metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 24000
- name: text
dtype: string
- name: duration
dtype: float64
- name: source
dtype: string
splits:
- name: train
num_bytes: 8610493692
num_examples: 20629
download_size: 8788602599
dataset_size: 8610493692
configs:
- config_name: default
data_files:
- split: train
path: data/train-*