search

nothinglo / Deep-Photo-Enhancer

TensorFlow implementation of the CVPR 2018 spotlight paper, Deep Photo Enhancer: Unpaired Learning for Image Enhancement from Photographs with GANs
MIT License
794 stars 110 forks source link

Trained model fail to be used for inference (LPGAN_exp_G3_577) #72

Closed manifischer closed 5 years ago

manifischer commented 5 years ago

I trained a model (using expert E) and when trying to use it in the inference code (LPGAN_exp_G3_577) the model did not load with error: NotFoundError: Restoring from checkpoint failed. This is most likely due to a Variable name or other graph key that is missing from the checkpoint. Please ensure that you have not altered the graph expected based on the checkpoint. I've noticed that in the expert C has 2 trained models, the one which is used has "new " in the name, when I do training only 1 model is generated (without "new"). What can I do to fix this?

nothinglo commented 5 years ago

I add the rename script to the download link. Sorry for your inconvenience.

manifischer commented 5 years ago

Thanks a lot. I will try it immediately!

JeffJassky commented 4 years ago

How do we use the rename script? I ran it and am still having the issue same issue. The passed save_path is not a valid checkpoint

MrRobot2211 commented 4 years ago

Any news regarding this I am having the same issue when loading checkpoints for inference

Youlenda commented 3 years ago

I have this error; how you handled it? @JeffJassky

The passed save_path is not a valid checkpoint: /Users/yalda/Desktop/avir/image_enhancement/[Online_Demo_Models]_Deep-Photo-Enhancer/LPGAN_exp_G3_736/model/90.000-new.ckpt