rebase with main

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

@@ -56,10 +56,10 @@ jobs:
         run: |
           docker run --rm \
           --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "ls -l /onnxruntime_src/build/gcc_cpu/release/test/"
+          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "ls -l /onnxruntime_src/build/cpu/test/"
 
       - name: Docker -- Run tests
         run: |
           docker run --rm \
           --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "/onnxruntime_src/build/gcc_cpu/release/test/unit_tests"   
+          -w /onnxruntime_src ort_genai_linux_arm64_gha bash -c "/onnxruntime_src/build/cpu/test/unit_tests"   

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

@@ -39,7 +39,7 @@ jobs:
 
       - name: Install the python wheel and test dependencies
         run: |
-          python3 -m pip install build/gcc_cpu/release/wheel/onnxruntime_genai*.whl
+          python3 -m pip install build/cpu/wheel/onnxruntime_genai*.whl
           python3 -m pip install -r test/python/requirements-nightly-cpu.txt --user
 
       - name: Get HuggingFace Token
@@ -59,9 +59,9 @@ jobs:
         if: always()
         continue-on-error: true
         run: |
-          ls -l ${{ github.workspace }}/build/gcc_cpu/release
+          ls -l ${{ github.workspace }}/build/cpu
 
       - name: Run tests
         run: |
           set -e -x
-          ./build/gcc_cpu/release/test/unit_tests
+          ./build/cpu/test/unit_tests

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

@@ -45,12 +45,12 @@ jobs:
         run: |
           set -e -x
           rm -rf build
-          cmake --preset linux_clang_cpu_release
-          cmake --build --preset linux_clang_cpu_release
+          cmake --preset linux_gcc_cpu_release
+          cmake --build --preset linux_gcc_cpu_release
 
       - name: Install the python wheel and test dependencies
         run: |
-          python3 -m pip install build/clang_cpu/release/wheel/onnxruntime_genai*.whl
+          python3 -m pip install build/cpu/wheel/onnxruntime_genai*.whl
           python3 -m pip install -r test/python/requirements-nightly-cpu.txt --user
 
       - name: Get HuggingFace Token

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

@@ -63,7 +63,7 @@ jobs:
             --gpus all \
             --rm \
             --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-            -w /onnxruntime_src ort_genai_linux_gpu_gha bash -c "python3 -m pip install /onnxruntime_src/build/gcc_cuda/release/wheel/onnxruntime_genai*.whl --user && python3 -m pip install -r test/python/requirements.txt --user && python3 test/python/test_onnxruntime_genai.py --cwd test/python --test_models test/test_models"
+            -w /onnxruntime_src ort_genai_linux_gpu_gha bash -c "python3 -m pip install /onnxruntime_src/build/cuda/wheel/onnxruntime_genai*.whl --user && python3 -m pip install -r test/python/requirements.txt --user && python3 test/python/test_onnxruntime_genai.py --cwd test/python --test_models test/test_models"
 
       - name: Docker -- Run tests
         run: |
@@ -72,4 +72,4 @@ jobs:
             --gpus all \
             --rm \
             --volume $GITHUB_WORKSPACE:/onnxruntime_src \
-            -w /onnxruntime_src ort_genai_linux_gpu_gha bash -c "/onnxruntime_src/build/gcc_cuda/release/test/unit_tests"
+            -w /onnxruntime_src ort_genai_linux_gpu_gha bash -c "/onnxruntime_src/build/cuda/test/unit_tests"

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

@@ -33,11 +33,12 @@ jobs:
         run: |
           mv ${{ env.ort_dir }} ort
 
-      - name: Build with CMake and Clang
+      - name: Configure CMake
         run: |
+          cmake --preset macos_cpu_release
+
       - name: Build with CMake
         run: |
-          cmake --preset macos_cpu_release
           cmake --build --preset macos_cpu_release --parallel
         continue-on-error: false
 

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

@@ -14,7 +14,7 @@ env:
   ort_dir: "onnxruntime-win-arm64-1.17.1"
   ort_zip: "$(ort_dir).zip"
   ort_url: "https://github.com/microsoft/onnxruntime/releases/download/v1.17.1/$(ort_zip)"
-  cmake_build_dir: 'build/release/cpu_default'
+  binaryDir: 'build/cpu'
 
 jobs:
   windows-cpu-arm64-build:
@@ -45,14 +45,17 @@ jobs:
       run: |
         Rename-Item -Path $env:ort_dir -NewName ort
 
-    - name: Build with CMake
+    - name: Configure CMake
       run: |
         cmake --preset windows_arm64_cpu_release
+
+    - name: Build with CMake
+      run: |
         cmake --build --preset windows_arm64_cpu_release --parallel
 
     - name: Install the Python Wheel and Test Dependencies
       run: |
-        python -m pip install (Get-ChildItem ("$env:cmake_build_dir\wheel\*.whl"))
+        python -m pip install (Get-ChildItem ("$env:binaryDir\wheel\*.whl"))
         python -m pip install -r test\python\requirements.txt
 
     - name: Run the Python Tests
@@ -62,15 +65,15 @@ jobs:
     - name: Build the C# API and Run the C# Tests
       run: |
         cd test\csharp
-        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:cmake_build_dir\Release"
+        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Release"
 
     - name: Verify Build Artifacts
       if: always()
       continue-on-error: true
       run: |
-        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:cmake_build_dir -Recurse
-        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:cmake_build_dir\test -Recurse
+        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir -Recurse
+        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir\test -Recurse
 
     - name: Run tests
       run: |
-        .\build\release\cpu_default\test\Release\unit_tests.exe
+        .\build\cpu\test\Release\unit_tests.exe

