Add Debug Pipelines

.github/workflows/linux-cpu-arm64-build.yml

@@ -70,26 +70,26 @@ jobs:
           --container-registry onnxruntimebuildcache \
           --repository ort_genai_linux_arm64_gha
 
-      - name: Docker -- Configure with CMake and GCC
+      - name: Build with Python
         run: |
-          docker run --rm \
-          --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "/usr/bin/cmake --preset linux_gcc_cpu_release"
-
-      - name: Docker -- Build with CMake and GCC
-        run: |
-          docker run --rm \
-          --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "/usr/bin/cmake --build --preset linux_gcc_cpu_release"
+          set -e -x
+          docker run \
+            --rm \
+            --volume $GITHUB_WORKSPACE:/onnxruntime_src \
+            -w /onnxruntime_src ort_genai_linux_arm64_gha \
+            bash -c " \
+              python3 -m pip install wheel && \
+              python3 -m pip install requests && \
+              python3 build.py --parallel --config=Debug"
 
       - name: Docker -- Check test directory
         run: |
           docker run --rm \
           --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "ls -l /onnxruntime_src/build/cpu/test/"
+          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "ls -l /onnxruntime_src/build/Linux/Debug/test/"
 
       - name: Docker -- Run tests
         run: |
           docker run --rm \
           --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/onnxruntime_src/ort/lib/ /onnxruntime_src/build/cpu/unit_tests"
+          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/onnxruntime_src/ort/lib/ /onnxruntime_src/build/Linux/Debug/unit_tests"

.github/workflows/linux-cpu-x64-build.yml

@@ -70,19 +70,17 @@ jobs:
           mv microsoft.ml.onnxruntime/${{ env.ORT_NIGHTLY_VERSION }}/runtimes/linux-x64/native/* ort/lib/
           cp ort/lib/libonnxruntime.so ort/lib/libonnxruntime.so.1
 
-      - name: Build with CMake and GCC
+      - name: Build with Python
         run: |
-          set -e -x
-          rm -rf build
-          cmake --preset linux_gcc_cpu_release
-          cmake --build --preset linux_gcc_cpu_release
+          python -m pip install requests
+          python3 build.py --parallel --config=Debug
 
       - name: Install the python wheel and test dependencies
         run: |
           python3 -m pip install -r test/python/requirements.txt --user
           python3 -m pip install -r test/python/cpu/torch/requirements.txt --user
           python3 -m pip install -r test/python/cpu/ort/requirements.txt --user
-          python3 -m pip install --user --no-index --no-deps --find-links build/cpu/wheel onnxruntime_genai
+          python3 -m pip install --user --no-index --no-deps --find-links build/Linux/Debug/wheel onnxruntime_genai
 
       - name: Use Dummy HuggingFace Token
         run: |
@@ -92,8 +90,8 @@ jobs:
         if: always()
         continue-on-error: true
         run: |
-          ls -l ${{ github.workspace }}/build/cpu
-          ls -l ${{ github.workspace }}/build/cpu/wheel
+          ls -l ${{ github.workspace }}/build/Linux/Debug
+          ls -l ${{ github.workspace }}/build/Linux/Debug/wheel
 
       # This will also download all the test models to the test/test_models directory
       # These models are used by the python tests as well as C#, C++ and others.
@@ -106,16 +104,16 @@ jobs:
         run: |
           export ORTGENAI_LOG_ORT_LIB=1
           cd test/csharp
-          dotnet test /p:Configuration=Release /p:NativeBuildOutputDir="../../build/cpu/" /p:OrtLibDir="../../ort/lib/" --verbosity normal
+          dotnet test /p:Configuration=Debug /p:NativeBuildOutputDir="../../build/Linux/Debug/" /p:OrtLibDir="../../ort/lib/" --verbosity normal
 
       - name: Build the Java API and Run the Java Tests
         run: |
           set -e -x
-          python3 build.py --config=Release --build_dir build/cpu --build_java --parallel --cmake_generator "Ninja"
+          python3 build.py --config=Debug --build_dir build/Linux/Debug --build_java --parallel
 
       - name: Run tests
         run: |
           set -e -x
           export ORTGENAI_LOG_ORT_LIB=1
           export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$GITHUB_WORKSPACE/ort/lib
-          ./build/cpu/unit_tests
+          ./build/Linux/Debug/unit_tests

