onnxruntime/BUILD.md

Build ONNX Runtime

Supported architectures

	x86_32	x86_64	ARM32v7	ARM64
Windows	YES	YES	YES	YES
Linux	YES	YES	YES	YES
Mac OS X	NO	YES	NO	NO

Supported dev environments

OS	Supports CPU	Supports GPU	Notes
Windows 10	YES	YES	Must use VS 2017 or the latest VS2015
Windows 10 Subsystem for Linux	YES	NO
Ubuntu 16.x	YES	YES	Also supported on ARM32v7 (experimental)

• Red Hat Enterprise Linux and CentOS are not supported.
• Other version of Ubuntu might work but we don't support them officially.
• GCC 4.x and below are not supported.

OS/Compiler Matrix:

OS/Compiler	Supports VC	Supports GCC
Windows 10	YES	Not tested
Linux	NO	YES(gcc>=5.0)

ONNX Runtime python binding only supports Python 3.5, 3.6 and 3.7.

Build

1. Checkout the source tree:

   git clone --recursive https://github.com/Microsoft/onnxruntime
   cd onnxruntime
   

2. Install cmake-3.13 or better from https://cmake.org/download/.
3. (optional) Install protobuf 3.6.1 from source code (cmake/external/protobuf). CMake flag protobuf_BUILD_SHARED_LIBS must be turned OFF on Windows and turned ON on Linux. After the installation, you should have the 'protoc' executable in your PATH. On Linux it is recommended to run ldconfig to make sure protobuf libraries are found.
4. If you installed your protobuf in a non standard location it would be helpful on Linux build to set the following env var: export CMAKE_ARGS="-DONNX_CUSTOM_PROTOC_EXECUTABLE=full path to protoc" so ONNX build can find it. On Linux also run ldconfig <protobuf lib folder path> so the linker can find protobuf libraries.
5. (optional) Install onnx from source code (cmake/external/onnx)

   export ONNX_ML=1
   python3 setup.py bdist_wheel
   pip3 install --upgrade dist/*.whl
   

6. Run ./build.sh --config RelWithDebInfo --build_wheel for Linux (or build.bat --config RelWithDebInfo --build_wheel for Windows). Upon successful build you should be able to find the wheel under dist folder.

The build script runs all unit tests by default (for native builds and skips tests by default for cross-compiled builds).

The complete list of build options can be found by running ./build.sh (or ./build.bat) --help

Build x86

1. For Windows, just add --x86 argument when launching build.bat
2. For Linux, it must be built out of a x86 os, --x86 argument also needs be specified to build.sh

Build ONNX Runtime Server on Linux

1. In the ONNX Runtime root folder, run ./build.sh --config RelWithDebInfo --build_server --use_openmp --parallel

Build/Test Flavors for CI

CI Build Environments

Build Job Name	Environment	Dependency	Test Coverage	Scripts
Linux_CI_Dev	Ubuntu 16.04	python=3.5	Unit tests; ONNXModelZoo	script
Linux_CI_GPU_Dev	Ubuntu 16.04	python=3.5; nvidia-docker	Unit tests; ONNXModelZoo	script
Windows_CI_Dev	Windows Server 2016	python=3.5	Unit tests; ONNXModelZoo	script
Windows_CI_GPU_Dev	Windows Server 2016	cuda=9.1; cudnn=7.1; python=3.5	Unit tests; ONNXModelZoo	script

Additional Build Flavors

The complete list of build flavors can be seen by running ./build.sh --help or ./build.bat --help. Here are some common flavors.

Windows CUDA Build

ONNX Runtime supports CUDA builds. You will need to download and install CUDA and CUDNN.

ONNX Runtime is built and tested with CUDA 9.1 and CUDNN 7.1 using the Visual Studio 2017 14.11 toolset (i.e. Visual Studio 2017 v15.3). CUDA versions from 9.1 up to 10.0, and CUDNN versions from 7.1 up to 7.4 should also work with Visual Studio 2017.

