2025-12-13 05:48:40 +00:00
8 changed files with 173 additions and 292 deletions
--- a/.github/workflows/backends.yml
+++ b/.github/workflows/backends.yml
@ -13,9 +13,12 @@ on:
        type: string

 jobs:
-  windows-cpu:
+  windows:
    runs-on: windows-latest
    
+    env:
+      VULKAN_VERSION: 1.3.261.1
+    
    strategy:
      matrix:
        include:
@ -27,16 +30,38 @@ jobs:
            defines: '-DGGML_NATIVE=OFF -DGGML_AVX2=ON -DSD_BUILD_SHARED_LIBS=ON'
          - build: 'avx512'
            defines: '-DGGML_NATIVE=OFF -DGGML_AVX512=ON -DSD_BUILD_SHARED_LIBS=ON'
+          - build: 'cuda12'
+            defines: '-DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61;89" -DSD_BUILD_SHARED_LIBS=ON'
+          - build: 'vulkan'
+            defines: "-DSD_VULKAN=ON -DSD_BUILD_SHARED_LIBS=ON"

    steps:
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
          submodules: recursive

+      - name: Install cuda-toolkit
+        id: cuda-toolkit-12
+        if: ${{ matrix.build == 'cuda12' }}
+        uses: Jimver/cuda-toolkit@v0.2.19
+        with:
+          cuda: '12.6.2'
+          method: network
+          sub-packages: '["nvcc", "cudart", "cublas", "cublas_dev", "thrust", "visual_studio_integration"]'
+
+      - name: Install Vulkan SDK
+        id: get_vulkan
+        if: ${{ matrix.build == 'vulkan' }}
+        run: |
+          curl.exe -o $env:RUNNER_TEMP/VulkanSDK-Installer.exe -L "https://sdk.lunarg.com/sdk/download/${env:VULKAN_VERSION}/windows/VulkanSDK-${env:VULKAN_VERSION}-Installer.exe"
+          & "$env:RUNNER_TEMP\VulkanSDK-Installer.exe" --accept-licenses --default-answer --confirm-command install
+          Add-Content $env:GITHUB_ENV "VULKAN_SDK=C:\VulkanSDK\${env:VULKAN_VERSION}"
+          Add-Content $env:GITHUB_PATH "C:\VulkanSDK\${env:VULKAN_VERSION}\bin"
+
      - name: Build
        id: cmake_build
        run: |
@ -47,129 +72,56 @@ jobs:
      
      - name: Upload artifact
        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
          name: windows-${{ matrix.build }}
          path: .\build\bin\Release\stable-diffusion.dll

-  windows-cuda12:
-    runs-on: windows-latest
-
-    steps:
-      - name: Checkout
-        id: checkout
-        uses: actions/checkout@v4.2.2
-        with:
-          repository: 'leejet/stable-diffusion.cpp'
-          ref: '${{ github.event.inputs.commit }}'
-          submodules: recursive
-
-      - name:  Install Cuda Toolkit 12.8
-        id: cuda-12
-        run: |
-          mkdir -p "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8"
-          choco install unzip -y
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_cudart/windows-x86_64/cuda_cudart-windows-x86_64-12.8.57-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvcc/windows-x86_64/cuda_nvcc-windows-x86_64-12.8.61-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvrtc/windows-x86_64/cuda_nvrtc-windows-x86_64-12.8.61-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/libcublas/windows-x86_64/libcublas-windows-x86_64-12.8.3.14-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvtx/windows-x86_64/cuda_nvtx-windows-x86_64-12.8.55-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_profiler_api/windows-x86_64/cuda_profiler_api-windows-x86_64-12.8.55-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/visual_studio_integration/windows-x86_64/visual_studio_integration-windows-x86_64-12.8.55-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvprof/windows-x86_64/cuda_nvprof-windows-x86_64-12.8.57-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_cccl/windows-x86_64/cuda_cccl-windows-x86_64-12.8.55-archive.zip"
-          unzip '*.zip' -d "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8"
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_cudart-windows-x86_64-12.8.57-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_nvcc-windows-x86_64-12.8.61-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_nvrtc-windows-x86_64-12.8.61-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\libcublas-windows-x86_64-12.8.3.14-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_nvtx-windows-x86_64-12.8.55-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_profiler_api-windows-x86_64-12.8.55-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\visual_studio_integration-windows-x86_64-12.8.55-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_nvprof-windows-x86_64-12.8.57-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\cuda_cccl-windows-x86_64-12.8.55-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\visual_studio_integration\MSBuildExtensions\*" "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\MSBuild\Microsoft\VC\v170\BuildCustomizations" /E /I /H /Y
-          echo "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
-          echo "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\libnvvp" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
-          echo "CUDA_PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" | Out-File -FilePath $env:GITHUB_ENV -Append -Encoding utf8
-          echo "CUDA_PATH_V12_8=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" | Out-File -FilePath $env:GITHUB_ENV -Append -Encoding utf8
-
-      - name: Build
-        id: cmake_build
-        shell: cmd
-        run: |
-          mkdir build
-          cd build
-          call "C:\Program Files\Microsoft Visual Studio\2022\Enterprise\VC\Auxiliary\Build\vcvars64.bat"
-          cmake .. -DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61;75;86;89;100" -DSD_BUILD_SHARED_LIBS=ON
-          cmake --build . --config Release
-      
-      - name: Upload artifact
-        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
-        with:
-          name: windows-cuda12
-          path: .\build\bin\Release\stable-diffusion.dll
-
-  windows-cuda11:
+  windows-2019:
    runs-on: windows-2019
+        
+    strategy:
+      matrix:
+        include:
+          - build: 'cuda11'
+            defines: '-DSD_CUDA=ON -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_CUDA_ARCHITECTURES="61" -DCMAKE_CUDA_FLAGS="-allow-unsupported-compiler"'

