Nyoko74

I train and try to test my model with the same code but when I test my model, I cannot load the checkpoint: 

How to train the model

1

I follow the parameter in your paper， I can get a good model in pre-training while fine-tuning is bad，it seems that the model can just learn the color at average...

I try to get Imdbs following [the link](https://github.com/NVlabs/imaginaire/blob/master/projects/coco_funit/README.md), but get an error: `Error while finding module specification for 'imaginaire.tools.build_lmdb' (ModuleNotFoundError: No module named 'imaginaire')` I have already install the third...

Resolution of output images

1

Hello, I tried your pretrained model on 256x256 images, but the output image is still 256x256. How should I modify the parameters to achieve higher resolution output? Thank you.