prepare_data.sh

#download LjSpeech
wget https://data.keithito.com/data/speech/LJSpeech-1.1.tar.bz2 -o /dev/null
mkdir data
tar -xvf LJSpeech-1.1.tar.bz2 >> /dev/null
mv LJSpeech-1.1 data/LJSpeech-1.1
echo "done 1"
rm LJSpeech-1.1.tar.bz2

gdown https://drive.google.com/u/0/uc?id=1-EdH0t0loc6vPiuVtXdhsDtzygWNSNZx
mv train.txt data/
echo "done 2"

#download Waveglow
gdown https://drive.google.com/u/0/uc?id=1WsibBTsuRg_SF2Z6L6NFRTT-NjEy1oTx
mkdir -p waveglow/pretrained_model/
mv waveglow_256channels_ljs_v2.pt hw_tts/waveglow/pretrained_model/waveglow_256channels.pt
echo "waveglow done"

gdown https://drive.google.com/u/0/uc?id=1cJKJTmYd905a-9GFoo5gKjzhKjUVj83j
tar -xvf mel.tar.gz
echo $(ls mels | wc -l)
rm mel.tar.gz

#download alignments
wget https://github.com/xcmyz/FastSpeech/raw/master/alignments.zip
unzip alignments.zip >> /dev/null
echo "alignments done"
rm alignments.zip