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

@@ -14,7 +14,7 @@ env:
   ort_dir: "onnxruntime-win-x64-1.17.1"
   ort_zip: "$(ort_dir).zip"
   ort_url: "https://github.com/microsoft/onnxruntime/releases/download/v1.17.1/$(ort_zip)"
-  cmake_build_dir: 'build/release/cpu_default'
+  binaryDir: 'build/cpu'
 
 jobs:
   windows-cpu-x64-build:
@@ -52,14 +52,17 @@ jobs:
       with:
         languages: 'cpp'
 
-    - name: Build with CMake
+    - name: Configure CMake
       run: |
         cmake --preset windows_x64_cpu_release
+
+    - name: Build with CMake
+      run: |
         cmake --build --preset windows_x64_cpu_release --parallel
 
     - name: Install the python wheel and test dependencies
       run: |
-        python -m pip install (Get-ChildItem ("$env:cmake_build_dir\wheel\*.whl"))
+        python -m pip install (Get-ChildItem ("$env:binaryDir\wheel\*.whl"))
         python -m pip install -r test\python\requirements-nightly-cpu.txt
 
     - name: Get HuggingFace Token
@@ -76,18 +79,18 @@ jobs:
     - name: Build the C# API and Run the C# Tests
       run: |
         cd test\csharp
-        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:cmake_build_dir\Release"
+        dotnet test /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Release"
 
     - name: Verify Build Artifacts
       if: always()
       continue-on-error: true
       run: |
-        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:cmake_build_dir -Recurse
-        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:cmake_build_dir\test -Recurse
+        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir -Recurse
+        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir\test -Recurse
 
     - name: Run tests
       run: |
-        .\build\release\cpu_default\test\Release\unit_tests.exe
+        .\build\cpu\test\Release\unit_tests.exe
 
     - name: Perform CodeQL Analysis
       uses: github/codeql-action/analyze@v3

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

@@ -14,7 +14,7 @@ env:
   cuda_dir: "${{ github.workspace }}\\cuda_sdk"
   cuda_version: "11.8"
   CUDA_PATH: ${{ github.workspace }}\\cuda_sdk\\v11.8
-  cmake_build_dir: 'build/release/cuda_default'
+  binaryDir: 'build/cuda'
 
 
 jobs:
@@ -47,9 +47,12 @@ jobs:
       run: |
         Rename-Item -Path $env:ort_dir -NewName ort
 
-    - name: Build with CMake
+    - name: Configure CMake
       run: |
         cmake --preset windows_x64_cuda_release -T cuda=${{ env.cuda_dir }}\\v${{ env.cuda_version }} -DTEST_PHI2=False
+
+    - name: Build with CMake
+      run: |
         cmake --build --preset windows_x64_cuda_release --parallel
 
     - name: Add CUDA to PATH
@@ -58,7 +61,7 @@ jobs:
 
     - name: Install the Python Wheel and Test Dependencies
       run: |
-        python -m pip install (Get-ChildItem ("$env:cmake_build_dir\wheel\*.whl"))
+        python -m pip install (Get-ChildItem ("$env:binaryDir\wheel\*.whl"))
         python -m pip install -r test\python\requirements-nightly-cpu.txt
 
     - name: Get HuggingFace Token
@@ -75,17 +78,17 @@ jobs:
     - name: Build the C# API and Run the C# Tests
       run: |
         cd test\csharp
-        dotnet test /p:Configuration=release /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:cmake_build_dir\Release"
+        dotnet test /p:Configuration=release /p:NativeBuildOutputDir="$env:GITHUB_WORKSPACE\$env:binaryDir\Release"
 
     - name: Verify Build Artifacts
       if: always()
       continue-on-error: true
       run: |
         
-        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:cmake_build_dir -Recurse
+        Get-ChildItem -Path $env:GITHUB_WORKSPACE\$env:binaryDir -Recurse
 
     - name: Prepend CUDA to PATH and Run tests
       run: |
         $env:PATH = "${{ env.cuda_dir }}\\v${{ env.cuda_version }}\\bin;" + $env:PATH 
         echo "Current PATH variable is: $env:PATH" 
-        .\build\release\cuda_default\test\Release\unit_tests.exe
+        .\build\cuda\test\Release\unit_tests.exe

.gitignore

@@ -10,7 +10,7 @@
 /test/test_models/*
 /cache_models
 /onnxruntime-linux-x64-*
-/*.csv
+*.csv
 .idea
 cache_dir
 example-models

.pipelines/nuget-publishing.yml

@@ -33,6 +33,11 @@ parameters:
   - '12.2'
   default: '11.8'
 
+- name: publish_to_ado_feed
+  displayName: 'Publish to Azure DevOps Feed'
+  type: boolean
+  default: false
+
 resources:
   repositories:
   - repository: manylinux
@@ -51,4 +56,4 @@ stages:
     enable_linux_cuda: ${{ parameters.enable_linux_cuda }}
     ort_version: ${{ parameters.ort_version }}
     cuda_version: ${{ parameters.cuda_version }}
-
+    publish_to_ado_feed: ${{ parameters.publish_to_ado_feed }}

.pipelines/pypl-publishing.yml

@@ -32,6 +32,11 @@ parameters:
   - '11.8'
   - '12.2'
 
+- name: publish_to_ado_feed
+  displayName: 'Whether to publish the packages to ADO feed.'
+  type: boolean
+  default: false
+
 resources:
   repositories:
   - repository: manylinux
@@ -50,3 +55,4 @@ stages:
     enable_win_cuda: ${{ parameters.enable_win_cuda }}
     ort_version: ${{ parameters.ort_version }}
     cuda_version: ${{ parameters.cuda_version }}
+    publish_to_ado_feed: ${{ parameters.publish_to_ado_feed }}