LLamaSharp/.github/workflows/compile.yml

name: Update Binaries

on:
  workflow_dispatch:
    inputs:
      llama_cpp_commit:
        description: 'Branch, tag, or commit to use for llama.cpp'
        required: true
        default: 'master'
  push:
    branches: [cron_job]
  #schedule:
  #  - cron: "22 22 * * 2"
concurrency:
  group: ${{ github.workflow }}-${{ github.ref }}-${{ github.event.inputs.llama_cpp_commit }}
  cancel-in-progress: true

env:
  # Compiler defines common to all platforms
  COMMON_DEFINE: -DLLAMA_NATIVE=OFF -DLLAMA_BUILD_TESTS=OFF -DLLAMA_BUILD_EXAMPLES=ON -DLLAMA_BUILD_SERVER=OFF -DBUILD_SHARED_LIBS=ON

jobs:
  compile-linux:
    name: Compile (Linux)
    strategy:
      fail-fast: true
      matrix:
        include:
          - build: 'noavx'
            defines: '-DLLAMA_AVX=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF'
          - build: 'avx2'
            defines: ''
          - build: 'avx'
            defines: '-DLLAMA_AVX2=OFF'
          - build: 'avx512'
            defines: '-DLLAMA_AVX512=ON'
    runs-on: ubuntu-20.04
    steps:
      - uses: actions/checkout@v4
        with:
          repository: ggerganov/llama.cpp
          fetch-depth: 0
          ref: '${{ github.event.inputs.llama_cpp_commit }}'
      - name: Build
        id: cmake_build
        run: |
          mkdir build
          cd build
          cmake .. ${{ env.COMMON_DEFINE }} ${{ matrix.defines }}
          cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}
      - uses: actions/upload-artifact@v3
        with:
          path: ./build/libllama.so
          name: llama-bin-linux-${{ matrix.build }}-x64.so
      - name: Upload Llava
        uses: actions/upload-artifact@v3
        with:
          path: ./build/examples/llava/libllava_shared.so
          name: llava-bin-linux-${{ matrix.build }}-x64.so

  compile-windows:
    name: Compile (Windows)
    strategy:
      fail-fast: true
      matrix:
        include:
          - build: 'noavx'
            defines: '-DLLAMA_AVX=OFF -DLLAMA_AVX2=OFF -DLLAMA_FMA=OFF'
          - build: 'avx2'
            defines: ''
          - build: 'avx'
            defines: '-DLLAMA_AVX2=OFF'
          - build: 'avx512'
            defines: '-DLLAMA_AVX512=ON -DLLAMA_AVX512_VBMI=ON -DLLAMA_AVX512_VNNI=ON'
    runs-on: windows-latest
    steps:
      - uses: actions/checkout@v4
        with:
          repository: ggerganov/llama.cpp
          fetch-depth: 0
          ref: '${{ github.event.inputs.llama_cpp_commit }}'

      - name: Build
        id: cmake_build
        run: |
          mkdir build
          cd build
          cmake .. ${{ env.COMMON_DEFINE }} ${{ matrix.defines }}
          cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}

      - name: Upload artifacts
        uses: actions/upload-artifact@v3
        with:
          path: .\build\bin\Release\llama.dll
          name: llama-bin-win-${{ matrix.build }}-x64.dll

      - name: Upload Llava
        uses: actions/upload-artifact@v3
        with:
          path: .\build\bin\Release\llava_shared.dll
          name: llava-bin-win-${{ matrix.build }}-x64.dll

  compile-clblast:
    name: Compile (clblast) - ${{ matrix.os }}
    strategy:
      fail-fast: false
      matrix:
        os: [
          ubuntu-22.04,
          windows-latest
        ]
    env:
      OPENBLAS_VERSION: 0.3.23
      OPENCL_VERSION: 2023.04.17
      CLBLAST_VERSION: 1.6.0
      VULKAN_VERSION: 1.3.261.1
    runs-on: ${{ matrix.os }}
    steps:
      - name: Clone
        id: checkout
        uses: actions/checkout@v4
        with:
          repository: ggerganov/llama.cpp
      - name: Download dependencies - Linux
        if: ${{ matrix.os == 'ubuntu-22.04' }}
        run: |
          sudo apt update
          sudo apt install libopencl-clang-dev libclblast-dev

