Build the generate() API from source
- Pre-requisites
- Clone the onnxruntime-genai repo
- Build the generate() API
- Install the library into your application
Pre-requisites
cmake.NET6(if building C#)
Clone the onnxruntime-genai repo
git clone https://github.com/microsoft/onnxruntime-genaicdonnxruntime-genaiBuild the generate() API
This step assumes that you are in the root of the onnxruntime-genai repo.
All of the build commands below have a--config argument, which takes the following options:
Releasebuilds release binariesDebugbuild binaries with debug symbolsRelWithDebInfobuilds release binaries with debug info
Build Python API
Windows CPU build
python build.py--config ReleaseWindows DirectML build
python build.py--use_dml--config ReleaseWindows NvTensorRtRtx build
python build.py--use_trt_rtx--config Release--cuda_home <cuda_path>Linux build
python build.py--config ReleaseLinux CUDA build
python build.py--use_cuda--config ReleaseMac build
python build.py--config ReleaseBuild Java API
python build.py--build_java--config ReleaseBuild for Android
If building on Windows, installninja.
pipinstallninjaRun the build script.
python build.py--build_java--android--android_home <path to your Android SDK>--android_ndk_path <path to your NDK installation>--android_abi[armeabi-v7a|arm64-v8a|x86|x86_64]--config ReleaseInstall the library into your application
Install Python wheel
# Change dir to the folder containing the onnxruntime-genai wheel# Example for Linux: cd build/Linux/Release/wheel/pipinstall*.whlInstall NuGet
Coming soon
Install JAR
Copybuild/Windows/Release/src/java/build/libs/*.jar into your application.
Install AAR
Copybuild/Android/Release/src/java/build/android/outputs/aar/onnxruntime-genai-release.aar into your application.
Install C/C++ header file and library
Windows
Use the header insrc\ort_genai.h and the libraries inbuild\Windows\Release
Linux
Use the header insrc/ort_genai.h and the libraries inbuild/Linux/Release