ONNX Runtime is a high-performance inference and training graph execution engine for deep learning models.
ONNX Runtime's C, C++ APIs offer an easy to use interface to onboard and execute onnx models.