    steps:
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
          submodules: recursive

-      - name:  Install Cuda Toolkit 11.8
-        id: cuda-11
-        run: |
-          mkdir -p "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8"
-          choco install unzip -y
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_cudart/windows-x86_64/cuda_cudart-windows-x86_64-11.8.89-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvcc/windows-x86_64/cuda_nvcc-windows-x86_64-11.8.89-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvrtc/windows-x86_64/cuda_nvrtc-windows-x86_64-11.8.89-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/libcublas/windows-x86_64/libcublas-windows-x86_64-11.8.1.74-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvtx/windows-x86_64/cuda_nvtx-windows-x86_64-11.8.86-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_profiler_api/windows-x86_64/cuda_profiler_api-windows-x86_64-11.8.86-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/visual_studio_integration/windows-x86_64/visual_studio_integration-windows-x86_64-11.8.86-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_nvprof/windows-x86_64/cuda_nvprof-windows-x86_64-11.8.87-archive.zip"
-          curl -O "https://developer.download.nvidia.com/compute/cuda/redist/cuda_cccl/windows-x86_64/cuda_cccl-windows-x86_64-11.8.89-archive.zip"
-          unzip '*.zip' -d "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8"
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_cudart-windows-x86_64-11.8.89-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_nvcc-windows-x86_64-11.8.89-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_nvrtc-windows-x86_64-11.8.89-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\libcublas-windows-x86_64-11.8.1.74-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_nvtx-windows-x86_64-11.8.86-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_profiler_api-windows-x86_64-11.8.86-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\visual_studio_integration-windows-x86_64-11.8.86-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_nvprof-windows-x86_64-11.8.87-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\cuda_cccl-windows-x86_64-11.8.89-archive\*" "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" /E /I /H /Y
-          xcopy "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\visual_studio_integration\MSBuildExtensions\*" "C:\Program Files (x86)\Microsoft Visual Studio\2019\Enterprise\MSBuild\Microsoft\VC\v160\BuildCustomizations" /E /I /H /Y
-          echo "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\bin" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
-          echo "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\libnvvp" | Out-File -FilePath $env:GITHUB_PATH -Encoding utf8 -Append
-          echo "CUDA_PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" | Out-File -FilePath $env:GITHUB_ENV -Append -Encoding utf8
-          echo "CUDA_PATH_V11_8=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8" | Out-File -FilePath $env:GITHUB_ENV -Append -Encoding utf8
+      - name: Install cuda-toolkit
+        id: cuda-toolkit-11
+        if: ${{ matrix.build == 'cuda11' }}
+        uses: Jimver/cuda-toolkit@v0.2.16
+        with:
+          cuda: '11.8.0'
+          method: network
+          sub-packages: '["nvcc", "cudart", "cublas", "cublas_dev", "thrust", "visual_studio_integration"]'
+          use-github-cache: false
+          use-local-cache: false

