Jiale Xu

Noise on generated high-resolution images

2

Hi, I have utilized your code to generate images successfully. But I notice that there are some noisy points on the generated images, as the following images shows. Is this...

Why do you use `BatchRenorm` instead of `nn.BatchNorm3d`?

1

I found that you implemented a `BatchRenorm` module in your code. I wonder that why you did't use `nn.BatchNorm3d` of pytorch directly? I hope you can explain this detail, thanks...

Request for training code

2

Hi, I want to train BiFuse model on a new dataset, but this repo contains no training code, such as loss function, metrics and training script, etc. So could you...

Pretrainig skinning net does not converge on customized dataset

2

Hi, I'm training SCANimate on ClothSeq dataset provided by Neural-GIF[(https://github.com/garvita-tiwari/neuralgif)](https://github.com/garvita-tiwari/neuralgif). However, I met some problems and got bad results. Here are the logs of the first stage when training on...

In `curriculums.py`, the `dataset_path`s of CARLA and CATS are reversed. [https://github.com/marcoamonteiro/pi-GAN/blob/master/curriculums.py#L126](https://github.com/marcoamonteiro/pi-GAN/blob/master/curriculums.py#L126) [https://github.com/marcoamonteiro/pi-GAN/blob/master/curriculums.py#L164](https://github.com/marcoamonteiro/pi-GAN/blob/master/curriculums.py#L164)

Hi, thanks for your fantastic work! This project is really useful for playing with generative models. Do you have any plans to add support for StyleGAN3?

The answer to exercise 32.1.3 says that "then we can get the T(time) = (n-m+1)*[a(1)*b(1)+a(2)*b(2)+,,,+a(m)*b(m)] So we get the answer.", but after I sum up the series, I find that...

TextMesh Implementation

2

Thanks a lot for presenting this amazing project! Do you have any plan to implement the recent [TextMesh](https://arxiv.org/abs/2304.12439) paper and integrate it into threestudio? Looking forward to future updates!

Computing jacobians is extremely slow

1

Hi, I find the process of computing jacobians is extremely slow when testing with a 512x512 pretrained stylegan2 model, is this normal?

Does this repo support fp16 training?

1

Thanks for you excellent work! I'm wondering if this repo supports mixed-precision training since I had a hard time training stable-diffusion or zero123 using fp16. Setting `Trainer(precision=16)` did not work....