I have discussed a brief review on strategies to improve the training of Generative Adversarial Networks (GAN) at the ML_INFN meeting. Here are the slides [public access].