README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: text
      dtype: string
    - name: duration
      dtype: float64
  splits:
    - name: train
      num_bytes: 4843356891.568
      num_examples: 19568
  download_size: 4460872532
  dataset_size: 4843356891.568
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
language:
  - zgh
  - ber
  - shi