Skip to content
Diacritic restoration tool for Croatian, Serbian and Slovene
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.


Diacritic restoration tool for Croatian, Serbian and Slovene

The tool expects tokenised / verticalised text. If not defined differently, the last column will be rediacritised.

$ echo 'Jucer nisam bio u skoli' | ../reldi-tokeniser/ hr | ./ hr	Jucer	Jučer	nisam	nisam	bio	bio	u	u	skoli	školi


python2.7 kenlm ( if you want to use language models for contextual disambiguation

Contextual disambiguation

There is an option for using large language models for contextual disambiguation. You can download the models from

If you have the language models in your working directory, you can disambiguate in context by adding the -l flag.


  author = {Nikola Ljubešić and Tomaž Erjavec and Darja Fišer},
  title = {Corpus-Based Diacritic Restoration for South Slavic Languages},
  booktitle = {Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC 2016)},
  year = {2016},
  month = {may},
  date = {23-28},
  location = {Portorož, Slovenia},
  editor = {Nicoletta Calzolari (Conference Chair) and Khalid Choukri and Thierry Declerck and Sara Goggi and Marko Grobelnik and Bente Maegaard and Joseph Mariani and Helene Mazo and Asuncion Moreno and Jan Odijk and Stelios Piperidis},
  publisher = {European Language Resources Association (ELRA)},
  address = {Paris, France},
  isbn = {978-2-9517408-9-1},
  language = {english}
You can’t perform that action at this time.