A machine learning pipeline that automatically restores missing diacritics in Slovak text. The project compares a zero-shot Masked Language Model baseline with a highly accurate, fine-tuned Token Classification model using SlovakBERT.
nlp machine-learning transformers diacritics text-processing wikipedia-dump slovak diacritics-restoration token-classification slovakbert
-
Updated
Mar 17, 2026 - Python