Skip to content
  • Pricing
Jonathan Hauller · PyLaia · Published July 20, 2026

Ratsmanuale_1798-1552

Text Recognition

Description

This model was trained on a set of 723 pages from the "Zurich Council Manuals" (Zürcher Ratsmanuale). They contain the minutes of the meetings of the authorities of the old city-state of Zurich until 1798, therefore the model works best on protocols. The data included (train and validation sets) covers a timespan from 1552 to 1798. The language of the protocols is Early New High German (Frühneuhochdeutsch) and the script style is German Kurrent (Frühneuzeitliche Kurrentschrift). Further parameters: Language: deu - German Base Model: The German Giant I Perform binarization: no Use existing line polygons for training: yes Omit lines by tag: unclear Train Abbrevs: no Train Tags: no

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.

Use this modelOpen in Transkribus
Very low error rate4.79% CER

Character Error Rate (CER) measures the percentage of characters incorrectly recognised. Lower is better. This model scored 4.79% on its validation set. As a rule of thumb, a CER below 10% is considered good for most handwritten material. This is a larger model trained on diverse material, which generally makes it more robust across different handwriting styles. That said, larger training sets also make it harder to push the CER down further.

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.

Words286,717
Lines65,844
Training Pages723
Model ID605869
Languages
German