Benchmark inference speed of CNNs with various quantization methods in Pytorch+TensorRT with Jetson Nano/Xavier - kentaroy47/benchmark-FP32-FP16-INT8-with-TensorRT