metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 20988505008.875
num_examples: 218521
download_size: 19838852943
dataset_size: 20988505008.875