• The path to the CUDA installation must be provided via the CUDA_PATH environment variable, or the --cuda_home parameter.
• The path to the CUDNN installation (include the cuda folder in the path) must be provided via the CUDNN_PATH environment variable, or --cudnn_home parameter. The CUDNN path should contain bin, include and lib directories.
• The path to the CUDNN bin directory must be added to the PATH environment variable so that cudnn64_7.dll is found.

You can build with:

./build.sh --use_cuda --cudnn_home /usr --cuda_home /usr/local/cuda (Linux)
./build.bat --use_cuda --cudnn_home <cudnn home path> --cuda_home <cuda home path> (Windows)

Depending on compatibility between the CUDA, CUDNN, and Visual Studio 2017 versions you are using, you may need to explicitly install an earlier version of the MSVC toolset.

• CUDA 10.0 is known to work with toolsets from 14.11 up to 14.16 (Visual Studio 2017 15.9), and should continue to work with future Visual Studio versions
  • https://devblogs.microsoft.com/cppblog/cuda-10-is-now-available-with-support-for-the-latest-visual-studio-2017-versions/
• CUDA 9.2 is known to work with the 14.11 MSVC toolset (Visual Studio 15.3 and 15.4)

To install the 14.11 MSVC toolset, see https://blogs.msdn.microsoft.com/vcblog/2017/11/15/side-by-side-minor-version-msvc-toolsets-in-visual-studio-2017/

To use the 14.11 toolset with a later version of Visual Studio 2017 you have two options:

1. Setup the Visual Studio environment variables to point to the 14.11 toolset by running vcvarsall.bat, prior to running the build script

   • e.g. if you have VS2017 Enterprise, an x64 build would use the following command "C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\VC\Auxiliary\Build\vcvarsall.bat" amd64 -vcvars_ver=14.11
   • For convenience, build.amd64.1411.bat will do this and can be used in the same way as build.bat.
     • e.g. .\build.amd64.1411.bat --use_cuda

2. Alternatively if you have CMake 3.12 or later you can specify the toolset version via the --msvc_toolset build script parameter.

   • e.g. .\build.bat --msvc_toolset 14.11

Side note: If you have multiple versions of CUDA installed on a Windows machine and are building with Visual Studio, CMake will use the build files for the highest version of CUDA it finds in the BuildCustomization folder. e.g. C:\Program Files (x86)\Microsoft Visual Studio\2017\Enterprise\Common7\IDE\VC\VCTargets\BuildCustomizations. If you want to build with an earlier version, you must temporarily remove the 'CUDA x.y.*' files for later versions from this directory.

MKL-DNN/MKLML

To build ONNX Runtime with MKL-DNN support, build it with ./build.sh --use_mkldnn To build ONNX Runtime using MKL-DNN built with dependency on MKL small libraries, build it with ./build.sh --use_mkldnn --use_mklml

nGraph

ONNX runtime with nGraph as an execution provider (released as preview) can be built on Linux as follows : ./build.sh --use_ngraph. Similarly, on Windows use .\build.bat --use_ngraph.

TensorRT

ONNX Runtime supports the TensorRT execution provider (released as preview). You will need to download and install CUDA, CUDNN and TensorRT.

The TensorRT execution provider for ONNX Runtime is built and tested with CUDA 9.0/CUDA 10.0, CUDNN 7.1 and TensorRT 5.0.2.6.

• The path to the CUDA installation must be provided via the CUDA_PATH environment variable, or the --cuda_home parameter. The CUDA path should contain bin, include and lib directories.
• The path to the CUDA bin directory must be added to the PATH environment variable so that nvcc is found.
• The path to the CUDNN installation (path to folder that contains libcudnn.so) must be provided via the CUDNN_PATH environment variable, or --cudnn_home parameter.
• The path to TensorRT installation must be provided via the --tensorrt_home parameter.