      - name: Build
        id: cmake_build
-        shell: cmd
        run: |
          mkdir build
          cd build
-          call "C:\Program Files (x86)\Microsoft Visual Studio\2019\Enterprise\VC\Auxiliary\Build\vcvars64.bat"
-          cmake .. -DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61;75;86;89" -DSD_BUILD_SHARED_LIBS=ON -DCMAKE_CUDA_FLAGS="-allow-unsupported-compiler"
+          cmake .. ${{ matrix.defines }}
          cmake --build . --config Release
      
      - name: Upload artifact
        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+        if: ${{ matrix.build != 'rocm5' }}
+        uses: actions/upload-artifact@v4.3.1
        with:
-          name: windows-cuda11
+          name: windows-${{ matrix.build }}
          path: .\build\bin\Release\stable-diffusion.dll
-
+          
  windows-sycl:
    runs-on: windows-latest
    
@ -185,7 +137,7 @@ jobs:
    steps:
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
@ -193,7 +145,7 @@ jobs:
          
      - name: Checkout Install Script
        id: checkoutInstallScript
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          path: scripts
          sparse-checkout: |
@ -211,7 +163,7 @@ jobs:
              
      - name: Upload artifact
        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
          name: windows-sycl
          path: .\build\bin\stable-diffusion.dll
@ -236,7 +188,7 @@ jobs:
          
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
@ -254,52 +206,13 @@ jobs:
              
      - name: Upload artifact
        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
          name: windows-rocm6
          path: .\build\bin\stable-diffusion.dll

-  windows-vulkan:
-    runs-on: windows-latest
-    
-    env:
-      VULKAN_VERSION: 1.3.261.1
-    
-    steps:
-      - name: Checkout
-        id: checkout
-        uses: actions/checkout@v4.2.2
-        with:
-          repository: 'leejet/stable-diffusion.cpp'
-          ref: '${{ github.event.inputs.commit }}'
-          submodules: recursive
-
-      - name: Install Vulkan SDK
-        id: get_vulkan
-        run: |
-          curl.exe -o $env:RUNNER_TEMP/VulkanSDK-Installer.exe -L "https://sdk.lunarg.com/sdk/download/${env:VULKAN_VERSION}/windows/VulkanSDK-${env:VULKAN_VERSION}-Installer.exe"
-          & "$env:RUNNER_TEMP\VulkanSDK-Installer.exe" --accept-licenses --default-answer --confirm-command install
-          Add-Content $env:GITHUB_ENV "VULKAN_SDK=C:\VulkanSDK\${env:VULKAN_VERSION}"
-          Add-Content $env:GITHUB_PATH "C:\VulkanSDK\${env:VULKAN_VERSION}\bin"
-
-      - name: Build
-        id: cmake_build
-        run: |
-          mkdir build
-          cd build
-          cmake .. -DSD_VULKAN=ON -DSD_BUILD_SHARED_LIBS=ON
-          cmake --build . --config Release
-      
-      - name: Upload artifact
-        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
-        with:
-          name: windows-vulkan
-          path: .\build\bin\Release\stable-diffusion.dll
-          
-  linux-cpu:
+  linux:
    runs-on: ubuntu-latest
-    
    strategy:
      matrix:
        include:
@ -311,115 +224,45 @@ jobs:
            defines: '-DGGML_NATIVE=OFF -DGGML_AVX2=ON -DSD_BUILD_SHARED_LIBS=ON'
          - build: 'avx512'
            defines: '-DGGML_NATIVE=OFF -DGGML_AVX512=ON -DSD_BUILD_SHARED_LIBS=ON'
+          - build: 'cuda11'
+            defines: '-DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61" -DSD_BUILD_SHARED_LIBS=ON'
+          - build: 'cuda12'
+            defines: '-DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61;89" -DSD_BUILD_SHARED_LIBS=ON'            
+          - build: 'sycl'
+            defines: '-DSD_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DSD_BUILD_SHARED_LIBS=ON'
            
