microsoft/onnxruntime-extensions

Public

mirrored from https://github.com/microsoft/onnxruntime-extensionsAvailable

CodeCommitsIssuesPull requestsActionsInsightsSecurity
v.0.11.0

Branches

Tags

  • No tags available.
0Branches0Tags
Go to file
Add file
Code

Clone

HTTPS

Download ZIP

docs/development.md

71lines · modeblame

0647ce6dWenbing Li3 years ago1# Build and Development
2
3This project supports Python and can be built from source easily, or a simple cmake build without Python dependency.
fb2a8c28Wenbing Li2 years ago4
0647ce6dWenbing Li3 years ago5## Python package
fb2a8c28Wenbing Li2 years ago6
0647ce6dWenbing Li3 years ago7The package contains all custom operators and some Python scripts to manipulate the ONNX models.
fb2a8c28Wenbing Li2 years ago8
972cb17aWenbing Li2 years ago9- Install Visual Studio with C++ development tools on Windows, or gcc(>8.0) for Linux or xcode for macOS, and cmake on the unix-like platform.
10- If running on Windows, ensure that long file names are enabled, both for the [operating system](https://docs.microsoft.com/en-us/windows/win32/fileio/maximum-file-path-limitation?tabs=cmd) and for git: `git config --system core.longpaths true`
11- Make sure the Python development header/library files be installed, (like `apt-get install python3-dev` for Ubuntu Linux)
12- `pip install .` to build and install the package.<br/> OR `pip install -e .` to install the package in the development mode, which is more friendly for the developer since the Python code change will take effect without having to copy the files to a different location in the disk.(**hints**: debug=1 in setup.cfg wil make C++ code be debuggable in a Python process.)
13- Add the following argument `--config-settings "ortx-user-option=use-cuda"` in the pip command line to enable **CUDA** kernels for the package.
14- The flags that can be used in --config-settings are:
fb2a8c28Wenbing Li2 years ago15- use-cuda: enable CUDA kernel build in Python package.
16- no-azure: disable AzureOp kernel build in Python package.
17- no-opencv: disable operators based on OpenCV in build.
5d558bd3Wenbing Li2 years ago18- cc-debug: generate debug info for extensions binaries and disable C/C++ compiler optimization.
19- cuda-archs: specify the CUDA architectures(like 70, 85, etc.), and the multiple values can be combined with semicolon. The default value is nvidia-smi util output of GPU-0
fb2a8c28Wenbing Li2 years ago20
5d558bd3Wenbing Li2 years ago21For example:`pip install . --config-settings "ortx-user-option=use-cuda,cc-debug" `, This command builds CUDA kernels into the package and installs it, accompanied by the generation of debug information.
0647ce6dWenbing Li3 years ago22
23Test:
fb2a8c28Wenbing Li2 years ago24
0647ce6dWenbing Li3 years ago25- 'pip install -r requirements-dev.txt' to install pip packages for development.
26- run `pytest test` in the project root directory.
27
28For a complete list of verified build configurations see [here](<./ci_matrix.md>)
29
30## Java package
fb2a8c28Wenbing Li2 years ago31
0647ce6dWenbing Li3 years ago32`bash ./build.sh -DOCOS_BUILD_JAVA=ON` to build jar package in out/<OS>/Release folder
33
34## Android package
fb2a8c28Wenbing Li2 years ago35
0647ce6dWenbing Li3 years ago36- pre-requisites: [Android Studio](https://developer.android.com/studio)
37
63fe165eEdward Chen3 years ago38Use `./tools/android/build_aar.py` to build an Android AAR package.
0647ce6dWenbing Li3 years ago39
40## iOS package
fb2a8c28Wenbing Li2 years ago41
63fe165eEdward Chen3 years ago42Use `./tools/ios/build_xcframework.py` to build an iOS xcframework package.
0647ce6dWenbing Li3 years ago43
5fd6bcf4Sayan Shaw2 years ago44## NuGet package
45
972cb17aWenbing Li2 years ago46In order to build a local NuGet package for testing, run `nuget.exe pack ./nuget/WinOnlyNuget.nuspec` to build a NuGet package for Windows.
fb2a8c28Wenbing Li2 years ago47
972cb17aWenbing Li2 years ago48Note: you might need to update the src paths in the ./nuget/WinOnlyNuget.nuspec file if the appropriate ortextensions.dll files do not exist/are not in the given location.
5fd6bcf4Sayan Shaw2 years ago49
0647ce6dWenbing Li3 years ago50## Web-Assembly
fb2a8c28Wenbing Li2 years ago51
972cb17aWenbing Li2 years ago52ONNXRuntime-Extensions will be built as a static library and linked with ONNXRuntime due to the lack of a good dynamic linking mechanism in WASM. Here are two additional arguments [–-use_extensions and --extensions_overridden_path](https://github.com/microsoft/onnxruntime/blob/860ba8820b72d13a61f0d08b915cd433b738ffdc/tools/ci_build/build.py#L416) on building onnxruntime to include ONNXRuntime-Extensions footprint in the ONNXRuntime package.
0647ce6dWenbing Li3 years ago53
a6012b38Wenbing Li3 years ago54## The C++ shared library
b5dce955Wenbing Li3 years ago55
fb2a8c28Wenbing Li2 years ago56For any alternative scenarios, execute the following commands:
57
58- On Windows: Run `build.bat`.
59- On Unix-based systems: Execute `bash ./build.sh`.
60
972cb17aWenbing Li2 years ago61The generated DLL or library is typically located in the `out/<OS>/<FLAVOR>` directory. To validate the build, utilize the unit tests available in the `test/test_static_test` and `test/shared_test` directories.
a6012b38Wenbing Li3 years ago62
5d558bd3Wenbing Li2 years ago63**CUDA Build**
64The cuda build can be enabled with -DOCOS_USE_CUDA=ON -DCMAKE_CUDA_ARCHITECTURES=<arch>
65
a6012b38Wenbing Li3 years ago66**VC Runtime static linkage**
972cb17aWenbing Li2 years ago67If you want to build the binary with VC Runtime static linkage, please add a parameter _-DCMAKE_MSVC_RUNTIME_LIBRARY="MultiThreaded$<$<CONFIG:Debug>:Debug>"_ when running build.bat
a6012b38Wenbing Li3 years ago68
b5dce955Wenbing Li3 years ago69## Copyright guidance
fb2a8c28Wenbing Li2 years ago70
972cb17aWenbing Li2 years ago71Check this link [here](https://docs.opensource.microsoft.com/releasing/general-guidance/copyright-headers/) for source file copyright header.