      - name: Download dependencies - Windows
        id: get_opencl
        if: ${{ matrix.os == 'windows-latest' }}
        run: |
          curl.exe -o $env:RUNNER_TEMP/opencl.zip -L "https://github.com/KhronosGroup/OpenCL-SDK/releases/download/v${env:OPENCL_VERSION}/OpenCL-SDK-v${env:OPENCL_VERSION}-Win-x64.zip"
          mkdir $env:RUNNER_TEMP/opencl
          tar.exe -xvf $env:RUNNER_TEMP/opencl.zip --strip-components=1 -C $env:RUNNER_TEMP/opencl
          curl.exe -o $env:RUNNER_TEMP/clblast.7z -L "https://github.com/CNugteren/CLBlast/releases/download/${env:CLBLAST_VERSION}/CLBlast-${env:CLBLAST_VERSION}-windows-x64.7z"
          curl.exe -o $env:RUNNER_TEMP/CLBlast.LICENSE.txt -L "https://github.com/CNugteren/CLBlast/raw/${env:CLBLAST_VERSION}/LICENSE"
          7z x "-o${env:RUNNER_TEMP}" $env:RUNNER_TEMP/clblast.7z
          rename-item $env:RUNNER_TEMP/CLBlast-${env:CLBLAST_VERSION}-windows-x64 clblast
          foreach ($f in (gci -Recurse -Path "$env:RUNNER_TEMP/clblast" -Filter '*.cmake')) {
            $txt = Get-Content -Path $f -Raw
            $txt.Replace('C:/vcpkg/packages/opencl_x64-windows/', "$($env:RUNNER_TEMP.Replace('\','/'))/opencl/") | Set-Content -Path $f -Encoding UTF8
          }

      - name: Build
        id: cmake_build
        if: ${{ matrix.os == 'windows-latest' }}
        run: |
          mkdir build
          cd build
          cmake .. ${{ env.COMMON_DEFINE }} -DLLAMA_CLBLAST=ON -DCMAKE_PREFIX_PATH="$env:RUNNER_TEMP/clblast"
          cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}
          copy $env:RUNNER_TEMP/clblast/lib/clblast.dll .\bin\Release\clblast.dll
          # # We should probably generate a sha256 sum in a file, and use that.
          # echo "78a8c98bcb2efe1a63318d901ab204d9ba96c3b29707b4ce0c4240bdcdc698d6  ./bin/Release/clblast.dll" >> tmp
          # sha256sum -c tmp || exit 255
          # rm tmp
          ls -R
      - name: Build
        if: ${{ matrix.os == 'ubuntu-22.04' }}
        run: |
          mkdir build
          cd build
          cmake .. ${{ env.COMMON_DEFINE }} -DLLAMA_CLBLAST=ON
          cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}
          # if we ever want to pull libclblast.so back into the packages, just uncomment this line, and the one below for the upload
          # cp $(ldconfig -p | grep libclblast.so | tail -n 1 | cut -d ' ' -f 4) ./
          ls -R
      - name: Upload artifacts (Windows)
        if: ${{ matrix.os == 'windows-latest' }}
        uses: actions/upload-artifact@v3
        with:
          path: |
            .\build\bin\Release\llama.dll
            .\build\bin\Release\clblast.dll
          name: llama-bin-win-clblast-x64.dll
      - name: Upload llava artifacts (Windows)
        if: ${{ matrix.os == 'windows-latest' }}
        uses: actions/upload-artifact@v3
        with:
          path: |
            .\build\bin\Release\llava_shared.dll
          name: llava-bin-win-clblast-x64.dll
      - name: Upload artifacts (linux)
        if: ${{ matrix.os == 'ubuntu-22.04' }}
        uses: actions/upload-artifact@v3
        with:
          path: |
            ./build/libllama.so
            # ./build/libclblast.so
          name: llama-bin-linux-clblast-x64.so
      - name: Upload llava artifacts (linux)
        if: ${{ matrix.os == 'ubuntu-22.04' }}
        uses: actions/upload-artifact@v3
        with:
          path: |
            ./build/examples/llava/libllava_shared.so
          name: llava-bin-linux-clblast-x64.so

          
  compile-cublas:
    name: Compile (cublas)
    strategy:
      fail-fast: false
      matrix:
        os: [ubuntu-20.04, windows-latest]
        cuda: ['12.1.0', '11.7.1']
    runs-on: ${{ matrix.os }}
    steps:
      - name: Clone
        id: checkout
        uses: actions/checkout@v4
        with:
          repository: ggerganov/llama.cpp
          fetch-depth: 0
          ref: '${{ github.event.inputs.llama_cpp_commit }}'

