qwerdbeta

Thank you! that worked but then I ran into the next issue when trying to train and it's not clear what cauase is. The error stack trace is: (inpainting_gmcnn) G:\pythonAI\inpainting_gmcnn\tensorflow>python...

is it because I don't have enough images? This is just a test run with 181 images

Above was with tensorflow implementation. I also tried the pytorch implementation in windows but it causes this error on training: (inpainting_gmcnn) G:\pythonAI\inpainting_gmcnn\pytorch>python train.py --dataset celeba --data_file G:\pythonAI\training_images\source_images\train_images.index ------------ Options -------------...

I finally got the pytorch implementation working after several tweaks.

I have a similar problem but instead of making the images and masks smaller, I made them 512x512 (images) and 256x256 (masks) and I get a similar error with different...

above @nywang2019 says he is using 512x512 training images but in the trace it says img sizes are 256x256. @shepnerd what would be needed to get the pretrain step to...

I gave up on pconv because I really just needed a on square rectangle which I achieved. However, I'm wondering what weights in the Json files might be better tuned...

...or do they scale dynamically?