You can build from source on Linux by using the following cmd from the onnxruntime directory:

./build.sh --cudnn_home <path to CUDNN e.g. /usr/lib/x86_64-linux-gnu/> --cuda_home <path to folder for CUDA e.g. /usr/local/cuda> --use_tensorrt --tensorrt_home <path to TensorRT home> (Linux)

OpenVINO Build

ONNX Runtime supports OpenVINO Execution Provider to enable deep learning inference using Intel^® OpenVINO^TM Toolkit. This execution provider supports several Intel hardware device types - CPU, integrated GPU, Intel^® Movidius^TM VPUs and Intel^® Vision accelerator Design with 8 Intel Movidius^TM MyriadX VPUs.

The OpenVINO Execution Provider can be built using the following commands:

• Install the OpenVINO 2018 R5.0.1 release along with its dependencies from (https://software.intel.com/en-us/openvino-toolkit).

• Install the model optimizer prerequisites for ONNX by running <openvino_install_dir>/deployment_tools/model_optimizer/install_prerequisites/install_prerequisites_onnx.sh

• Initialize the OpenVINO environment by running the setupvars.sh in <openvino_install_directory>/bin using the below command:

  source setupvars.sh

• To configure Intel^® Processor Graphics(GPU), please follow the installation steps from (https://docs.openvinotoolkit.org/2018_R5/_docs_install_guides_installing_openvino_linux.html#GPU-steps)

• To configure Intel^® Movidius^TM USB, please follow the getting started guide from (https://docs.openvinotoolkit.org/2018_R5/_docs_install_guides_installing_openvino_linux.html#Movidius-steps)

• To configure Intel^® Vision Accelerator Design based on 8 Movidius^TM MyriadX VPUs, please follow the configuration guide from (https://docs.openvinotoolkit.org/2018_R5/_docs_install_guides_installing_openvino_linux.html#Vision-Accelerator-Design-steps)

• Build ONNX Runtime using the below command.

  ./build.sh --config RelWithDebInfo --use_openvino <hardware_option>

  --use_openvino: Builds the OpenVINO Execution Provider in ONNX Runtime.

  <hardware_option>: Specifies the hardware target for building OpenVINO Execution Provider. Below are the options for different Intel target devices.

Hardware Option	Target Device
CPU_FP32	Intel® CPUs
GPU_FP32	Intel® Integrated Graphics
GPU_FP16	Intel® Integrated Graphics with FP16 quantization of models
MYRIAD_FP16	Intel® MovidiusTM USB sticks
VAD-R_FP16	Intel® Vision Accelerator Design based on 8 MovidiusTM MyriadX VPUs

For more information on OpenVINO Execution Provider's ONNX Layer support, Topology support, and Intel hardware enabled, please refer to the document OpenVINO-ExecutionProvider.md in $onnxruntime_root/docs/execution_providers

OpenBLAS

Windows

Instructions how to build OpenBLAS for windows can be found here https://github.com/xianyi/OpenBLAS/wiki/How-to-use-OpenBLAS-in-Microsoft-Visual-Studio#build-openblas-for-universal-windows-platform.

Once you have the OpenBLAS binaries, build ONNX Runtime with ./build.bat --use_openblas

Linux

For Linux (e.g. Ubuntu 16.04), install libopenblas-dev package sudo apt-get install libopenblas-dev and build with ./build.sh --use_openblas

OpenMP

./build.sh --use_openmp (for Linux)
./build.bat --use_openmp (for Windows)

Build with Docker on Linux

Install Docker: https://docs.docker.com/install/

CPU

cd tools/ci_build/github/linux/docker
docker build -t onnxruntime_dev --build-arg OS_VERSION=16.04 -f Dockerfile.ubuntu .
docker run --rm -it onnxruntime_dev /bin/bash

GPU

If you need GPU support, please also install:

1. nvidia driver. Before doing this please add nomodeset rd.driver.blacklist=nouveau to your linux kernel boot parameters.
2. nvidia-docker2: Install doc

To test if your nvidia-docker works:

docker run --runtime=nvidia --rm nvidia/cuda nvidia-smi

Then build a docker image. We provided a sample for use:

cd tools/ci_build/github/linux/docker
docker build -t cuda_dev -f Dockerfile.ubuntu_gpu .

