---
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: transcript
    dtype: string
  splits:
  - name: train
    num_bytes: 3245378777.43
    num_examples: 21690
  download_size: 3671209553
  dataset_size: 3245378777.43
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---