This repository provides the official TensorFlow implementation of the U-GAT-IT paper accepted at ICLR 2020. The model performs unsupervised image-to-image translation using a novel attention module and adaptive layer-instance normalization (AdaLIN). The attention module guides the model to focus on important regions distinguishing between source and target domains, while AdaLIN helps control the balance between shape and texture changes. Pre-trained checkpoints are provided for selfie-to-anime translation.