.github/workflows/linux-cpu-x64-nightly-build.yml

@@ -22,8 +22,6 @@ jobs:
       - name: Checkout OnnxRuntime GenAI repo
         uses: actions/checkout@v2
 
-
-
       - name: Download OnnxRuntime
         run: |
           curl -L -o ${{ env.ort_zip }} ${{ env.ort_url }} 
@@ -41,19 +39,17 @@ jobs:
         run: |
           git submodule update --init --recursive
 
-      - name: Build with CMake and clang
+      - name: Build with Python
         run: |
-          set -e -x
-          rm -rf build
-          cmake --preset linux_gcc_cpu_release
-          cmake --build --preset linux_gcc_cpu_release
+          python3 -m pip install requests
+          python3 build.py --parallel --config=Release
 
       - name: Install the python wheel and test dependencies
         run: |
           python3 -m pip install -r test/python/requirements.txt --user
           python3 -m pip install -r test/python/cpu/torch/requirements.txt --user
           python3 -m pip install -r test/python/cpu/ort/requirements.txt --user
-          python3 -m pip install build/cpu/wheel/onnxruntime_genai*.whl --no-deps
+          python3 -m pip install build/Linux/Release/wheel/onnxruntime_genai*.whl --no-deps
 
       - name: Use Dummy HuggingFace Token
         run: |

.github/workflows/linux-gpu-x64-build.yml

@@ -85,7 +85,7 @@ jobs:
             --multiple_repos \
             --repository onnxruntimecudabuildx64
 
-      - name: Config with Cmake in Docker
+      - name: Build with Python in Docker
         run: |
           set -e -x
           docker run \
@@ -94,20 +94,9 @@ jobs:
             --volume $GITHUB_WORKSPACE:/ort_genai_src \
             -w /ort_genai_src onnxruntimecudabuildx64 \
             bash -c " \
-              /usr/bin/cmake --preset linux_gcc_cuda_release \
-                -DMANYLINUX=ON \
-                -DPYTHON_EXECUTABLE=${{ env.PYTHON_EXECUTABLE }} "
-
-      - name: Build with Cmake in Docker
-        run: |
-          set -e -x
-          docker run \
-            --gpus all \
-            --rm \
-            --volume $GITHUB_WORKSPACE:/ort_genai_src \
-            -w /ort_genai_src onnxruntimecudabuildx64 \
-            bash -c " \
-              /usr/bin/cmake --build --preset linux_gcc_cuda_release"
+              ${{ env.PYTHON_EXECUTABLE }} -m pip install requests && \
+              ${{ env.PYTHON_EXECUTABLE }} -m pip install wheel && \
+              ${{ env.PYTHON_EXECUTABLE }} build.py --use_cuda --cuda_home=/usr/local/cuda-12.2 --parallel --config=Debug"
 
       - name: Use Dummy HuggingFace Token
         run: |
@@ -126,7 +115,7 @@ jobs:
               ${{ env.PYTHON_EXECUTABLE }} -m pip install -r test/python/requirements.txt --user && \
               ${{ env.PYTHON_EXECUTABLE }} -m pip install -r test/python/cuda/torch/requirements.txt --user && \
               ${{ env.PYTHON_EXECUTABLE }} -m pip install -r test/python/cuda/ort/requirements.txt --user && \
-              ${{ env.PYTHON_EXECUTABLE }} -m pip install /ort_genai_src/build/cuda/wheel/onnxruntime_genai*manylinux*.whl --no-deps --user && \
+              ${{ env.PYTHON_EXECUTABLE }} -m pip install /ort_genai_src/build/Linux/Debug/wheel/onnxruntime_genai*manylinux*.whl --no-deps --user && \
               ${{ env.PYTHON_EXECUTABLE }} test/python/test_onnxruntime_genai.py --cwd test/python --test_models test/test_models --e2e"
 
       # TODO: Enable this by adding dotnet to the docker image
@@ -148,4 +137,4 @@ jobs:
             --rm \
             --volume /data/ortgenai_pytorch_models:/data/ortgenai_pytorch_models \
             --volume $GITHUB_WORKSPACE:/ort_genai_src \
-            -w /ort_genai_src onnxruntimecudabuildx64 bash -c "ORTGENAI_LOG_ORT_LIB=1 LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/ort_genai_src/build/cuda/ /ort_genai_src/build/cuda/unit_tests"
+            -w /ort_genai_src onnxruntimecudabuildx64 bash -c "ORTGENAI_LOG_ORT_LIB=1 LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/ort_genai_src/build/Linux/Debug/ /ort_genai_src/build/Linux/Debug/unit_tests"

