metadata
dataset_info:
features:
- name: audio
sequence: float32
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 3882577961
num_examples: 18450
- name: test
num_bytes: 40879128
num_examples: 377
download_size: 3938134546
dataset_size: 3923457089