XXXXRT666
XXXXRT666
And if you want to use the pre-trained models, I think have a few hours dataset would be better
It's said that learning rate is 1, and to train a pre-trained model, there is one more step.[111](https://github.com/RVC-Boss/GPT-SoVITS/issues/1002#issuecomment-2076884843)
> It's said that learning rate is 1, and to train a pre-trained model, there is one more step.[111](https://github.com/RVC-Boss/GPT-SoVITS/issues/1002#issuecomment-2076884843) I think it means that we first train sovits model and...
直接封装wav是最快的方式，不进行任何压缩
ffmpeg
你的api就不应该在runtime里
硬件快比我大了，太旧的无法支持
[？？？](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/issues/1967)，提两遍？
没N卡
> ``` > > "D:\workspace\GPT-SoVITS-beta0217\runtime\python.exe" GPT_SoVITS/prepare_datasets/1-get-text.py > > "D:\workspace\GPT-SoVITS-beta0217\runtime\python.exe" GPT_SoVITS/prepare_datasets/1-get-text.py > > Traceback (most recent call last): > > File "D:\workspace\GPT-SoVITS-beta0217\GPT_SoVITS\prepare_datasets\1-get-text.py", line 18, in > > from text.cleaner import clean_text...