README.md

metadata

dataset_info:
  features:
    - name: speaker
      dtype: string
    - name: audio_path
      dtype: string
  splits:
    - name: train_small
      num_bytes: 6391000
      num_examples: 83000
    - name: test
      num_bytes: 1856610
      num_examples: 26523
    - name: train_unclean
      num_bytes: 30364189
      num_examples: 384347
    - name: train
      num_bytes: 11463447
      num_examples: 161457
  download_size: 5855968
  dataset_size: 50075246
configs:
  - config_name: default
    data_files:
      - split: train_small
        path: data/train_small-*
      - split: test
        path: data/test-*
      - split: train_unclean
        path: data/train_unclean-*
      - split: train
        path: data/train-*