metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 44100
- name: text
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 15201359165
num_examples: 26916
download_size: 13335555160
dataset_size: 15201359165
configs:
- config_name: default
data_files:
- split: train
path: data/train-*