How to use it in python after TRT Build?

[davodogster]

Hi I just did the TRT build using
./build.sh --cudnn_home /usr/local/cuda/ --cuda_home /usr/local/cuda-11.6 --use_tensorrt --tensorrt_home /home/scion.local/davidsos/Documents/onnxruntime/TensorRT-8.4.1.5

It passed most of the tests but failed the QOrdered tests.

But how does python recognize it when I run import onnxruntime?

I have a Pytorch lightning model I want to convert to ONNX and then to TRT for Onnxruntime-GPU TRT, CUDA Execution accelerated inference large scale (1M+ images). Is there a tutorial to convert a pytorch lightning model to ONNX-TRT then use it for TRT inference?
Thanks, Sam @abudup