[Nightly][Refactor]Migrate nightly single-node model tests from .py to .yaml (#6503)

### What this PR does / why we need it?
This PR refactors the nightly single-node model test by migrating test
configurations from Python scripts to a more maintainable `YAML-based`
format.

| Original PR | Python (`.py`) | YAML (`.yaml`) |
| :--- | :--- | :--- |
| [#3568](https://github.com/vllm-project/vllm-ascend/pull/3568) |
`test_deepseek_r1_0528_w8a8_eplb.py` | `DeepSeek-R1-0528-W8A8.yaml` |
| [#3631](https://github.com/vllm-project/vllm-ascend/pull/3631) |
`test_deepseek_r1_0528_w8a8.py` | `DeepSeek-R1-0528-W8A8.yaml` |
| [#5874](https://github.com/vllm-project/vllm-ascend/pull/5874) |
`test_deepseek_r1_w8a8_hbm.py` | `DeepSeek-R1-W8A8-HBM.yaml` |
| [#3908](https://github.com/vllm-project/vllm-ascend/pull/3908) |
`test_deepseek_v3_2_w8a8.py` | `DeepSeek-V3.2-W8A8.yaml` |
| [#5682](https://github.com/vllm-project/vllm-ascend/pull/5682) |
`test_kimi_k2_thinking.py` | `Kimi-K2-Thinking.yaml` |
| [#4111](https://github.com/vllm-project/vllm-ascend/pull/4111) |
`test_mtpx_deepseek_r1_0528_w8a8.py` | `MTPX-DeepSeek-R1-0528-W8A8.yaml`
|
| [#3733](https://github.com/vllm-project/vllm-ascend/pull/3733) |
`test_prefix_cache_deepseek_r1_0528_w8a8.py` |
`Prefix-Cache-DeepSeek-R1-0528-W8A8.yaml` |
| [#6543](https://github.com/vllm-project/vllm-ascend/pull/6543) |
`test_qwen3_235b_w8a8.py` | `Qwen3-235B-A22B-W8A8.yaml` |
| [#6543](https://github.com/vllm-project/vllm-ascend/pull/6543) |
`test_qwen3_235b_a22b_w8a8_eplb.py` | `Qwen3-235B-A22B-W8A8.yaml` |
| [#3973](https://github.com/vllm-project/vllm-ascend/pull/3973) |
`test_qwen3_30b_w8a8.py` | `Qwen3-30B-A3B-W8A8.yaml` |
| [#3541](https://github.com/vllm-project/vllm-ascend/pull/3541) |
`test_qwen3_32b_int8.py` | `Qwen3-32B-Int8.yaml` |
| [#3757](https://github.com/vllm-project/vllm-ascend/pull/3757) |
`test_qwq_32b.py` | `QwQ-32B.yaml` |
| [#5616](https://github.com/vllm-project/vllm-ascend/pull/5616) |
`test_qwen3_next_w8a8.py` | `Qwen3-Next-80B-A3B-Instruct-W8A8.yaml` |
| [#3541](https://github.com/vllm-project/vllm-ascend/pull/3541) |
`test_qwen2_5_vl_7b.py` | `Qwen2.5-VL-7B-Instruct.yaml` |
| [#5301](https://github.com/vllm-project/vllm-ascend/pull/5301) |
`test_qwen2_5_vl_7b_epd.py` | `Qwen2.5-VL-7B-Instruct-EPD.yaml` |
| [#3707](https://github.com/vllm-project/vllm-ascend/pull/3707) |
`test_qwen2_5_vl_32b.py` | `Qwen2.5-VL-32B-Instruct.yaml` |
| [#3676](https://github.com/vllm-project/vllm-ascend/pull/3676) |
`test_qwen3_32b_int8_a3_feature_stack3.py` |
`Qwen3-32B-Int8-A3-Feature-Stack3.yaml` |
| [#3709](https://github.com/vllm-project/vllm-ascend/pull/3709) |
`test_prefix_cache_qwen3_32b_int8.py` |
`Prefix-Cache-Qwen3-32B-Int8.yaml` |
| [#5395](https://github.com/vllm-project/vllm-ascend/pull/5395) |
`test_qwen3_next.py` | `Qwen3-Next-80B-A3B-Instruct-A2.yaml` |
| [#3474](https://github.com/vllm-project/vllm-ascend/pull/3474) |
`test_qwen3_32b.py` | `Qwen3-32B.yaml` |
| [#3541](https://github.com/vllm-project/vllm-ascend/pull/3541) |
`test_qwen3_32b_int8.py` | `Qwen3-32B-Int8-A2.yaml` |
### Does this PR introduce _any_ user-facing change?

