Post training 4-bit quantization of convolutional networks for rapid-deployment

Open in new window