---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: transcripts
    dtype: string
  splits:
  - name: train
    num_bytes: 53133867240.215996
    num_examples: 299636
  - name: test
    num_bytes: 6736967357.531417
    num_examples: 37455
  - name: valid
    num_bytes: 6484620568.886582
    num_examples: 37454
  download_size: 62734833313
  dataset_size: 66355455166.633995
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: valid
    path: data/valid-*
---