search

marcoslucianops / DeepStream-Yolo

NVIDIA DeepStream SDK 7.0 / 6.4 / 6.3 / 6.2 / 6.1.1 / 6.1 / 6.0.1 / 6.0 / 5.1 implementation for YOLO models
MIT License
1.39k stars 344 forks source link

Cross platform engine export #442

Open tomikpik opened 10 months ago

tomikpik commented 10 months ago

Hi there, could you please give me a hint on how to perform cross-platform engine export? I'd like to export the engine for the Jetson AGX Xavier on the x86 machine with NVIDIA A10 GPU x86 machine.

Thanks :)

marcoslucianops commented 10 months ago

You can use the ONNX model in x86 and Jetson platforms, but the engine isn't possible.

tomikpik commented 10 months ago

So that basically means I need to distribute the model as ONNX with my application and perform the engine export on target device?

marcoslucianops commented 10 months ago

Yes, you will need to build the engine using the distributed ONNX model in each platform.