---
dataset_info:
  features:
  - name: audio
    struct:
    - name: bytes
      dtype: binary
  - name: transcript
    dtype: string
  splits:
  - name: train
    num_bytes: 11455747249
    num_examples: 20616
  download_size: 8862384047
  dataset_size: 11455747249
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---