A collection of models (and the dataset used to train them) that are trained for summarizing arabic text.