      - uses: Jimver/cuda-toolkit@v0.2.14
        if: runner.os == 'Windows'
        id: cuda-toolkit-windows
        with:
          cuda: ${{ matrix.cuda }}
          method: 'network'
          sub-packages: '["nvcc", "cudart", "cublas", "cublas_dev", "thrust", "visual_studio_integration"]'

      - uses: Jimver/cuda-toolkit@v0.2.14
        if: runner.os == 'Linux'
        id: cuda-toolkit-linux
        with:
          cuda: ${{ matrix.cuda }}
          method: 'network'
          linux-local-args: '["--toolkit"]'

      - name: Build
        id: cmake_build
        run: |
          mkdir build
          cd build
          cmake .. ${{ env.COMMON_DEFINE }} -DLLAMA_CUBLAS=ON
          cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}
          ls -R

      - name: Upload artifacts (Windows)
        if: ${{ matrix.os == 'windows-latest' }}
        uses: actions/upload-artifact@v3
        with:
          path: .\build\bin\Release\llama.dll
          name: llama-bin-win-cublas-cu${{ matrix.cuda }}-x64.dll
      - name: Upload llava artifacts (Windows)
        if: ${{ matrix.os == 'windows-latest' }}
        uses: actions/upload-artifact@v3
        with:
          path: .\build\bin\Release\llava_shared.dll
          name: llava-bin-win-cublas-cu${{ matrix.cuda }}-x64.dll
      - name: Upload artifacts (Linux)
        if: ${{ matrix.os == 'ubuntu-20.04' }}
        uses: actions/upload-artifact@v3
        with:
          path: ./build/libllama.so
          name: llama-bin-linux-cublas-cu${{ matrix.cuda }}-x64.so
      - name: Upload llava artifacts (Linux)
        if: ${{ matrix.os == 'ubuntu-20.04' }}
        uses: actions/upload-artifact@v3
        with:
          path: ./build/examples/llava/libllava_shared.so
          name: llava-bin-linux-cublas-cu${{ matrix.cuda }}-x64.so
 
    
  compile-macos:
    name: Compile (MacOS)
    strategy:
      fail-fast: true
      matrix:
        include:
          - build: 'arm64'
            defines: '-DCMAKE_OSX_ARCHITECTURES=arm64'
          - build: 'x64'
            defines: '-DCMAKE_OSX_ARCHITECTURES=x86_64 -DLLAMA_METAL=OFF -DLLAMA_AVX=ON -DLLAMA_AVX2=ON'
    runs-on: macos-latest   
    steps:
      - uses: actions/checkout@v4
        with:
          repository: ggerganov/llama.cpp
          fetch-depth: 0
          ref: '${{ github.event.inputs.llama_cpp_commit }}'
      - name: Dependencies
        continue-on-error: true
        run: |
          brew update
      - name: Build
        id: cmake_build
        run: |
          mkdir build
          cd build
          cmake .. ${{ env.COMMON_DEFINE }} ${{ matrix.defines }}
          cmake --build . --config Release -j ${env:NUMBER_OF_PROCESSORS}
      - name: Upload artifacts
        uses: actions/upload-artifact@v3
        with:
          path: ./build/libllama.dylib
          name: llama-bin-osx-${{ matrix.build }}.dylib
      - name: Upload Llava
        uses: actions/upload-artifact@v3
        with:
          path: ./build/examples/llava/libllava_shared.dylib
          name: llava-bin-osx-${{ matrix.build }}.dylib
      - name: Upload Metal
        if: ${{ matrix.build != 'x64' }}
        uses: actions/upload-artifact@v3
        with:
          path: ./build/bin/ggml-metal.metal
          name: ggml-metal.metal
  
  build-deps:
    runs-on: ubuntu-latest
    name: "Gather Binaries"
    if: ${{ always() }}
    needs: [
      "compile-linux",
      "compile-macos",
      "compile-windows",
      "compile-cublas",
      "compile-clblast"
    ]
    steps:      
      - uses: actions/download-artifact@v4
        with:
          path: artifacts
      - name: List Files
        run: ls -R
      - name: Rearrange Files
        run: |
          # Make all directories at once
          mkdir --parents deps/{avx,avx2,avx512,osx-arm64,osx-x64,cu11.7.1,cu12.1.0,clblast}

