README.md

metadata

license: apache-2.0
dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: transcription
      dtype: string
  splits:
    - name: train
      num_bytes: 54770494383.876
      num_examples: 278766
    - name: validation
      num_bytes: 7864058140.98
      num_examples: 34726
    - name: test
      num_bytes: 7572538413.28
      num_examples: 35166
  download_size: 74257809304
  dataset_size: 70207090938.136
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*