### How was this patch tested?

- vLLM version: v0.15.0
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.15.0

---------

Signed-off-by: MrZ20 <2609716663@qq.com>
This commit is contained in:
SILONG ZENG
2026-03-03 20:13:43 +08:00
committed by GitHub
parent a0a904a3d4
commit 859f2c25b9
51 changed files with 2265 additions and 2336 deletions

View File

@@ -28,7 +28,10 @@ on:
type: string
default: "swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.5.1-910b-ubuntu22.04-py3.11"
tests:
required: true
required: false
type: string
config_file_path:
required: false
type: string
name:
required: false
@@ -44,12 +47,12 @@ defaults:
# only cancel in-progress runs of the same workflow
# and ignore the lint / 1 card / 4 cards test type
concurrency:
group: ascend-nightly-${{ github.workflow_ref }}-${{ github.ref }}-${{ inputs.tests }}
group: ascend-nightly-${{ github.workflow_ref }}-${{ github.ref }}-${{ inputs.config_file_path || inputs.tests }}
cancel-in-progress: true
jobs:
e2e-nightly:
name: ${{ inputs.tests }}
name: ${{ inputs.name || inputs.config_file_path || inputs.tests }}
runs-on: ${{ inputs.runner }}
timeout-minutes: 600
container:
@@ -114,14 +117,33 @@ jobs:
update-alternatives --install /usr/bin/clang clang /usr/bin/clang-15 20
update-alternatives --install /usr/bin/clang++ clang++ /usr/bin/clang++-15 20
- name: Run vllm-project/vllm-ascend test
- name: Validate Inputs
run: |
if [[ -z "${{ inputs.tests }}" && -z "${{ inputs.config_file_path }}" ]]; then
echo "Error: Either 'tests' or 'config_file_path' must be provided."
exit 1
fi
- name: Run Pytest (py-driven)
if: ${{ inputs.tests != '' }}
env:
VLLM_WORKER_MULTIPROC_METHOD: spawn
VLLM_USE_MODELSCOPE: True
VLLM_CI_RUNNER: ${{ inputs.runner }}
BENCHMARK_HOME: /vllm-workspace/vllm-ascend/benchmark
working-directory: /vllm-workspace/vllm-ascend
run: |
# ignore test_dispatch_ffn_combine until the test is fixed
pytest -sv ${{ inputs.tests }} \
echo "Running pytest with tests path: ${{ inputs.tests }}"
pytest -sv "${{ inputs.tests }}" \
--ignore=tests/e2e/nightly/single_node/ops/singlecard_ops/test_fused_moe.py
- name: Run Pytest (YAML-driven)
if: ${{ always() && inputs.config_file_path != '' }}
env:
VLLM_WORKER_MULTIPROC_METHOD: spawn
VLLM_USE_MODELSCOPE: True
VLLM_CI_RUNNER: ${{ inputs.runner }}
CONFIG_YAML_PATH: ${{ inputs.config_file_path }}
working-directory: /vllm-workspace/vllm-ascend
run: |
echo "Running YAML-driven test with config: ${{ inputs.config_file_path }}"
pytest -sv tests/e2e/nightly/single_node/models/scripts/test_single_node.py

View File

@@ -49,15 +49,6 @@ jobs:
fail-fast: false
matrix:
test_config:
- name: qwen3-next
os: linux-aarch64-a2b3-4
tests: tests/e2e/nightly/single_node/models/test_qwen3_next.py
- name: qwen3-32b
os: linux-aarch64-a2b3-4
tests: tests/e2e/nightly/single_node/models/test_qwen3_32b.py
- name: qwen3-32b-in8-a2
os: linux-aarch64-a2b3-4
tests: tests/e2e/nightly/single_node/models/test_qwen3_32b_int8.py
- name: test_custom_op
os: linux-aarch64-a2b3-1
tests: tests/e2e/nightly/single_node/ops/singlecard_ops
@@ -71,10 +62,33 @@ jobs:
name: ${{ matrix.test_config.name }}
image: 'swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/vllm-ascend:nightly-a2'
single-node-yaml-tests:
name: single-node
if: always() && (github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')
strategy:
fail-fast: false
matrix:
test_config:
- name: qwen3-32b
os: linux-aarch64-a2b3-4
config_file_path: Qwen3-32B.yaml
- name: qwen3-next-80b-a3b-instruct
os: linux-aarch64-a2b3-4
config_file_path: Qwen3-Next-80B-A3B-Instruct-A2.yaml
- name: qwen3-32b-int8
os: linux-aarch64-a2b3-4
config_file_path: Qwen3-32B-Int8-A2.yaml
uses: ./.github/workflows/_e2e_nightly_single_node.yaml
with:
runner: ${{ matrix.test_config.os }}
image: 'swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/vllm-ascend:nightly-a2'
config_file_path: ${{ matrix.test_config.config_file_path }}
name: ${{ matrix.test_config.name }}
multi-node-tests:
name: multi-node
if: always() && (github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')
needs: single-node-tests
needs: [single-node-tests, single-node-yaml-tests]
strategy:
fail-fast: false
max-parallel: 1