          cp artifacts/llama-bin-linux-noavx-x64.so/libllama.so  deps/libllama.so
          cp artifacts/llama-bin-linux-avx-x64.so/libllama.so    deps/avx/libllama.so
          cp artifacts/llama-bin-linux-avx2-x64.so/libllama.so   deps/avx2/libllama.so
          cp artifacts/llama-bin-linux-avx512-x64.so/libllama.so deps/avx512/libllama.so

          cp artifacts/llama-bin-win-noavx-x64.dll/llama.dll  deps/llama.dll
          cp artifacts/llama-bin-win-avx-x64.dll/llama.dll    deps/avx/llama.dll
          cp artifacts/llama-bin-win-avx2-x64.dll/llama.dll   deps/avx2/llama.dll
          cp artifacts/llama-bin-win-avx512-x64.dll/llama.dll deps/avx512/llama.dll

          cp artifacts/llava-bin-linux-noavx-x64.so/libllava_shared.so  deps/libllava_shared.so
          cp artifacts/llava-bin-linux-avx-x64.so/libllava_shared.so    deps/avx/libllava_shared.so
          cp artifacts/llava-bin-linux-avx2-x64.so/libllava_shared.so   deps/avx2/libllava_shared.so
          cp artifacts/llava-bin-linux-avx512-x64.so/libllava_shared.so deps/avx512/libllava_shared.so

          cp artifacts/llama-bin-win-noavx-x64.dll/llama.dll  deps/libllama.dll
          cp artifacts/llama-bin-win-avx-x64.dll/llama.dll    deps/avx/libllama.dll
          cp artifacts/llama-bin-win-avx2-x64.dll/llama.dll   deps/avx2/libllama.dll
          cp artifacts/llama-bin-win-avx512-x64.dll/llama.dll deps/avx512/libllama.dll

          cp artifacts/llava-bin-win-noavx-x64.dll/llava_shared.dll  deps/llava_shared.dll
          cp artifacts/llava-bin-win-avx-x64.dll/llava_shared.dll    deps/avx/llava_shared.dll
          cp artifacts/llava-bin-win-avx2-x64.dll/llava_shared.dll   deps/avx2/llava_shared.dll
          cp artifacts/llava-bin-win-avx512-x64.dll/llava_shared.dll deps/avx512/llava_shared.dll


          cp artifacts/llama-bin-osx-arm64.dylib/libllama.dylib deps/osx-arm64/libllama.dylib
          cp artifacts/llava-bin-osx-arm64.dylib/libllava_shared.dylib deps/osx-arm64/libllava_shared.dylib
          cp artifacts/ggml-metal.metal/ggml-metal.metal        deps/osx-arm64/ggml-metal.metal
          cp artifacts/llama-bin-osx-x64.dylib/libllama.dylib   deps/osx-x64/libllama.dylib
          cp artifacts/llava-bin-osx-x64.dylib/libllava_shared.dylib deps/osx-x64/libllava_shared.dylib

          cp artifacts/llama-bin-win-cublas-cu11.7.1-x64.dll/llama.dll    deps/cu11.7.1/llama.dll
          cp artifacts/llava-bin-win-cublas-cu11.7.1-x64.dll/llava_shared.dll    deps/cu11.7.1/llava_shared.dll

          cp artifacts/llama-bin-linux-cublas-cu11.7.1-x64.so/libllama.so deps/cu11.7.1/libllama.so
          cp artifacts/llava-bin-linux-cublas-cu11.7.1-x64.so/libllava_shared.so deps/cu11.7.1/libllama_shared.so
          
          cp artifacts/llama-bin-win-cublas-cu12.1.0-x64.dll/llama.dll    deps/cu12.1.0/llama.dll
          cp artifacts/llava-bin-win-cublas-cu12.1.0-x64.dll/llava_shared.dll    deps/cu12.1.0/llava_shared.dll

          cp artifacts/llama-bin-linux-cublas-cu12.1.0-x64.so/libllama.so deps/cu12.1.0/libllama.so
          cp artifacts/llava-bin-linux-cublas-cu12.1.0-x64.so/libllava_shared.so deps/cu12.1.0/libllava_shared.so
          
          cp artifacts/llama-bin-win-clblast-x64.dll/{llama,clblast}.dll deps/clblast/
          
          cp artifacts/llama-bin-linux-clblast-x64.so/libllama.so deps/clblast/


      - name: Upload artifacts
        uses: actions/upload-artifact@v3
        with:
          path: deps/
          name: deps


      - name: Remove Artifacts
        uses: geekyeggo/delete-artifact@v2
        with:
          name: |
            llama-*
            llava-*
            *.metal