metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 3216337594.21
num_examples: 27922
- name: eval
num_bytes: 315293337.242
num_examples: 3082
- name: test
num_bytes: 548066740.984
num_examples: 5016
download_size: 3948072414
dataset_size: 4079697672.436
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*
- split: test
path: data/test-*