metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: sentence
dtype: string
splits:
- name: train
num_bytes: 2548963106.592838
num_examples: 21946
- name: test
num_bytes: 58073522.79916243
num_examples: 500
download_size: 2598203829
dataset_size: 2607036629.392
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*