metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 44100
- name: speaker
dtype: string
splits:
- name: train
num_bytes: 22104265
num_examples: 4
download_size: 21444820
dataset_size: 22104265
configs:
- config_name: default
data_files:
- split: train
path: data/train-*