README.md

---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: song_id
    dtype: string
  - name: structure
    dtype: string
  - name: start_time
    dtype: float64
  - name: end_time
    dtype: float64
  - name: tempos
    dtype: string
  - name: keys
    dtype: string
  - name: instruments
    dtype: string
  - name: genres
    dtype: string
  - name: moods
    dtype: string
  - name: metadata
    dtype: string
  splits:
  - name: train
    num_bytes: 797925040.0
    num_examples: 831
  - name: eval
    num_bytes: 15363261.0
    num_examples: 16
  download_size: 801646719
  dataset_size: 813288301.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: eval
    path: data/eval-*
---