README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: logits
      sequence: float32
  splits:
    - name: train
      num_bytes: 1198241856.0912256
      num_examples: 37376
  download_size: 1117454012
  dataset_size: 1198241856.0912256
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*