metadata
dataset_info:
features:
- name: audio_path
dtype: string
- name: text
sequence: string
- name: start_times
sequence: 'null'
splits:
- name: train
num_bytes: 551891
num_examples: 115
download_size: 282967
dataset_size: 551891
configs:
- config_name: default
data_files:
- split: train
path: data/train-*