Multilingual and open-source version of ModernBERT trained by LISN/CNRS laboratory on open-source data: wikipedia and OPUS corpora