- We are attempting to convert an AutoML Edge .pb model which was trained using Google AutoML Vision (Exporting Edge models | AutoML Vision Object Detection | Google Cloud) into a .uff file so that we can perform inference using the TensorRT engine.
- We encountered the above problem (Imgur: The magic of the Internet) during conversion of the frozen .pb into .uff
- Setup details: Ubuntu 18.04, TensorRT-7.0.0.11, Quadro P1000 GPU
Has anyone solved this problem before?