Industrial-scale music recommendation dataset with organic/recommendation interactions and audio embeddings
📌 Overview • 🔑 Key Features • 📊 Statistics • 📝 Format • 🏆 Benchmark • ⬇️ Download • ❓ FAQ
The Yambda-5B dataset is a large-scale open database comprising 4.79 billion user-item interactions
collected from 1 million users and spanning 9.39 million tracks. The dataset… See the full description on the dataset page:
https://huggingface.co/datasets/yandex/yambda.
