This model is one of the artifacts of the paper Massively Multilingual Lexical Specialization of Multilingual Transformers.

It was obtained by fine-tuning the representations of xlm-roberta-base on the dataset babelbert-dataset.