-aiter.outputs.aiter_ref }} (${{ needs.resolve-aiter.outputs.aiter_short_sha }})"
|