A variant of WGAN which uses a gradient penalty term in the loss function to improve training stability. 27.07.2023 17:54 aior