German Electra Model

The German Electra model: german-nlp-group/electra-base-german-uncased

We released an improved German Electra model. It was trained on a total of 1,500,000 steps.

Uncase and Umlauts (‘Ö’, ‘Ä’, ‘Ü’)

This model is uncased. This helps especially for domains where colloquial terms with uncorrect capitalization is often used.

The special characters ‘ö’, ‘ü’, ‘ä’ are included through the strip_accent=False option, as this leads to an improved precision.


This model was trained and open sourced in conjunction with the German NLP Group in equal parts by: