Created by: luotao1

In https://github.com/PaddlePaddle/models/tree/develop/deep_speech_2#preparing-data

cd datasets
sh run_all.sh
cd ..

But there is no datasets directory currently, how do we prepare the data?