Zenobia

@innerlee missing keys: `size mismatch for conv1.conv.weight_orig: copying a param with shape torch.Size([32, 6, 3, 3]) from checkpoint, the shape in current model is torch.Size([32, 4, 3, 3]). size mismatch...

@zcswdt Hello, may I ask if the above video can be converted normally when you use it? There are some problems with my conversion and I feel there is some...

@AliaksandrSiarohin Thank you for your reply, but I still don't know how to choose the appropriate video frame. Is the head kept as still as possible, the mouth shape is...

@AliaksandrSiarohin Sorry to bother you, but I also want to ask if my processing steps are correct and do not need other face alignment

@hangzhaomit Hello, when i use custom dataset with 5 class,getting this error: RuntimeError: cuda runtime error (710) : device-side assert triggered at /pytorch/aten/src/THC/generic/THCTensorMath.cu:226 You mentioned earlier that you can change...

1. I used the reconstruction results of train mode to calculate the L1 loss and the reconstruction results of avd mode are almost the same, so I think avd mode...

> 1. Reconstruction does not make sense for avd, since it specifically designed for cross identity, where the shapes of the objects could be different. > 2. There are no...

Thank you very much for providing the training log. Is it convenient to provide the training log of train-avd @yoyo-nb

May I ask if you have found the evaluation method and verification data set

@thiner Do you see a way to evaluate the model? I've been poking around the code for a long time without looking at the code and introduction to model evaluation...