@InProceedings{marmonier-EtAl:2026:wmt,
  author    = {Marmonier, Malik  and  Favre, Alain  and  Sagot, Benoît  and  Bawden, Rachel},
  title     = {A Savoyard Version of the FLORES+ Corpus},
  booktitle      = {Proceedings of the Eleventh Conference on Machine Translation},
  month          = {October},
  year           = {2026},
  address        = {Budapest, Hungary},
  publisher      = {Association for Computational Linguistics},
  pages     = {2526--2545},
  abstract  = {We present a Savoyard version of the FLORES+ corpus, a contribution to the Open Language Data Initiative (OLDI) shared task at WMT 2026. Savoyard, an endangered Arpitan dialect of the Gallo-Romance language family, is currently undergoing active revitalization. We give an overview of its linguistic history, describe our data collection process, and report baseline scores for a range of systems on the resulting French-Savoyard evaluation set as a point of comparison for future work. We hope that this dataset will stimulate further data collection efforts for Savoyard and the broader Arpitan dialect continuum, supporting the development of NLP resources for these endangered languages and contributing to the preservation of humanity's linguistic heritage.},
  url       = {https://aclanthology.org/2026.wmt-1.197}
}