KIHONG KIM

Thanks for your reply These are the error which I got Is this a cuda library related error? Now, I'm using GeForce RTX 2080, CUDA 10.2 Version ```text 2021-01-08 16:07:22.553627:...

Thanks, I solve the problem by changing pytorch-lighting version but there was another error while running `SBERT-to-AE` ``` Traceback (most recent call last): File "translation.py", line 521, in trainer.fit(model, data)...

> Hi, try to set 1.5x or 2x of style image size, then it may smooth your pic I'm experiencing similar problems. May I ask you why it works out?

Hi, @lxj616 Thanks for your interesting Make A Stable Diffusion Video I wonder how to train your Pretrained toy model Could you explain a little bit more how to trained...

@lxj616 cool Is this method fine-tuning approach which applying a model trained with image to video? or Using video model and data from the training stage? Actually I don't read...

@lxj616 Thank you for your reply. I understand what you did. To make a video, adding temporal consistency layer and train them similar to dreambooth. Pseudo3DConv and pseudo 3d attention...

> tensor.backward(tensor_a)는 tensor에 대한 gradient를 tensor_a로 지정해주는 효과가 있습니다. wgan discriminator loss는 fake - real이므로 loss에 대한 real의 gradient는 -1이 됩니다. 그냥 (-real_predict).backward()를 써도 되지만요. Wgan loss였군요 ! 빠른 답변...

I also struggle in finding the good results of APNet. Should I use the renderer, landmark loss as well?

> > @lubovbyc I am still confused with the problem. I did not find this problem on my dataset. I attached the dataset with 4K images and the corresponding checkpoint....

Thank you for your answer @Alex-experiments. However, I was unable to render because the following error occurred. ``` Traceback (most recent call last): File "gen/cad_img.py", line 73, in main() File...