vai_q_pytorch is designed to work as a PyTorch plugin. Xilinx provides the simplest APIs to introduce the FPGA-friendly quantization feature. For a well-defined model, you only need to add a few lines to get a quantize model object. To do so, follow these steps: