---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: transcripts
    dtype: string
  splits:
  - name: train
    num_bytes: 269069437737.89664
    num_examples: 1472446
  - name: test
    num_bytes: 34539219446.625694
    num_examples: 184056
  - name: valid
    num_bytes: 35417778368.76969
    num_examples: 184056
  download_size: 334429258612
  dataset_size: 339026435553.29205
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: valid
    path: data/valid-*
---