metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio
dtype: audio
- name: source
dtype: string
- name: ref_audio
dtype: audio
splits:
- name: train
num_bytes: 162478640
num_examples: 443
download_size: 159020897
dataset_size: 162478640
configs:
- config_name: default
data_files:
- split: train
path: data/train-*