Datasets:

jiaheillu
/

sovits_audio_preview

# imallryt语音模型训练参数



duration of dataset：22min 

epoch: 2960

batch_size :24

segments = 122

step:14200

Losses vs step



![截屏2023-04-08 11.46.24](./README.assets/%E6%88%AA%E5%B1%8F2023-04-08%2011.46.24.png)