metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: text
dtype: string
- name: audio
sequence: float32
splits:
- name: train
num_bytes: 19504038988
num_examples: 100000
download_size: 17224747032
dataset_size: 19504038988