ONNX Runtime
Run exported ONNX models with a common inference interface and hardware-specific execution providers.
Published by Microsoft
How to use it
Use ONNX Runtime to run an exported model from Python, C++, C#, or a supported mobile client. Start with the CPU provider and compare the model's output against the original framework before enabling an accelerator.
Providers are distributed in different packages and support different operators. Check the provider's platform and driver requirements, then verify the active providers and any CPU fallback on the target device.
- Hardware
- CPU, NVIDIA CUDA, Apple Core ML, DirectML
- Model formats
- ONNX, ORT model
- Operating systems
- Linux, Windows, macOS, Android, iOS
- Access
- Runs locally. Install a build for the intended execution provider.