    steps:
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
          submodules: recursive
-
-      - name: Build
-        id: cmake_build
-        run: |
-          mkdir build
-          cd build
-          cmake .. ${{ matrix.defines }}
-          cmake --build . --config Release
-                  
-      - name: Upload artifact
-        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+   
+      - name: Install cuda-toolkit
+        id: cuda-toolkit-11
+        if: ${{ matrix.build == 'cuda11' }}
+        uses: Jimver/cuda-toolkit@v0.2.14
        with:
-          name: linux-${{ matrix.build }}
-          path: ./build/bin/libstable-diffusion.so
-          
-  linux-cuda12:
-    runs-on: ubuntu-latest
-    container: nvidia/cuda:12.8.0-devel-ubuntu24.04
-    
-    steps:
-      - name: Install dependencies
-        env:
-          DEBIAN_FRONTEND: noninteractive
-        run: |
-          apt update
-          apt install -y cmake build-essential ninja-build libgomp1 git
-    
-      - name: Checkout
-        id: checkout
-        uses: actions/checkout@v4.2.2
+          cuda: '11.7.1'
+          method: network
+          use-github-cache: false
+          use-local-cache: false
+   
+      - name: Install cuda-toolkit
+        id: cuda-toolkit-12
+        if: ${{ matrix.build == 'cuda12' }}
+        uses: Jimver/cuda-toolkit@v0.2.14
        with:
-          repository: 'leejet/stable-diffusion.cpp'
-          ref: '${{ github.event.inputs.commit }}'
-          submodules: recursive
-
-      - name: Build
-        id: cmake_build
-        run: |
-          mkdir build
-          cd build
-          cmake .. -DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61;75;86;89;100" -DSD_BUILD_SHARED_LIBS=ON
-          cmake --build . --config Release
-     
-      - name: Upload artifact
-        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
-        with:
-          name: linux-cuda12
-          path: ./build/bin/libstable-diffusion.so
-          
-  linux-cuda11:
-    runs-on: ubuntu-latest
-    container: nvidia/cuda:11.8.0-devel-ubuntu22.04
-    
-    steps:
-      - name: Install dependencies
-        env:
-          DEBIAN_FRONTEND: noninteractive
-        run: |
-          apt update
-          apt install -y cmake build-essential ninja-build libgomp1 git
-    
-      - name: Checkout
-        id: checkout
-        uses: actions/checkout@v4.2.2
-        with:
-          repository: 'leejet/stable-diffusion.cpp'
-          ref: '${{ github.event.inputs.commit }}'
-          submodules: recursive
-
-      - name: Build
-        id: cmake_build
-        run: |
-          mkdir build
-          cd build
-          cmake .. -DSD_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES="61;75;86;89" -DSD_BUILD_SHARED_LIBS=ON
-          cmake --build . --config Release
-     
-      - name: Upload artifact
-        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
-        with:
-          name: linux-cuda11
-          path: ./build/bin/libstable-diffusion.so
-          
-  linux-sycl:
-    runs-on: ubuntu-latest
-    
-    steps:
-      - name: Checkout
-        id: checkout
-        uses: actions/checkout@v4.2.2
-        with:
-          repository: 'leejet/stable-diffusion.cpp'
-          ref: '${{ github.event.inputs.commit }}'
-          submodules: recursive
-
+          cuda: '12.2.0'
+          method: network
+          use-github-cache: false
+          use-local-cache: false
+   
      - name: Install Sycl tools
        id: installSyclCompiler
+        if: ${{ matrix.build == 'sycl' }}
        shell: bash
        run: |        
          cd /tmp
@ -431,21 +274,31 @@ jobs:
          sudo apt install intel-oneapi-compiler-dpcpp-cpp intel-oneapi-mkl-devel
   
      - name: Build
-        id: cmake_build
+        id: cmake_build_sycl
+        if: ${{ matrix.build == 'sycl' }}
        run: |
          source /opt/intel/oneapi/setvars.sh
          mkdir build
          cd build
-          cmake .. -DSD_SYCL=ON -DCMAKE_C_COMPILER=icx -DCMAKE_CXX_COMPILER=icpx -DSD_BUILD_SHARED_LIBS=ON
+          cmake .. ${{ matrix.defines }}
          cmake --build . --config Release
-     
+   
+      - name: Build
+        id: cmake_build
+        if: ${{ matrix.build != 'sycl' }}
+        run: |
+          mkdir build
+          cd build
+          cmake .. ${{ matrix.defines }}
+          cmake --build . --config Release
+                  
      - name: Upload artifact
        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
-          name: linux-sycl
+          name: linux-${{ matrix.build }}
          path: ./build/bin/libstable-diffusion.so
-          
+
  linux-hip:
    runs-on: ubuntu-22.04
    container: rocm/dev-ubuntu-22.04:6.0.2
@ -460,7 +313,7 @@ jobs:
          
      - name: Clone
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v3
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
@ -476,12 +329,12 @@ jobs:
               
