metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
splits:
- name: train
num_bytes: 194471647079.83072
num_examples: 2015684
- name: validation
num_bytes: 28040296.365272578
num_examples: 279
download_size: 202898228069
dataset_size: 194499687376.19598
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*