metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 5260665634.825
num_examples: 14729
- name: test
num_bytes: 595249339.986
num_examples: 1637
download_size: 5274156815
dataset_size: 5855914974.811
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*