.github/workflows/mac-cpu-arm64-build.yml

@@ -52,14 +52,11 @@ jobs:
           mv ${{ env.ORT_PACKAGE_NAME }}/build/native/include ort/
           mv ${{ env.ORT_PACKAGE_NAME }}/runtimes/osx-arm64/native/* ort/lib/
 
-      - name: Configure CMake
+      - name: Build with Python
         run: |
-          cmake --preset macos_arm64_cpu_release
-
-      - name: Build with CMake
-        run: |
-          cmake --build --preset macos_arm64_cpu_release --parallel
-        continue-on-error: false
+          python3 -m pip install requests
+          python3 -m pip install wheel
+          python3 build.py --parallel --config=Release --cmake_generator "Unix Makefiles"
 
       - name: Install the python wheel and test dependencies
         run: |
@@ -68,7 +65,7 @@ jobs:
           python3 -m pip install -r test/python/requirements.txt
           python3 -m pip install -r test/python/macos/torch/requirements.txt
           python3 -m pip install -r test/python/macos/ort/requirements.txt
-          python3 -m pip install build/cpu/osx-arm64/wheel/onnxruntime_genai*.whl --no-deps
+          python3 -m pip install build/MacOS/Release/wheel/onnxruntime_genai*.whl --no-deps
 
       - name: Remove the ort lib and header files
         run: |
@@ -78,8 +75,8 @@ jobs:
         if: always()
         continue-on-error: true
         run: |
-          ls -l ${{ github.workspace }}/build/cpu/osx-arm64
-          ls -l ${{ github.workspace }}/build/cpu/osx-arm64/wheel
+          ls -l ${{ github.workspace }}/build/MacOS/Release
+          ls -l ${{ github.workspace }}/build/MacOS/Release/wheel
 
       # This will also download all the test models to the test/test_models directory
       # These models are used by the python tests as well as C#, C++ and others.
@@ -95,17 +92,18 @@ jobs:
         run: |
           export ORTGENAI_LOG_ORT_LIB=1
           cd test/csharp
-          dotnet test /p:Configuration=Release /p:NativeBuildOutputDir="../../build/cpu/osx-arm64" --verbosity normal
+          dotnet test /p:Configuration=Release /p:NativeBuildOutputDir="../../build/MacOS/Release/Release" --verbosity normal
 
       - name: Build the Java API and Run the Java Tests
         run: |
           set -e -x
           source genai-macos-venv/bin/activate
-          python3 build.py --config=Release --build_dir build/cpu/osx-arm64 --build_java --parallel --cmake_generator "Unix Makefiles"
+          python3 build.py --config=Release --build_java --parallel --cmake_generator "Unix Makefiles"
+          dotnet test /p:Configuration=Release /p:NativeBuildOutputDir="../../build/MacOS/Release/Release" --verbosity normal
 
       - name: Run tests
         run: |
           set -e -x
           export ORTGENAI_LOG_ORT_LIB=1
-          export DYLD_LIBRARY_PATH=$DYLD_LIBRARY_PATH:$GITHUB_WORKSPACE/build/cpu/osx-arm64
-          ./build/cpu/osx-arm64/unit_tests
+          export DYLD_LIBRARY_PATH=$DYLD_LIBRARY_PATH:$GITHUB_WORKSPACE/build/MacOS/Release/Release
+          ./build/MacOS/Release/Release/unit_tests

.github/workflows/win-cpu-arm64-build.yml

@@ -11,7 +11,7 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
   cancel-in-progress: true
 env:
-  binaryDir: 'build/cpu/win-arm64'
+  binaryDir: 'build/Windows/Release'
   ORT_NIGHTLY_REST_API: "https://feeds.dev.azure.com/aiinfra/PublicPackages/_apis/packaging/Feeds/ORT-Nightly/packages?packageNameQuery=Microsoft.ML.OnnxRuntime&api-version=6.0-preview.1"
   ORT_PACKAGE_NAME: "Microsoft.ML.OnnxRuntime"
 
