Need to optimize a DL model on a target platform for availability and best performance
User friendly c++ library based on TensorRT
Optimize options : batch size, precision, memory, INT8