Anna Frühstück
Anna Frühstück
You need to import that function from torch.hub: `from torch.hub import download_url_to_file`
Which GPU are you attempting to run the code with? We recommend a GPU with at least 12GB of memory to accommodate the models.
A 3080 shouldn't have issues memory-wise. This is speculative, but if the process previously crashed, some memory may still be retained from zombie python processes. Have you checked `nvidia-smi`?
I can replicate this error and will look into it when I have time. For a quick fix, please create an environment and attempt to manually install the packages specified...
You should be able to use your own trained model files by following these steps https://github.com/afruehstueck/tileGAN#using-your-own-datanetwork-optional