[Test] Refactor accuracy test to nightly test

Signed-off-by: hfadzxy <starmoon_zhang@163.com>

Author: zhangxinyuehfad

.github/workflows/_accuracy_test.yaml renamed to .github/workflows/_e2e_nightly_single_node_accuracy.yaml

@@ -1,4 +1,21 @@
-name: 'accuracy test'
+#
+# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+# This file is a part of the vllm-ascend project.
+#
+
+name: 'e2e nightly accuracy test'
 
 on:
   workflow_call:
@@ -24,38 +41,44 @@ on:
         type: boolean
         default: false
 
-jobs:
-  accuracy_tests:
+# 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.
+# It's used to activate ascend-toolkit environment variables.
+defaults:
+  run:
+    shell: bash -el {0}
 
-    runs-on: ${{ inputs.runner }}
+# only cancel in-progress runs of the same workflow
+# and ignore the lint / 1 card / 2 cards / 4 cards test type
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  e2e-nightly:
     name: ${{ inputs.model_name }} accuracy
+    runs-on: ${{ inputs.runner }}
     container:
-      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-910b-ubuntu22.04-py3.11
+      image: ${{ inputs.image }}
       env:
         VLLM_USE_MODELSCOPE: True
-        # 1. If version specified (work_dispatch), do specified branch accuracy test
-        # 2. If no version (labeled PR), do accuracy test by default ref:
-        # The branch, tag or SHA to checkout. When checking out the repository that
-        # triggered a workflow, this defaults to the reference or SHA for that event.
-        # Otherwise, uses the default branch.
         GHA_VLLM_ASCEND_VERSION: ${{ inputs.vllm-ascend }}
-
     steps:
-      - name: Checkout repository
-        uses: actions/checkout@v4
-
-      - name: Set model name as output
-        id: set_output
+      - name: Check npu and CANN info
         run: |
-          echo "model_name=${{ inputs.model_name }}" >> $GITHUB_OUTPUT
+          npu-smi info
+          cat /usr/local/Ascend/ascend-toolkit/latest/"$(uname -i)"-linux/ascend_toolkit_install.info
 
       - name: Config mirrors
         run: |
-          sed -Ei 's@(ports|archive).ubuntu.com@cache-service.nginx-pypi-cache.svc.cluster.local:8081@g' /etc/apt/sources.list
-          pip config set global.index-url http://cache-service.nginx-pypi-cache.svc.cluster.local/pypi/simple
-          pip config set global.trusted-host cache-service.nginx-pypi-cache.svc.cluster.local
+          sed -i 's|ports.ubuntu.com|mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list
+          pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple
           apt-get update -y
           apt install git -y
+          git config --global url."https://gh-proxy.test.osinfra.cn/https://github.com/".insteadOf https://github.com/
+
+      - name: Checkout vllm-project/vllm-ascend repo
+        uses: actions/checkout@v4
 
       - name: Install system dependencies
         run: |
@@ -73,7 +96,14 @@ jobs:
         working-directory: ./vllm-empty
         run: |
           VLLM_TARGET_DEVICE=empty pip install -e .
-        
+
+      - name: Install vllm-project/vllm-ascend
+        env:
+          PIP_EXTRA_INDEX_URL: https://mirrors.huaweicloud.com/ascend/repos/pypi
+        run: |
+          pip install -r requirements-dev.txt
+          pip install -v -e .
+
       - name: Install Ascend toolkit & triton_ascend (for Qwen3-Next-80B-A3B-Instruct)
         if: ${{ inputs.model_name == 'Qwen3-Next-80B-A3B-Instruct' }}
         shell: bash -l {0}
@@ -149,11 +179,12 @@ jobs:
             pip show vllm | grep "Version:" | awk '{print "GHA_VLLM_VERSION="$2}' | sed 's/+.*//'
           } >> "$GITHUB_ENV"
 
-      - name: Run accuracy test
+      - name: Run vllm-project/vllm-ascend accuracy test
         id: report
         env:
           VLLM_WORKER_MULTIPROC_METHOD: spawn
           VLLM_USE_MODELSCOPE: True
+          VLLM_CI_RUNNER: ${{ inputs.runner }}
           VLLM_VERSION: ${{ env.GHA_VLLM_VERSION }}
           VLLM_COMMIT: ${{ env.VLLM_COMMIT }}
           VLLM_ASCEND_VERSION: ${{ env.GHA_VLLM_ASCEND_VERSION || github.ref }}
@@ -182,4 +213,4 @@ jobs:
           path: ./benchmarks/accuracy/${{ steps.report.outputs.markdown_name }}.md
           if-no-files-found: warn
           retention-days: 90
-          overwrite: true
+          overwrite: true

.github/workflows/accuracy_test.yaml

@@ -23,9 +23,6 @@
 name: ascend test / models
 
 on:
-  schedule:
-    # Runs every 6 hours
-    - cron:  '0 */6 * * *'
   pull_request:
     branches:
       - 'main'
@@ -75,7 +72,7 @@ jobs:
           - model_name: DeepSeek-V2-Lite
             runner: a2-2
       fail-fast: false
-    uses: ./.github/workflows/_accuracy_test.yaml
+    uses: ./.github/workflows/_e2e_nightly_single_node_accuracy.yaml
     with:
       vllm: v0.11.0
       runner:  linux-aarch64-${{ matrix.runner }}

.github/workflows/vllm_ascend_test_models.yaml

@@ -27,6 +27,10 @@ on:
   pull_request: 
     branches:
       - 'main'
+  push:
+    branches:
+      - 'main'
+      - '*-dev'
 
 # 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.
@@ -83,3 +87,30 @@ jobs:
       config_file_path: ${{ matrix.test_config.config_file_path }}
     secrets:
       KUBECONFIG_B64: ${{ secrets.KUBECONFIG_A2_B64 }}
+
+  single-node-accuracy-tests:
+    strategy:
+      fail-fast: false
+      matrix:
+        test_config:
+          - name: Qwen3-8B
+            os: linux-aarch64-a2-1
+          - name: Qwen2.5-VL-7B-Instruct
+            os: linux-aarch64-a2-1
+          - name: Qwen2-Audio-7B-Instruct
+            os: linux-aarch64-a2-1
+          - name: Qwen3-30B-A3B
+            os: linux-aarch64-a2-2
+          - name: Qwen3-VL-30B-A3B-Instruct
+            os: linux-aarch64-a2-2
+          - name: DeepSeek-V2-Lite
+            os: linux-aarch64-a2-2
+          - name: Qwen3-Next-80B-A3B-Instruct
+            os: linux-aarch64-a2-4
+    uses: ./.github/workflows/_e2e_nightly_single_node_accuracy.yaml
+    with:
+      vllm: v0.11.0
+      runner: ${{ matrix.test_config.os }}
+      model_name: ${{ matrix.test_config.name }}
+      image: swr.cn-southwest-2.myhuaweicloud.com/base_image/ascend-ci/cann:8.2.rc1-910b-ubuntu22.04-py3.11
+      upload: false