jerryuhoo/VISinger

Use VITS and Opencpop to develop singing voice synthesis; Different from VISinger.

Init

Unofficial Implement of VISinger

Reference Repos

https://github.com/jaywalnut310/vits

https://github.com/MoonInTheRiver/DiffSinger

https://wenet.org.cn/opencpop/

https://github.com/PlayVoice/VI-SVS

Data Preprocess

export PYTHONPATH=.

Generate ../VISinger_data/label_vits_phn/XXX._label.npy|XXX._label_dur.npy|XXX_score.npy|XXX_score_dur.npy|XXX_pitch.npy|XXX_slurs.npy

python prepare/data_vits_phn.py

Generate filelists/vits_file.txt Format: wave path|label path|label duration path|score path|score duration path|pitch path|slurs path;

python prepare/preprocess.py

VISinger training

python train.py -c configs/singing_base.json -m singing_base

or

./train.sh

Inference

./evaluate_score.sh

Samples