problem with int8 quantization of tensorrt for models trained with adam optimizer

problem with int8 quantization of tensorrt for models trained with adam optimizer #3284