This model includes Transkribus German Kurrent M2 and some sets containing german language Latin script. It works best for documents/collections with Kurrent/Sütterlin and Latin script mixed. For only Kurrent or only Latin script the dedicated models might work better, so please test and compare before use.
This is a generic model trained on a diverse dataset. Such models provide good results without the need for any extra training work. However, the best results can usually be achieved by training a special model for homogenous material, e. g. texts written by the same person or from a narrow historical period.
Image source: Frey, Hans ( vermutlich 1885-1886): Geologie, Vorlesung von Albert Heim. (ETH-Bibliothek Zürich, Hs 1290:1, https://doi.org/10.7891/e-manuscripta-666 / Public Domain Mark) (bearbeitet)