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: 204206434
num_examples: 671
download_size: 198986468
dataset_size: 204206434
configs:
- config_name: default
data_files:
- split: train
path: data/train-*