Improving embedding with contrastive fine-tuning on small datasets with expert-augmented scores

Open in new window