README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: id
      dtype: string
    - name: transcription
      dtype: string
    - name: segment_start
      dtype: int64
    - name: segment_end
      dtype: int64
  splits:
    - name: train
      num_bytes: 4478662375.432
      num_examples: 30616
    - name: validation
      num_bytes: 845929172.888
      num_examples: 5808
    - name: test
      num_bytes: 909076999.72
      num_examples: 5484
  download_size: 6187026845
  dataset_size: 6233668548.04
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*