[TEST]Add initial multi modal cases for nightly test and deepseek-r1 tests (#3631)

### What this PR does / why we need it?
This PR adds the initial multi modal model for nightly test, including 3
cases for Qwen2.5-vl-7b acc/perf test on A3, we need test them daily. It
also inclues 8 cases for deepseek-r1-0528-w8a8 func, acc and perf tests
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
by running the test


- vLLM version: v0.11.0rc3
- vLLM main: https://github.com/vllm-project/vllm/commit/v0.11.0

---------

Signed-off-by: jiangyunfan1 <jiangyunfan1@h-partners.com>
This commit is contained in:
jiangyunfan1
2025-10-23 17:18:49 +08:00
committed by GitHub
parent 427b17e2da
commit 9434f24ded
4 changed files with 192 additions and 25 deletions

View File

@@ -26,7 +26,7 @@ on:
branches:
- 'main'
- '*-dev'
types: [labeled]
types: [labeled,opened,synchronize]
# Bash shells do not use ~/.profile or ~/.bashrc so these shells need to be explicitly
# declared as "shell: bash -el {0}" on steps that need to be properly activated.
@@ -80,10 +80,7 @@ jobs:
if: contains(github.event.pull_request.labels.*.name, 'run-nightly')
strategy:
matrix:
# should add A3 chip runner when available
os: [ linux-aarch64-a3-16 ]
# Note (yikun): If CI resource are limited we can split job into two chain jobs
# only trigger e2e test after lint passed and the change is e2e related with pull request.
uses: ./.github/workflows/_e2e_nightly.yaml
with:
vllm: v0.11.0
@@ -94,15 +91,32 @@ jobs:
if: contains(github.event.pull_request.labels.*.name, 'run-nightly')
strategy:
matrix:
# should add A3 chip runner when available
os: [ linux-aarch64-a3-16 ]
# Note (yikun): If CI resource are limited we can split job into two chain jobs
# only trigger e2e test after lint passed and the change is e2e related with pull request.
uses: ./.github/workflows/_e2e_nightly.yaml
with:
vllm: v0.11.0
runner: ${{ matrix.os }}
image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-a3-ubuntu22.04-py3.11
tests: tests/e2e/nightly/models/test_deepseek_r1_w8a8_eplb.py
qwen2-5-vl-7b:
if: contains(github.event.pull_request.labels.*.name, 'run-nightly')
strategy:
matrix:
os: [ linux-aarch64-a3-4 ]
uses: ./.github/workflows/_e2e_nightly.yaml
with:
vllm: v0.11.0
runner: ${{ matrix.os }}
image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-a3-ubuntu22.04-py3.11
tests: tests/e2e/nightly/models/test_qwen2_5_vl_7b.py
deepseek-r1-0528-w8a8:
if: contains(github.event.pull_request.labels.*.name, 'run-nightly')
strategy:
matrix:
os: [ linux-aarch64-a3-16 ]
uses: ./.github/workflows/_e2e_nightly.yaml
with:
vllm: v0.11.0
runner: ${{ matrix.os }}
image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-a3-ubuntu22.04-py3.11
tests: tests/e2e/nightly/models/test_deepseek_r1_0528_w8a8.py