Datasets for benchmark testing embedding models in Azerbaijani for sentence similarity tasks.