@@ -34,18 +34,6 @@ jobs:
       with:
         nuget-version: '5.x'
 
-    - name: Setup Java 21
-      uses: actions/setup-java@v4
-      with:
-        java-version: '21'
-        distribution: 'temurin'
-        cache: 'gradle'
-
-    - name: Setup Gradle
-      uses: gradle/actions/setup-gradle@v3
-      with:
-        gradle-version: '8.6'
-
     - name: Download OnnxRuntime Nightly
       shell: powershell
       run: |
@@ -64,15 +52,16 @@ jobs:
         move ${{ env.ORT_PACKAGE_NAME }}/build/native/include ort/
         move ${{ env.ORT_PACKAGE_NAME }}/runtimes/win-arm64/native/* ort/lib/
 
-    - name: Configure CMake
+    - name: Build with Python
       run: |
-        python -m pip install wheel requests
+        python -m pip install requests
+        python -m pip install wheel
+        python build.py --parallel --config=Release
 
-        cmake --preset windows_arm64_cpu_release
-
-    - name: Build with CMake
+    - name: Build the C# API and Run the C# Tests
       run: |
-        cmake --build --preset windows_arm64_cpu_release --parallel
+        cd test\csharp
+        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Release" /p:OrtLibDir="$env:GITHUB_WORKSPACE\ort\lib"
 
     - name: Install the Python Wheel and Test Dependencies
       run: |
@@ -84,15 +73,6 @@ jobs:
       run: |
         python test/python/test_onnxruntime_genai.py --cwd "test\python" --test_models "test\test_models"
 
-    - name: Build the C# API and Run the C# Tests
-      run: |
-        cd test\csharp
-        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Release" /p:OrtLibDir="$env:GITHUB_WORKSPACE\ort\lib"
-
-    - name: Build the Java API and Run the Java Tests
-      run: |
-        python build.py --config=Release --build_dir $env:binaryDir --build_java --parallel
-
     - name: Verify Build Artifacts
       if: always()
       continue-on-error: true

.github/workflows/win-cpu-x64-build.yml

@@ -11,7 +11,7 @@ concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
   cancel-in-progress: true
 env:
-  binaryDir: 'build/cpu/win-x64'
+  binaryDir: 'build/Windows/Debug'
   ORT_NIGHTLY_REST_API: "https://feeds.dev.azure.com/aiinfra/PublicPackages/_apis/packaging/Feeds/ORT-Nightly/packages?packageNameQuery=Microsoft.ML.OnnxRuntime&api-version=6.0-preview.1"
   ORT_PACKAGE_NAME: "Microsoft.ML.OnnxRuntime"
 
@@ -76,13 +76,11 @@ jobs:
       with:
         languages: 'cpp'
 
-    - name: Configure CMake
+    - name: Build with Python
       run: |
-        cmake --preset windows_x64_cpu_release
-
-    - name: Build with CMake
-      run: |
-        cmake --build --preset windows_x64_cpu_release --parallel
+        python3 -m pip install requests
+        python3 -m pip install wheel
+        python3 build.py --parallel --config=Debug --build_java
 
     - name: Install the python wheel and test dependencies
       run: |
@@ -102,23 +100,19 @@ jobs:
     - name: Build the C# API and Run the C# Tests
       run: |
         cd test\csharp
-        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Release" /p:OrtLibDir="$env:GITHUB_WORKSPACE\ort\lib" --verbosity normal
-
-    - name: Build the Java API and Run the Java Tests
-      run: |
-        python3 build.py --config=Release --build_dir $env:binaryDir --build_java --parallel
+        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Debug" /p:OrtLibDir="$env:GITHUB_WORKSPACE\ort\lib" --verbosity normal
 
     - name: Verify Build Artifacts
       if: always()
       continue-on-error: true
       run: |
-        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir -Recurse
+        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir\Debug -Recurse
         Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir\test -Recurse
 
     - name: Run tests
       run: |
-        copy $env:GITHUB_WORKSPACE\ort\lib\* .\$env:binaryDir\Release
-        & .\$env:binaryDir\Release\unit_tests.exe
+        copy $env:GITHUB_WORKSPACE\ort\lib\* .\$env:binaryDir\Debug
+        & .\$env:binaryDir\Debug\unit_tests.exe
 
     - name: Perform CodeQL Analysis
       uses: github/codeql-action/analyze@v3