# This workflow will:
# - Create a new Github release
# - Build wheels for supported architectures
# - Deploy the wheels to the Github release
# - Release the static code to PyPi
# For more information see: https://help.github.com/en/actions/language-and-framework-guides/using-python-with-github-actions#publishing-to-package-registries

name: Build wheels and deploy

on:
  create:
    tags:
      - '**'

jobs:
  setup_release:
    name: Create Release
    runs-on: ubuntu-latest
    steps:
      - name: Get the tag version
        id: extract_branch
        run: echo ::set-output name=branch::${GITHUB_REF#refs/tags/}
        shell: bash

      - name: Create Release
        id: create_release
        uses: actions/create-release@v1
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        with:
          tag_name: ${{ steps.extract_branch.outputs.branch }}
          release_name: ${{ steps.extract_branch.outputs.branch }}

  build_wheels:
    name: Build Wheel
    runs-on: ${{ matrix.os }}
    needs: setup_release

    strategy:
      fail-fast: false
      matrix:
          os: [ubuntu-20.04, ubuntu-22.04]
          python-version: ['3.7', '3.8', '3.9', '3.10']
          torch-version: ['1.11.0', '1.12.0', '1.13.0']
          cuda-version: ['113', '116', '117']
          exclude:
            # Nvidia only supports 11.7+ for ubuntu-22.04
            - os: ubuntu-22.04
              cuda-version: '116'
            - os: ubuntu-22.04
              cuda-version: '113'
            # Torch only builds cuda 117 for 1.13.0+
            - cuda-version: '117'
              torch-version: '1.11.0'
            - cuda-version: '117'
              torch-version: '1.12.0'
            # Torch only builds cuda 116 for 1.12.0+
            - cuda-version: '116'
              torch-version: '1.11.0'
            # 1.13.0 drops support for cuda 11.3
            - cuda-version: '113'
              torch-version: '1.13.0'
            # Fails with "Validation Error" on artifact upload
            - cuda-version: '117'
              torch-version: '1.13.0'
              os: ubuntu-20.04

    steps:
      - name: Checkout
        uses: actions/checkout@v3
      
      - name: Set up Python
        uses: actions/setup-python@v3
        with:
          python-version: ${{ matrix.python-version }}

      - name: Set up Linux Env
        if: ${{ runner.os == 'Linux' }}
        run: |
          sudo rm -rf /usr/share/dotnet
          bash .github/workflows/env.sh
          echo ${{ needs.create_release.outputs.upload_url }} 
          echo ${{ needs.steps.extract_branch.outputs.upload_url }} 
        shell:
          bash

      - name: Install CUDA ${{ matrix.cuda-version }}
        if: ${{ matrix.cuda-version != 'cpu' }}
        run: |
          bash .github/workflows/cuda/cu${{ matrix.cuda-version }}-${{ runner.os }}.sh
        shell:
          bash

      - name: Check GPU Env
        if: ${{ matrix.cuda-version != 'cpu' }}
        run: |
          source .github/workflows/cuda/cu${{ matrix.cuda-version }}-${{ runner.os }}-env.sh
          nvcc --version
        shell:
          bash

      - name: Install PyTorch ${{ matrix.torch-version }}+cu${{ matrix.cuda-version }}
        run: |
          pip install numpy pyyaml scipy ipython mkl mkl-include ninja cython typing pandas typing-extensions dataclasses && conda clean -ya
          pip install --no-index --no-cache-dir torch==${{ matrix.torch-version }} -f https://download.pytorch.org/whl/cu${{ matrix.cuda-version }}/torch_stable.html
          python --version
          python -c "import torch; print('PyTorch:', torch.__version__)"
          python -c "import torch; print('CUDA:', torch.version.cuda)"
          python -c "from torch.utils import cpp_extension; print (cpp_extension.CUDA_HOME)"
        shell:
          bash
      
      - name: Get the tag version
        id: extract_branch
        run: echo ::set-output name=branch::${GITHUB_REF#refs/tags/}

      - name: Get Release with tag
        id: get_current_release
        uses: joutvhu/get-release@v1
        with:
          tag_name: ${{ steps.extract_branch.outputs.branch }}
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

      - name: Build wheel
        run: |
          export FLASH_ATTENTION_FORCE_BUILD="TRUE"
          export FORCE_CUDA="1"
          export PATH=/usr/local/nvidia/bin:/usr/local/nvidia/lib64:$PATH
          export LD_LIBRARY_PATH=/usr/local/nvidia/lib64:/usr/local/cuda/lib64:$LD_LIBRARY_PATH
          export CUDA_INSTALL_DIR=/usr/local/cuda-11.3$CUDA_INSTALL_DIR
          pip install ninja packaging setuptools wheel
          python setup.py bdist_wheel --dist-dir=dist
          tmpname=cu${{ matrix.cuda-version }}torch${{ matrix.torch-version }}
          wheel_name=$(ls dist/*whl | xargs -n 1 basename | sed "s/-/+$tmpname-/2")
          ls dist/*whl |xargs -I {} mv {} dist/${wheel_name}
          echo "wheel_name=${wheel_name}" >> $GITHUB_ENV

      - name: Log Built Wheels
        run: |
          ls dist

      - name: Upload Release Asset
        id: upload_release_asset 
        uses: actions/upload-release-asset@v1
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        with:
          upload_url: ${{ steps.get_current_release.outputs.upload_url }}
          asset_path: ./dist/${{env.wheel_name}}
          asset_name: ${{env.wheel_name}}
          asset_content_type: application/*

  publish_package:
    name: Publish package
    needs: [build_wheels]

    runs-on: ubuntu-latest

    steps:
      - uses: actions/checkout@v3

      - uses: actions/setup-python@v4
        with:
          python-version: '3.10'

      - name: Install dependencies
        run: |
          pip install ninja packaging setuptools wheel twine
          pip install torch

      - name: Build core package
        env:
          FLASH_ATTENTION_SKIP_CUDA_BUILD: "TRUE"
        run: |
          python setup.py sdist --dist-dir=dist

      - name: Deploy
        env:
          TWINE_USERNAME: ${{ secrets.PYPI_USERNAME }}
          TWINE_PASSWORD: ${{ secrets.PYPI_TOKEN }}
        run: |
          python -m twine upload dist/*