tokenized/mls/metadata.json

{
  "dataset": "mls",
  "total_samples": 258213,
  "total_hours": 1076.58,
  "num_shards_tokenized": 517,
  "num_shards_audio_brut": 517,
  "codec": "mimi",
  "num_codebooks": 8,
  "frame_rate": 12.5,
  "sample_rate": 24000
}