README.md

metadata

dataset_info:
  features:
    - name: client_id
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: sentence_id
      dtype: string
    - name: sentence
      dtype: string
    - name: up_votes
      dtype: int64
    - name: down_votes
      dtype: int64
    - name: age
      dtype: string
    - name: gender
      dtype: string
    - name: accents
      dtype: string
    - name: variant
      dtype: float64
    - name: locale
      dtype: string
  splits:
    - name: train
      num_bytes: 3160485788.475912
      num_examples: 30769
    - name: test
      num_bytes: 257121524.65190506
      num_examples: 2153
  download_size: 3263975273
  dataset_size: 3417607313.127817
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
license: cc-by-4.0
task_categories:
  - automatic-speech-recognition
  - text-to-speech
language:
  - zh