there are some strange white areas in my result.

[liangzimei]

Thanks for your exciting work firstly.
when i use -cn lama-fourier to train my own dataset . i find there are some white areas in some train and test images (not all images, and according to my observation it is irrelevant to mask size), like below( these two images are selected from epoch33/40):

and

Do you know how to avoid this situation? Thanks in advance.

PS:
my dataset is a food image set and there are 150,000 images.
And i use this command to train my model
CUDA_VISIBLE_DEVICES=0,1,2,3 python bin/train.py -cn lama-fourier location=food data.batch_size=10 data.num_workers=8 trainer.kwargs.gpus=[0,1,2,3] trainer.kwargs.limit_train_batches=12360 optimizers.generator.lr=0.001 optimizers.discriminator.lr=0.0001

[windj007]

Hi! Sorry for the late reply!

Did you resolved the issue?

What are the training and testing resolutions?

[liangzimei]

Hi! Sorry for the late reply!

Did you resolved the issue?

What are the training and testing resolutions?

sorry, the issue has not been solved. training resolution is 256 and testing resolution is 512.

[windj007]

Do these artifacts appear on training images as well?

[liangzimei]

Yes, here is some training samples in last epoch.

[liangzimei]

Do these artifacts appear on training images as well?

Hi, do you have any suggestions? Thanks.

[windj007]

Honestly, I did not have time to think about your issue thoroughly.

A couple of questions:

1. Could you please show training metrics from tensorboard? Especially discriminator scores and R1 gradient penalty, but others should be informative as well.
2. Did you try changing coefficient for R1 gradient penalty?

[windj007]

I'm closing the issue for now. If you have any other questions, feel free to reopen it or create a new one.