Zeqiang Lai

This is a stylegan2 issue, mostly it might be caused by incorrect installation of CUDA. For more reference, I recommend to take a look at https://github.com/rosinality/stylegan2-pytorch/issues

> **Q: RuntimeError: PytorchStreamReader fail reading zip archive** A: The checkpoint is broken. Download again by removing the broken one. Optionally, you could download the checkpoint manually from [here](https://huggingface.co/aaronb/StyleGAN2/tree/main), and...

> StyleGAN Custom OP, DLL not found - #12 - #40 - #38 - #31 - #35 - #30

> Q: No Cuda Runtime - #32 - #51 - #37

> Q: Requirements - #11 - #10 - #13

> **Q: Does this project support uploading custom image ?** A: Yes, this project support custom image.

## Performance - #52 - #44

> **Q: The custom image is of low quality, distorted, fail to drag.** A. The results might be distorted, this is an issue of the algorithms instead of the implementation....

Hardware Requirements

> Q: AttributeError: module torchvision.transforms has no attribute "Resize" A: Then version of torchvision is too low. Upgrade it via ```bash pip install torchvision --upgrade ```