A Korean sentence BERT model fine-tuned on KLUE-NLI and augmented STS datasets for semantic textual similarity.