---
dataset_info:
  features:
  - name: transcription
    dtype: string
  - name: speaker_id
    dtype: int64
  - name: duration
    dtype: float64
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  splits:
  - name: train
    num_bytes: 20255773612.548
    num_examples: 82371
  - name: valid
    num_bytes: 668655282.274
    num_examples: 2782
  - name: test
    num_bytes: 678701136.122
    num_examples: 2779
  download_size: 19796572013
  dataset_size: 21603130030.944
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: valid
    path: data/valid-*
  - split: test
    path: data/test-*
---