Augment, Drop & Swap: Improving Diversity in LLM Captions for Efficient Music-Text Representation Learning

Open in new window