Code for our paper at ECCV 2020: Post-Training Piecewise Linear Quantization for Deep Neural Networks