Hi, I’m loading a 52 MB uff model with uffParser in C++ and it spends about 80 seconds to do this job. My code is based on the sample sampleUffMNIST, and the issue is in the function loadModelAndCreateEngine in the code line parser->parse(uffFile, *network, nvinfer1::DataType::kFLOAT).
How could I solve this issue?. I’m using TensorRT-188.8.131.52 on windows with Visual Studio 2015.
I appreciate any help.