How should I adjust the hyperparameters to train a strong DDPM model based on CelebA?

[LinWeiJeff]

I want to use this github repository to train a DDPM model based on CelebA dataset consisting of 202,599 align&cropped human face images (each image is 218(height)x178(width) pixels) for human face image inpainting task. Is there anyone who can suggest me how to adjust the hyperparameters used in this repository to get a strong denoising model in DDPM? I hope the strong denoising model (UNet model) in DDPM learns the features of human face well so it can be used for human face image synthesis task and human face image inpainting task (i.e. recovering the masked parts of a masked human face image).

I want to know how to adjust the values of the hyperparameters in diffusion_defaults() of script_util.py, model_and_diffusion_defaults() of
script_util.py and create_argparser() of image_train.py to get a strong denoising model in DDPM based on CelebA dataset.

Thanks a lot!!! p.s. I directly set up the values of the hyperparameters in the codes not through flags

[honne07]

hello.I saw that you had the problem of setting the hyparameters,and now I still have the same problem. And I create a .sh profile and when I run the code,it didn't make sense. So could you tell me something about it?I've been troubled by this question for long.