metadata
dataset_info:
features:
- name: audio
sequence: float32
- name: text
dtype: string
splits:
- name: train
num_bytes: 6275123142
num_examples: 9480
- name: test
num_bytes: 1558405374
num_examples: 2370
download_size: 3705957610
dataset_size: 7833528516
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*