1 install miniconda or anaconda
2 Move "models" to the project root directory
3 install pytorch==》https://pytorch.org/get-started/locally/
4 pip install -e.
5 python startweb.py or python -X utf8 tools/diffusion/inference.py --config configs/svc_cn_hubert_soft_ms.py --checkpoint models/ys35/ys35.ckpt --speaker_mapping models/ys35/spk_map.json --gradioS