metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 97440783596.831
num_examples: 20507
download_size: 127972695307
dataset_size: 97440783596.831
configs:
- config_name: default
data_files:
- split: train
path: data/train-*