Audio Generation on Classical music, 5 seconds at 12 kHz

Metric: Bits per byte (higher is better)

LeaderboardDataset
Loading chart...
#ModelBits per byteExtra DataPaperDateCode
1VAB-Encodec (Ours)40NoFrom Vision to Audio and Beyond: A Unified Model...2024-09-27Code
2Sparse Transformer 152M (strided)1.97NoGenerating Long Sequences with Sparse Transformers2019-04-23Code