search

lipku / metahuman-stream

Real time interactive streaming digital human
https://livetalking-doc.readthedocs.io/
Apache License 2.0
3.56k stars 502 forks source link

替换成自己训练的模型报错 #45

Open feipengheart opened 6 months ago

feipengheart commented 6 months ago

image

SevenKous commented 6 months ago

The default audio encoder in ER-NeRF is deepspeech, you have to choose the model wav2vec esperanto

zhuxiu1234 commented 6 months ago

请问ngp_kf.pth 这么文件在哪呢？为啥我训练出来的模型里没这个文件

lipku commented 6 months ago

就是训练后的pth文件，不一定是这个名字

jt-z commented 3 months ago

好像是模型加载参数不匹配？似乎必须要用 wav2vec 的 audio encoder