metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: sentence
dtype: string
- name: path
dtype: string
splits:
- name: train
num_bytes: 1795947809
num_examples: 16672
download_size: 1660739004
dataset_size: 1795947809
configs:
- config_name: default
data_files:
- split: train
path: data/train-*