Gabelsberger Natural

The model is based on various manuscripts written in Gabelsberger shorthand. The training data includes a part of the diaries and supplementary sheets of Michael Cardinal von Faulhaber (https://www.faulhaber-edition.de/index.html), a part of the minutes of the Council of Ministers from 1900(mrp (oeaw.ac.at)), some war diaries of Carl Schmitt (Arbeitsgruppe Carl Schmitt / Carl Schmitt Tagebücher · GitLab) and some ego documents from the private estates of various people. The model was trained by Milanka Matić-Chalkitis as part of the MultiHTR project (project leader: Prof. Dr. Achim Rabus) at the Department of Slavic Languages and Literatures of the University of Freiburg (Germany). We would like to thank Dr. Philipp Gahn and Dr. Michael Pilarski (Institute of Contemporary History, Munich-Berlin), Dr. Stephan Kurz (Austrian Academy of Sciences) and Prof. Dr. Florian Meinl (University of Göttingen) and his team for kindly providing the GT data and for their close cooperation. The model is intended to offer assistance to those who have little or no expertise in Gabelsberger shorthand, but who wish to explore the context of their documents themselves. It should be noted that, although this model is based on a variety of different training data, the automatic transcription of individual manuscripts varies in quality. We recommend comparing the transcription results of the model with and without the language model.