metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: transcript
dtype: string
splits:
- name: train
num_bytes: 3245378777.43
num_examples: 21690
download_size: 3671209553
dataset_size: 3245378777.43
configs:
- config_name: default
data_files:
- split: train
path: data/train-*