This model, developed by Microsoft, is designed for document understanding tasks, such as form understanding and receipt understanding. It incorporates layout information into its training, making it particularly suited to tasks where the visual presentation of a document is important.