huawei-ascend-npu.yml

name: huawei-ascend-npu

on:
  workflow_dispatch:
  pull_request:
    paths:
      - '.github/workflows/huawei-ascend-npu.yml'
      - 'requirements/**'
  schedule:
    - cron: "0 0 * * *"
  
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}
  cancel-in-progress: true

permissions:
  contents: read
  issues: write

jobs:
  unit-tests:
    runs-on: [self-hosted, ascend, npu]
    container:
      image: ascendai/cann
      ports:
        - 80
      options: --device /dev/davinci4
               --device /dev/davinci_manager
               --device /dev/devmm_svm
               --device /dev/hisi_hdc
               --shm-size "20g"
               -v /usr/local/dcmi:/usr/local/dcmi
               -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi 
               -v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/
               -v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info
               -v /etc/ascend_install.info:/etc/ascend_install.info      

    steps:
    - uses: actions/checkout@v4

    - name: Install pytorch 
      run: |
        npu-smi info
        apt-get update
        echo "export LD_LIBRARY_PATH=/usr/local/Ascend/driver/lib64/common:/usr/local/Ascend/driver/lib64/driver:${LD_LIBRARY_PATH}" >>~/.bashrc
        echo "source /usr/local/Ascend/ascend-toolkit/set_env.sh" >>~/.bashrc
        cat ~/.bashrc

        pip install torch==2.2.0 torchvision==0.17.0 torch_npu==2.2.0 numpy==1.26.4 cloudpickle tornado -i https://pypi.tuna.tsinghua.edu.cn/simple
        python -c "import torch; print('torch:', torch.__version__)"
        python -c "import torch,torch_npu; print('torch_npu:', torch.npu.is_available(),',version:', torch_npu.__version__)"

    - name: Install deepspeed
      run: |
        pip install . -i https://pypi.tuna.tsinghua.edu.cn/simple
        ds_report

    - name: Python environment
      run: |
        pip list

    - name: Multi-card parallel training
      env:
        GITHUB_TOKEN: ${{ secrets.DEEPSPEED_TOKEN }}
      run: |
        echo "y" | apt-get install git
        git clone https://github.com/microsoft/DeepSpeedExamples.git
        cd DeepSpeedExamples
        deepspeed ./training/cifar/cifar10_deepspeed.py