metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 522183991.84
num_examples: 19600
- name: test
num_bytes: 10656816.16
num_examples: 400
download_size: 521928306
dataset_size: 532840808
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*