metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: duration
dtype: float32
- name: text
dtype: string
- name: source
dtype: 'null'
splits:
- name: train
num_bytes: 12085840949.625
num_examples: 26099
download_size: 12057872775
dataset_size: 12085840949.625
configs:
- config_name: default
data_files:
- split: train
path: data/train-*