View File

@@ -109,65 +109,11 @@ jobs:
single-node-tests:
name: single-node
if: always() && (github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')
needs: multi-node-tests
needs: [multi-node-tests]
strategy:
fail-fast: false
matrix:
test_config:
- name: qwen3-32b-in8-a3
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwen3_32b_int8.py
- name: qwen3-32b-int8-a3-feature-stack3
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwen3_32b_int8_a3_feature_stack3.py
- name: qwen3-235b-a22b-w8a8-eplb
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_qwen3_235b_a22b_w8a8_eplb.py
- name: deepseek-r1-w8a8-eplb
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_deepseek_r1_0528_w8a8_eplb.py
- name: deepseek-r1-w8a8-mtpx
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_mtpx_deepseek_r1_0528_w8a8.py
- name: qwen2-5-vl-7b
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwen2_5_vl_7b.py
- name: qwen2-5-vl-7b-epd
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwen2_5_vl_7b_epd.py
- name: qwen2-5-vl-32b
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwen2_5_vl_32b.py
- name: qwen3-32b-int8-prefix-cache
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_prefix_cache_qwen3_32b_int8.py
- name: deepseek-r1-0528-w8a8
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_deepseek_r1_0528_w8a8.py
- name: deepseek-r1-0528-w8a8-prefix-cache
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_prefix_cache_deepseek_r1_0528_w8a8.py
- name: qwq-32b-a3
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwq_32b.py
- name: qwen3-30b-w8a8
os: linux-aarch64-a3-2
tests: tests/e2e/nightly/single_node/models/test_qwen3_30b_w8a8.py
- name: qwen3-235b-w8a8
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_qwen3_235b_w8a8.py
- name: qwen3-next-w8a8
os: linux-aarch64-a3-4
tests: tests/e2e/nightly/single_node/models/test_qwen3_next_w8a8.py
- name: kimi-k2-thinking
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_kimi_k2_thinking.py
- name: deepseek-r1-w8a8-hbm
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_deepseek_r1_w8a8_hbm.py
- name: deepseek3_2-w8a8
os: linux-aarch64-a3-16
tests: tests/e2e/nightly/single_node/models/test_deepseek_v3_2_w8a8.py
- name: qwen3-30b-acc
os: linux-aarch64-a3-4
tests: tests/e2e/weekly/single_node/models/test_qwen3_30b_acc.py
@@ -178,6 +124,70 @@ jobs:
tests: ${{ matrix.test_config.tests }}
name: ${{ matrix.test_config.name }}
single-node-yaml-tests:
name: single-node
if: always() && (github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')
needs: [multi-node-tests]
strategy:
fail-fast: false
matrix:
test_config:
# YAML-driven tests
- name: deepseek-r1-0528-w8a8
os: linux-aarch64-a3-16
config_file_path: DeepSeek-R1-0528-W8A8.yaml
- name: deepseek-r1-w8a8-hbm
os: linux-aarch64-a3-16
config_file_path: DeepSeek-R1-W8A8-HBM.yaml
- name: deepseek-v3-2-w8a8
os: linux-aarch64-a3-16
config_file_path: DeepSeek-V3.2-W8A8.yaml
- name: kimi-k2-thinking
os: linux-aarch64-a3-16
config_file_path: Kimi-K2-Thinking.yaml
- name: mtpx-deepseek-r1-0528-w8a8
os: linux-aarch64-a3-16
config_file_path: MTPX-DeepSeek-R1-0528-W8A8.yaml
- name: qwen3-235b-a22b-w8a8
os: linux-aarch64-a3-16
config_file_path: Qwen3-235B-A22B-W8A8.yaml
- name: qwen3-30b-a3b-w8a8
os: linux-aarch64-a3-4
config_file_path: Qwen3-30B-A3B-W8A8.yaml
- name: qwen3-next-80b-a3b-instruct-w8a8
os: linux-aarch64-a3-4
config_file_path: Qwen3-Next-80B-A3B-Instruct-W8A8.yaml
- name: qwq-32b
os: linux-aarch64-a3-4
config_file_path: QwQ-32B.yaml
- name: qwen3-32b-int8
os: linux-aarch64-a3-4
config_file_path: Qwen3-32B-Int8.yaml
- name: qwen2-5-vl-7b
os: linux-aarch64-a3-4
config_file_path: Qwen2.5-VL-7B-Instruct.yaml
- name: qwen2-5-vl-7b-epd
os: linux-aarch64-a3-4
config_file_path: Qwen2.5-VL-7B-Instruct-EPD.yaml
- name: qwen2-5-vl-32b
os: linux-aarch64-a3-4
config_file_path: Qwen2.5-VL-32B-Instruct.yaml
- name: qwen3-32b-int8-a3-feature-stack3
os: linux-aarch64-a3-4
config_file_path: Qwen3-32B-Int8-A3-Feature-Stack3.yaml
- name: qwen3-32b-int8-prefix-cache
os: linux-aarch64-a3-4
config_file_path: Prefix-Cache-Qwen3-32B-Int8.yaml
- name: deepseek-r1-0528-w8a8-prefix-cache
os: linux-aarch64-a3-16
config_file_path: Prefix-Cache-DeepSeek-R1-0528-W8A8.yaml
uses: ./.github/workflows/_e2e_nightly_single_node.yaml
with:
runner: ${{ matrix.test_config.os }}
image: 'swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/vllm-ascend:nightly-a3'
config_file_path: ${{ matrix.test_config.config_file_path }}
name: ${{ matrix.test_config.name }}
custom-ops-tests:
name: test ops
if: always() && (github.event_name == 'schedule' || github.event_name == 'workflow_dispatch')