---
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
---