---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: transcription
    dtype: string
  - name: duration
    dtype: float64
  splits:
  - name: train
    num_bytes: 2787784507.326612
    num_examples: 25498
  - name: eval
    num_bytes: 298604497.95664775
    num_examples: 2810
  - name: test
    num_bytes: 552707369.3965499
    num_examples: 4741
  download_size: 3602942250
  dataset_size: 3639096374.6798096
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: eval
    path: data/eval-*
  - split: test
    path: data/test-*
---