Skip to content
  • Pricing
mirjam.el-attal · PyLaia · Published June 22, 2026

German Bastarda XV. century

Text Recognition

Description

This model is based on a collection of manuscripts written by Heinrich Haller between approx. 1460 and 1480 in the carthusian monastery Allerengelberg in Schnals, South-Tyrol, Italy. The texts are written in an early New High German dialect from the Upper German/Bavarian linguistic area. The original manuscripts are held in the collection of the ULB Tirol, where they have been digitised and the digital reproductions can be accessed here: https://ulb-digital.uibk.ac.at/topic/view/8687549 The typeface is a Bastarda (where stems with loops alternate with those without), though the script might be described more as a Hybrida, as well. Virgules are used as punctuation marks, with full stops on the centre line being rare; word breaks are indicated by dashes at the end and beginning of lines. Abbreviations consist almost exclusively of a nasal mark above vowels (vowel + n/m); a nasal mark above n or m (= nn/mm or mb) is rare. As for other special characters, there is generally only a diaeresis above vowels, which is usually simply rendered as an umlaut in transcriptions. The ground truth set for this model was created by the ULB Tirol, using the HTR model Kurrent_1515_ENHG_v3 (ID: 46140) as a basemodel during the training process.

Try this model

Drag an image here

Select a file...

PNG or JPG up to 10 Mb

Wolpi
AI Assistant

By uploading an image, you accept our terms and privacy policy.

German Bastarda XV. century
Use this modelOpen in Transkribus
Very low error rate1.01% CER

Character Error Rate (CER) measures the percentage of characters incorrectly recognised. Lower is better. This model scored 1.01% on its validation set. As a rule of thumb, a CER below 10% is considered good for most handwritten material.

Measured on the model's own validation data. Results on your documents may differ depending on handwriting style, document condition, language, and how closely your material resembles the training data.

Words12,148
Lines1,534
Training Pages60
Model ID590029
Languages
German
Centuries
14th c.15th c.