Update README.md
Browse files
README.md
CHANGED
|
@@ -106,6 +106,8 @@ configs:
|
|
| 106 |
|
| 107 |
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 includes both implicit feedback, such as listening events, and explicit feedback, in the form of likes and dislikes. Additionally, it provides distinctive markers for organic versus recommendation-driven interactions, along with precomputed audio embeddings to facilitate content-aware recommendation systems.
|
| 108 |
|
|
|
|
|
|
|
| 109 |
## Key Features
|
| 110 |
|
| 111 |
- 🎵 4.79B user-music interactions (listens, likes, dislikes, unlikes, undislikes)
|
|
|
|
| 106 |
|
| 107 |
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 includes both implicit feedback, such as listening events, and explicit feedback, in the form of likes and dislikes. Additionally, it provides distinctive markers for organic versus recommendation-driven interactions, along with precomputed audio embeddings to facilitate content-aware recommendation systems.
|
| 108 |
|
| 109 |
+
Preprint: https://arxiv.org/abs/2505.22238
|
| 110 |
+
|
| 111 |
## Key Features
|
| 112 |
|
| 113 |
- 🎵 4.79B user-music interactions (listens, likes, dislikes, unlikes, undislikes)
|