metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 3820438322.1
num_examples: 22500
- name: test
num_bytes: 424493146.9
num_examples: 2500
download_size: 4110984538
dataset_size: 4244931469
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*