Jun-Yan Zhu
Jun-Yan Zhu
It's possible. Please refer to [Training and Test Tips](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/docs/tips.md#trainingtesting-with-rectangular-images).
See #1468 for a detailed discussion.
See this [FAQ](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/docs/qa.md#runtimeerror-errors-in-loading-state_dict-812-671461-296).
It's hard to know. Maybe this image is corrupt, or this image is too small (smaller than your crop size). Which `--preprocess` flag did you use?
Sometimes, your cropping might get unlucky. It not only depends on the image sizes, but also depends on where you crop the patches.
`--norm` refers to using instance normalization or batch normalization [instance | batch | none]. See this [line](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/options/base_options.py#L37). You probably don't want to touch it. `--preprocess` is for data preprocessing. See...
See [Training/Test Tips](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/docs/tips.md#fine-tuningresume-training).
This [FAQ](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/docs/qa.md#runtimeerror-errors-in-loading-state_dict-812-671461-296) might be related to your issue.
It's hard to know. Every case is different. Would it be possible to share with us some console output / debugging output?
+1. please refer to [training & test tips](https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/docs/tips.md#trainingtesting-with-high-res-images) for high-res model training.