xc-llm-ascend/.github/workflows/vllm_ascend_test_long_term.yaml

#
# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
# This file is a part of the vllm-ascend project.
#
# 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.
#
name: 'e2e test / long-term-test'

on:
  schedule:
    # Runs at 23:00 UTC (7:00 AM Beijing) every day
    - cron: '0 23 * * *'
  pull_request:
    types: [ labeled ]

# 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}

# only cancel in-progress runs of the same workflow
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

jobs:
  long-term-test:
    # long-term-test will be triggered when tag 'long-term-test' & 'ready-for-test' or schedule job
    if: ${{ contains(github.event.pull_request.labels.*.name, 'long-term-test')  && contains(github.event.pull_request.labels.*.name, 'ready-for-test') || github.event_name == 'schedule' }}
    strategy:
      max-parallel: 2
      matrix:
        os: [linux-arm64-npu-1, linux-arm64-npu-4]
        vllm_version: [main, v0.9.1]
    name: vLLM Ascend long term test
    runs-on: ${{ matrix.os }}
    container:
      # TODO(yikun): Remove m.daocloud.io prefix when infra proxy ready
      image: m.daocloud.io/quay.io/ascend/cann:8.1.rc1-910b-ubuntu22.04-py3.10
      env:
        HF_ENDPOINT: https://hf-mirror.com
        HF_TOKEN: ${{ secrets.HF_TOKEN }}
        VLLM_LOGGING_LEVEL: ERROR
    steps:
      - name: Check npu and CANN info
        run: |
          npu-smi info
          cat /usr/local/Ascend/ascend-toolkit/latest/"$(uname -i)"-linux/ascend_toolkit_install.info

      - name: Config mirrors
        run: |
          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: |
          apt-get -y install `cat packages.txt`
          apt-get -y install gcc g++ cmake libnuma-dev

      - name: Checkout vllm-project/vllm repo
        uses: actions/checkout@v4
        with:
          repository: vllm-project/vllm
          ref: ${{ matrix.vllm_version }}
          path: ./vllm-empty

      - name: Install vllm-project/vllm from source
        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: Run vllm-project/vllm-ascend long term test
        run: |
          if [[ "${{ matrix.os }}" == "linux-arm64-npu-1" ]]; then
            # v0 spec decode test
            VLLM_USE_MODELSCOPE=True pytest -sv tests/e2e/long_term/spec_decode_v0/e2e/test_mtp_correctness.py  # it needs a clean process
            pytest -sv tests/e2e/long_term/spec_decode_v0 --ignore=tests/e2e/long_term/spec_decode_v0/e2e/test_mtp_correctness.py
            # v1 spec decode test
            VLLM_USE_MODELSCOPE=True pytest -sv tests/e2e/long_term/spec_decode_v1/test_v1_mtp_correctness.py
            # TODO: revert me when test_v1_spec_decode.py::test_ngram_correctness is fixed
            VLLM_USE_MODELSCOPE=True pytest -sv tests/e2e/long_term/spec_decode_v1/test_v1_spec_decode.py
            # accuracy test single card
            pytest -sv tests/e2e/long_term/test_accuracy.py
          else
            # accuracy test multi card
            VLLM_USE_MODELSCOPE=True pytest -sv tests/e2e/long_term/test_deepseek_v2_lite_tp2_accuracy.py
          fi