Building the ML SDK Model Converter

The build system must have:

  • CMake 3.25 or later.

  • C/C++ 17 compiler: GCC, or optionally Clang on Linux and MSVC on Windows®.

  • Python 3.10 or later. Required python libraries for building are listed in tooling-requirements.txt.

  • Flatbuffers flatc compiler 25.2.10 or later.

The following dependencies are also needed:

For the preferred dependency versions see the manifest file.

Providing Flatc

There are 3 options for providing the flatc binary and headers.

  1. Using the default path. When the repository is initialized using the repo manifest, the flatbuffers source is checked out in <repo-root>/dependencies/flatbuffers/. The VGF Library cmake scripts automatically find and build flatc in this location.

  2. Providing a custom flatc path. If flatc cannot be found in the default <repo-root>/dependencies/flatbuffers path, you can provide a custom binary file path to the build script using the --flatc-path <path_to_flatc> option, see Building with the script.

  3. Installing flatc to the system. If flatc cannot be found in the default path and no custom path is provided, it will be searched using PATH. To install flatc to the system and make it available on the searchable PATH, see the flatbuffers documentation. For example, on Linux navigate to the flatbuffers checkout location and run the following commands:

cmake -B build -DCMAKE_INSTALL_PREFIX=/usr/local
cmake --build build --target install

Building with the script

Arm® provides a python build script to make build configuration options easily discoverable. When the script is run from a git-repo manifest checkout, the script uses default paths and does not require any additional arguments. Otherwise the paths to the dependencies must be specified.

To build on Linux, run:

SDK_PATH="path/to/sdk"
python3 ${SDK_PATH}/sw/model-converter/scripts/build.py -j $(nproc) \
    --vgf-lib-path ${SDK_PATH}/sw/vgf-lib \
    --flatbuffers-path ${SDK_PATH}/dependencies/flatbuffers \
    --argparse-path ${SDK_PATH}/dependencies/argparse \
    --tosa-mlir-translator-path ${SDK_PATH}/dependencies/tosa_mlir_translator \
    --external-llvm ${SDK_PATH}/dependencies/llvm-project

To build on Windows®, run:

$env:SDK_PATH="path\to\sdk"
$cores = [System.Environment]::ProcessorCount
python3 "$env:SDK_PATH\sw\model-converter\scripts\build.py" -j $cores `
    --vgf-lib-path "$env:SDK_PATH\sw\vgf-lib" `
    --flatbuffers-path "$env:SDK_PATH\dependencies\flatbuffers" `
    --argparse-path "$env:SDK_PATH\dependencies\argparse" `
    --tosa-mlir-translator-path "$env:SDK_PATH\dependencies\tosa_mlir_translator" `
    --external-llvm "$env:SDK_PATH\dependencies\llvm-project"

If the components are in their default locations, it is not necessary to specify the --vgf-lib-path, --flatbuffers-path, --argparse-path, --tosa-mlir-translator-path, and --external-llvm options.

Tests can be enabled and run with --test and linting by --lint. The documentation can be built with --doc. To build the documentation, sphinx and doxygen must be installed on the machine.

You can install the project build artifacts into a specified location by passing the option --install with the required path.

To create an archive with the build artifacts option, add --package. The archive will be stored in the provided location.”

For more information, see the help output:

python3 scripts/build.py --help