metadata
dataset_info:
features:
- name: transcript
dtype: string
- name: audio
dtype: audio
splits:
- name: train
num_bytes: 3642940434
num_examples: 4500
download_size: 3527507569
dataset_size: 3642940434
configs:
- config_name: default
data_files:
- split: train
path: data/train-*