Open qxforever opened 1 year ago
Thanks for your great work!
I found that $\lambda_{lr}$ is different in the paper and code. In the paper it is set as $\lambda_{lr} = 0.1$. But in lib/trainer/AvatarTrainer.py it is set to $1.0$ .
\lambda_{lr}
\lambda_{lr} = 0.1
lib/trainer/AvatarTrainer.py
https://github.com/YuelangX/LatentAvatar/blob/257aea312dc1322ac9fd5994a61ce6def702fcca/lib/trainer/AvatarTrainer.py#L57
Should I change the code if I want to reproduce your results on demo dataset?
Thanks for your great work!
I found that $
\lambda_{lr}$ is different in the paper and code. In the paper it is set as $\lambda_{lr} = 0.1$. But inlib/trainer/AvatarTrainer.pyit is set to $1.0$ .https://github.com/YuelangX/LatentAvatar/blob/257aea312dc1322ac9fd5994a61ce6def702fcca/lib/trainer/AvatarTrainer.py#L57
Should I change the code if I want to reproduce your results on demo dataset?