Then run it

./tools/ci_build/github/linux/run_dockerbuild.sh

ARM Builds

We have experimental support for Linux ARM builds. Windows on ARM is well tested.

Cross compiling for ARM with Docker (Linux/Windows - FASTER, RECOMMENDED)

This method allows you to compile using a desktop or cloud VM. This is much faster than compiling natively and avoids out-of-memory issues that may be encountered when on lower-powered ARM devices. The resulting ONNX Runtime Python wheel (.whl) file is then deployed to an ARM device where it can be invoked in Python 3 scripts.

The Dockerfile used in these instructions specifically targets Raspberry Pi 3/3+ running Raspbian Stretch. The same approach should work for other ARM devices, but may require some changes to the Dockerfile such as choosing a different base image (Line 0: FROM ...).

1. Install DockerCE on your development machine by following the instructions here

2. Create an empty local directory

   mkdir onnx-build
   cd onnx-build
   

3. Save the Dockerfile to your new directory

   • Dockerfile.arm32v7

4. Run docker build

   This will build all the dependencies first, then build ONNX Runtime and its Python bindings. This will take several hours.

   docker build -t onnxruntime-arm32v7 -f Dockerfile.arm32v7 .
   

5. Note the full path of the .whl file

   • Reported at the end of the build, after the # Build Output line.
   • It should follow the format onnxruntime-0.3.0-cp35-cp35m-linux_armv7l.whl, but version number may have changed. You'll use this path to extract the wheel file later.

6. Check that the build succeeded

   Upon completion, you should see an image tagged onnxruntime-arm32v7 in your list of docker images:

   docker images
   

7. Extract the Python wheel file from the docker image

   (Update the path/version of the .whl file with the one noted in step 5)

   docker create -ti --name onnxruntime_temp onnxruntime-arm32v7 bash
   docker cp onnxruntime_temp:/code/onnxruntime/build/Linux/MinSizeRel/dist/onnxruntime-0.3.0-cp35-cp35m-linux_armv7l.whl .
   docker rm -fv onnxruntime_temp
   

   This will save a copy of the wheel file, onnxruntime-0.3.0-cp35-cp35m-linux_armv7l.whl, to your working directory on your host machine.

8. Copy the wheel file (onnxruntime-0.3.0-cp35-cp35m-linux_armv7l.whl) to your Raspberry Pi or other ARM device

9. On device, install the ONNX Runtime wheel file

   sudo apt-get update
   sudo apt-get install -y python3 python3-pip
   pip3 install numpy
   
   # Install ONNX Runtime
   # Important: Update path/version to match the name and location of your .whl file
   pip3 install onnxruntime-0.3.0-cp35-cp35m-linux_armv7l.whl
   

10. Test installation by following the instructions here

Cross compiling on Linux (without Docker)

1. Get the corresponding toolchain. For example, if your device is Raspberry Pi and the device os is Ubuntu 16.04, you may use gcc-linaro-6.3.1 from https://releases.linaro.org/components/toolchain/binaries

2. Setup env vars

      export PATH=/opt/gcc-linaro-6.3.1-2017.05-x86_64_arm-linux-gnueabihf/bin:$PATH
      export CC=arm-linux-gnueabihf-gcc
      export CXX=arm-linux-gnueabihf-g++
   

3. Get a pre-compiled protoc:

   You may get it from https://github.com/protocolbuffers/protobuf/releases/download/v3.6.1/protoc-3.6.1-linux-x86_64.zip . Please unzip it after downloading.

