DeepStream for TensorRT 8 for dGPU and Jetson

I would like to use TensorRT 8 with DeepStream 5.1, to use Efficient Det model in the DeepStream pipeline. Is it possible to integrate libnvinfer8 with DeepStream as it currently tires to find libnvinfer7.

We never experience for that, I would suggest you don’t do this since there are some libs complied with TRT 7.x, so there may cause some unexpected behaviour if you use TRT8

Got it. I have compiled TRT 7 with EfficientNMSPlugin so now it supports the EfficientDet model.

Great work.