metadata
dataset_info:
features:
- name: caption
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: test
num_bytes: 4897858523.041246
num_examples: 15463
- name: train
num_bytes: 58584881989.45539
num_examples: 183631
download_size: 61916988256
dataset_size: 63482740512.496635
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: train
path: data/train-*