4. (optional) Setup sysroot for enabling python extension. (TODO: will add details later)

5. Save the following content as tool.cmake

   set(CMAKE_SYSTEM_NAME Linux)
   set(CMAKE_SYSTEM_PROCESSOR arm)
   set(CMAKE_CXX_COMPILER arm-linux-gnueabihf-c++)
   set(CMAKE_C_COMPILER arm-linux-gnueabihf-gcc)
   set(CMAKE_FIND_ROOT_PATH_MODE_PROGRAM NEVER)
   set(CMAKE_FIND_ROOT_PATH_MODE_LIBRARY ONLY)
   set(CMAKE_FIND_ROOT_PATH_MODE_INCLUDE ONLY)
   set(CMAKE_FIND_ROOT_PATH_MODE_PACKAGE ONLY)
   

6. Append -DONNX_CUSTOM_PROTOC_EXECUTABLE=/path/to/protoc -DCMAKE_TOOLCHAIN_FILE=path/to/tool.cmake to your cmake args, run cmake and make to build it.

Native compiling on Linux ARM device (SLOWER)

Docker build runs on a Raspberry Pi 3B with Raspbian Stretch Lite OS (Desktop version will run out memory when linking the .so file) will take 8-9 hours in total.

sudo apt-get update
sudo apt-get install -y \
    sudo \
    build-essential \
    curl \
    libcurl4-openssl-dev \
    libssl-dev \
    wget \
    python3 \
    python3-pip \
    python3-dev \
    git \
    tar

pip3 install --upgrade pip
pip3 install --upgrade setuptools
pip3 install --upgrade wheel
pip3 install numpy

# Build the latest cmake
mkdir /code
cd /code
wget https://cmake.org/files/v3.12/cmake-3.12.3.tar.gz;
tar zxf cmake-3.12.3.tar.gz

cd /code/cmake-3.12.3
./configure --system-curl
make
sudo make install

# Prepare onnxruntime Repo
cd /code
git clone --recursive https://github.com/Microsoft/onnxruntime

# Start the basic build
cd /code/onnxruntime
./build.sh --config MinSizeRel --arm --update --build

# Build Shared Library
./build.sh --config MinSizeRel --arm --build_shared_lib

# Build Python Bindings and Wheel
./build.sh --config MinSizeRel --arm --enable_pybind --build_wheel

# Build Output
ls -l /code/onnxruntime/build/Linux/MinSizeRel/*.so
ls -l /code/onnxruntime/build/Linux/MinSizeRel/dist/*.whl

Cross compiling on Windows

Using Visual C++ compilers

1. Download and install Visual C++ compilers and libraries for ARM(64). If you have Visual Studio installed, please use the Visual Studio Installer (look under the section Individual components after choosing to modify Visual Studio) to download and install the corresponding ARM(64) compilers and libraries.

2. Use build.bat and specify --arm or --arm64 as the build option to start building. Preferably use Developer Command Prompt for VS or make sure all the installed cross-compilers are findable from the command prompt being used to build using the PATH environmant variable.

Using other compilers

(TODO)

Android Builds

Cross compiling on Linux

1. Get Android NDK from https://developer.android.com/ndk/downloads. Please unzip it after downloading.

2. Get a pre-compiled protoc:

   You may get it from https://github.com/protocolbuffers/protobuf/releases/download/v3.6.1/protoc-3.6.1-linux-x86_64.zip. Please unzip it after downloading.

3. Denote the unzip destination in step 1 as $ANDROID_NDK, append -DCMAKE_TOOLCHAIN_FILE=$ANDROID_NDK/build/cmake/android.toolchain.cmake -DANDROID_ABI=arm64-v8a -DONNX_CUSTOM_PROTOC_EXECUTABLE=path/to/protoc to your cmake args, run cmake and make to build it.

Note: For 32-bit devices, replace -DANDROID_ABI=arm64-v8a to -DANDROID_ABI=armeabi-v7a.