metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 16512354260
num_examples: 96000
- name: test
num_bytes: 4128088565
num_examples: 24000
download_size: 20017668018
dataset_size: 20640442825
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*