How to install onnx-light#

This page aligns the most common onnx-light installation workflows side by side for Python and C++. Each row shows the equivalent commands for local development, faster builds, and basic post-install validation.

Common install patterns#

Install for local development#

pip install -e .[dev] -v

cmake -S . -B build-install \
      -DCMAKE_BUILD_TYPE=Release \
      -DONNX_LIGHT_BUILD_PYTHON=OFF \
      -DCMAKE_INSTALL_PREFIX=/usr/local
cmake --build build-install
cmake --install build-install

Speed up the build#

CMAKE_BUILD_PARALLEL_LEVEL=8 pip install -e .[dev]

cmake --build build-install --parallel 8

Verify the install#

python -c "import onnx_light; print(onnx_light.__version__)"

find_package(onnx_light REQUIRED)
target_link_libraries(my_target PRIVATE onnx_light::onnx_light)

Load a model with the proto-only target#

import onnx_light.onnx as onnxl

model = onnxl.load("model.onnx")

find_package(onnx_light REQUIRED)
add_executable(load_model main.cc)
target_link_libraries(load_model PRIVATE onnx_light::lib_onnx_proto)

The quoted headers come from the installed onnx_light target’s exported include directories.

#include "onnx.h"
#include "stream.h"

#include <exception>
#include <iostream>

int main() {
  try {
    onnx::ModelProto model;
    onnx::utils::FileStream stream("model.onnx");
    onnx::ParseOptions options;
    options.num_threads = 1;
    onnx::ParseModelProtoFromStream(model, stream, options);
    if (model.has_graph()) {
      // ref_graph() is safe after confirming the graph is present.
      std::cout << "Loaded " << model.ref_graph().node_size() << " nodes\n";
    }
    return 0;
  } catch (const std::exception &e) {
    std::cerr << "Error loading model.onnx: " << e.what() << "\n";
    return 1;
  }
}

Notes#

  • Python editable installs pull build dependencies from pyproject.toml and build the extension in place for local development.

  • The C++ install flow exports CMake targets under onnx_light::... and does not require building the Python extension when -DONNX_LIGHT_BUILD_PYTHON=OFF is set.

  • If the C++ library is installed to a non-standard prefix, configure the downstream project with -DCMAKE_PREFIX_PATH=<prefix> before calling find_package(onnx_light REQUIRED).

  • Use onnx_light::onnx_light for higher-level ONNX features such as checker, shape inference, or version conversion. Use onnx_light::lib_onnx_proto when the downstream code only needs proto parsing and serialization.

See also#