metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: text
dtype: string
splits:
- name: train
num_bytes: 5198083099.513551
num_examples: 21242
- name: test
num_bytes: 460428062.6684487
num_examples: 1848
download_size: 5668661981
dataset_size: 5658511162.181999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*