A sentence embedding model based on RoBERTa-large, fine-tuned for semantic textual similarity and clustering.