README.md

metadata

dataset_info:
  features:
    - name: sentence
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: prediction_language
      dtype: string
    - name: prediction
      dtype: string
  splits:
    - name: test
      num_bytes: 459322390.375
      num_examples: 3901
  download_size: 454570104
  dataset_size: 459322390.375
configs:
  - config_name: default
    data_files:
      - split: test
        path: data/test-*