metadata
dataset_info:
features:
- name: audio
sequence: float32
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 3663786480
num_examples: 10000
- name: test
num_bytes: 726295950
num_examples: 771
download_size: 3970829647
dataset_size: 4390082430
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*