      - name: Upload artifact Rocm
        id: upload_artifact_rocm
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
          name: linux-rocm6
          path: ./build/bin/libstable-diffusion.so
-
-  osx-cpu:
+          
+  osx:
    runs-on: macos-latest
          
    strategy:
@ -499,7 +352,7 @@ jobs:
    steps:          
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
        with:
          repository: 'leejet/stable-diffusion.cpp'
          ref: '${{ github.event.inputs.commit }}'
@ -515,36 +368,30 @@ jobs:
                  
      - name: Upload artifact
        id: upload_artifact
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
          name: osx-${{ matrix.build }}
          path: ./build/bin/libstable-diffusion.dylib
-          
+
  release:
-    runs-on: windows-latest
+    runs-on: ubuntu-latest

    needs:
-      - windows-cpu
-      - windows-cuda12
-      - windows-cuda11
-      - windows-sycl
+      - windows
      - windows-hip
-      - windows-vulkan
-      - linux-cpu
-      - linux-cuda12
-      - linux-cuda11
-      - linux-sycl
+      - windows-sycl
+      - linux
      - linux-hip
-      - osx-cpu
+      - osx

    steps:
      - name: Checkout
        id: checkout
-        uses: actions/checkout@v4.2.2
+        uses: actions/checkout@v4.1.2
      
      - name: Download artifacts
        id: download_artifacts
-        uses: actions/download-artifact@v4.1.9
+        uses: actions/download-artifact@v4
        with:
            path: Backends
      
@ -553,7 +400,7 @@ jobs:

      - name: Setup nuget
        id: setup_nuget
-        uses: NuGet/setup-nuget@v2.0.1
+        uses: NuGet/setup-nuget@v2.0.0

      - name: Pack
        id: pack
@ -570,11 +417,11 @@ jobs:
          
      - name: Upload artifacts
        id: upload_artifacts
-        uses: actions/upload-artifact@v4.6.1
+        uses: actions/upload-artifact@v4.3.1
        with:
          name: StableDiffusion.NET.Backend-Nugets
          path: ./*.nupkg
          
      - name: Nuget Push
        id: nuget_push
-        run: dotnet nuget push **\*.nupkg --skip-duplicate --api-key ${{ secrets.NUGET_TOKEN }} --source https://api.nuget.org/v3/index.json
+        run: dotnet nuget push **\*.nupkg --skip-duplicate --api-key ${{ secrets.NUGET_TOKEN }} --source https://api.nuget.org/v3/index.json
--- a/Examples/ImageCreationUI/ImageCreationUI.csproj
+++ b/Examples/ImageCreationUI/ImageCreationUI.csproj
@ -11,9 +11,9 @@

  <ItemGroup>
    <PackageReference Include="HPPH.System.Drawing" Version="1.0.0" />
-    <PackageReference Include="StableDiffusion.NET.Backend.Cpu" Version="3.4.1" />
-    <PackageReference Include="StableDiffusion.NET.Backend.Cuda" Version="3.4.1" />
-    <PackageReference Include="StableDiffusion.NET.Backend.Rocm" Version="3.4.1" />
+    <PackageReference Include="StableDiffusion.NET.Backend.Cpu" Version="3.3.1" />
+    <PackageReference Include="StableDiffusion.NET.Backend.Cuda" Version="3.3.1" />
+    <PackageReference Include="StableDiffusion.NET.Backend.Rocm" Version="3.3.1" />
  </ItemGroup>

  <ItemGroup>
--- a/StableDiffusion.NET/Enums/Quantization.cs
+++ b/StableDiffusion.NET/Enums/Quantization.cs
@ -34,14 +34,11 @@ public enum Quantization
    F64 = 28,
    IQ1_M = 29,
    BF16 = 30,
-    //Q4_0_4_4 = 31, support has been removed from gguf files
-    //Q4_0_4_8 = 32,
-    //Q4_0_8_8 = 33,
+    Q4_0_4_4 = 31,
+    Q4_0_4_8 = 32,
+    Q4_0_8_8 = 33,
    TQ1_0 = 34,
    TQ2_0 = 35,
-    // SD_TYPE_IQ4_NL_4_4 = 36,
-    // SD_TYPE_IQ4_NL_4_8 = 37,
-    // SD_TYPE_IQ4_NL_8_8 = 38,

