Convert deep learning frameworks such as PyTorch and TensorFlow/Keras and run them directly from c++ - github

Windows distributions of this project may collect usage data and send it to Microsoft

What is ONNX C++ Runtime?

It is simply the C++ API of ONNX Runtime that allows you to:

• Load ONNX models in a C++ program
• Run inference on them
• Feed inputs and fetch outputs
• Configure execution providers (CPU, CUDA, etc.)
• Optimize models

Essentially, you can embed fast ML inference inside native C++ applications.