README.md

metadata

dataset_info:
  features:
    - name: audio
      struct:
        - name: bytes
          dtype: 'null'
        - name: path
          dtype: string
    - name: sentence
      dtype: string
  splits:
    - name: train
      num_bytes: 237278
      num_examples: 3002
    - name: valid
      num_bytes: 29745
      num_examples: 375
    - name: test
      num_bytes: 29683
      num_examples: 376
  download_size: 160291
  dataset_size: 296706
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: valid
        path: data/valid-*
      - split: test
        path: data/test-*