-    Unspecified = 39
+    Unspecified
 }
--- a/StableDiffusion.NET/Enums/Sampler.cs
+++ b/StableDiffusion.NET/Enums/Sampler.cs
@ -11,7 +11,5 @@ public enum Sampler
    DPMPP2Mv2,
    IPNDM,
    IPNDM_V,
-    LCM,
-    DDIM_Trailing,
-    TCD
+    LCM
 }
--- a/StableDiffusion.NET/Models/DiffusionModel.cs
+++ b/StableDiffusion.NET/Models/DiffusionModel.cs
@ -113,7 +113,6 @@ public sealed unsafe class DiffusionModel : IDisposable
                                            parameter.ClipSkip,
                                            parameter.CfgScale,
                                            parameter.Guidance,
-                                            parameter.Eta,
                                            parameter.Width,
                                            parameter.Height,
                                            parameter.SampleMethod,
@ -149,7 +148,6 @@ public sealed unsafe class DiffusionModel : IDisposable
                                            parameter.ClipSkip,
                                            parameter.CfgScale,
                                            parameter.Guidance,
-                                            parameter.Eta,
                                            parameter.Width,
                                            parameter.Height,
                                            parameter.SampleMethod,
@ -177,7 +175,6 @@ public sealed unsafe class DiffusionModel : IDisposable
                                    parameter.ClipSkip,
                                    parameter.CfgScale,
                                    parameter.Guidance,
-                                    parameter.Eta,
                                    parameter.Width,
                                    parameter.Height,
                                    parameter.SampleMethod,
--- a/StableDiffusion.NET/Models/Parameter/DiffusionParameter.cs
+++ b/StableDiffusion.NET/Models/Parameter/DiffusionParameter.cs
@ -90,10 +90,5 @@ public sealed class DiffusionParameter
    /// </summary>
    public float Guidance { get; set; } = 3.5f;

-    /// <summary>
-    /// eta in DDIM, only for DDIM and TCD (default: 0)
-    /// </summary>
-    public float Eta { get; set; } = 0f;
-
    #endregion
 }
--- a/StableDiffusion.NET/Native/Native.cs
+++ b/StableDiffusion.NET/Native/Native.cs
@ -82,7 +82,6 @@ internal unsafe partial class Native
                                                int clip_skip,
                                                float cfg_scale,
                                                float guidance,
-                                                float eta,
                                                int width,
                                                int height,
                                                sample_method_t sample_method,
--- a/build.bat
+++ b/build.bat
@ -0,0 +1,48 @@
+if not exist stable-diffusion.cpp (
+    git clone --recursive https://github.com/leejet/stable-diffusion.cpp
+)
+
+cd stable-diffusion.cpp
+git fetch
+git checkout 73c2176648898b3a223c581ed138e5593f8fded5
+git submodule init
+git submodule update
+
+if not exist build (
+    mkdir build
+)
+
+cd build
+
+Rem ----------------------------------------------------------------------------
+rem Pick one of the builds below.
+
+rem # cuda12 #
+cmake .. -DSD_CUBLAS=ON -DSD_BUILD_SHARED_LIBS=ON -DSD_BUILD_EXAMPLES=OFF
+
+rem # rocm5.5 #
+rem cmake .. -G Ninja -DCMAKE_C_COMPILER=clang -DCMAKE_CXX_COMPILER=clang++ -DSD_HIPBLAS=ON -DCMAKE_BUILD_TYPE=Release -DAMDGPU_TARGETS="gfx1100;gfx1102;gfx1030" -DSD_BUILD_SHARED_LIBS=ON -DSD_BUILD_EXAMPLES=OFF
+
+rem # avx512 #
+rem cmake .. -DGGML_AVX512=ON -DSD_BUILD_SHARED_LIBS=ON -DSD_BUILD_EXAMPLES=OFF
+
+rem # avx2 #
+rem cmake .. -DGGML_AVX2=ON -DSD_BUILD_SHARED_LIBS=ON -DSD_BUILD_EXAMPLES=OFF
+
+rem # avx #
+rem cmake .. -DGGML_AVX2=OFF -DSD_BUILD_SHARED_LIBS=ON -DSD_BUILD_EXAMPLES=OFF
+
+rem # noavx #
+rem cmake .. -DGGML_AVX=OFF -DGGML_AVX2=OFF -DGGML_FMA=OFF -DSD_BUILD_SHARED_LIBS=ON -DSD_BUILD_EXAMPLES=OFF
+
+Rem ----------------------------------------------------------------------------
+
+cmake --build . --config Release
+
+cd ..\..
+
+dotnet publish -c Release -o bin
+
+copy .\stable-diffusion.cpp\build\bin\Release\*.dll .\bin\
+
+pause