metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: score
dtype: float64
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 5346976563.35
num_examples: 41433
download_size: 5145514547
dataset_size: 5346976563.35
configs:
- config_name: default
data_files:
- split: train
path: data/train-*