Train model on small dataset

autonomousvision / stylegan-xl

[SIGGRAPH'22] StyleGAN-XL: Scaling StyleGAN to Large Diverse Datasets

MIT License

961 stars 113 forks source link

Train model on small dataset #81

Open ADAMS12345678 opened 2 years ago

ADAMS12345678 commented 2 years ago

Hello author! I train an unconditional class image generation model on a small dataset (under 100 images), however the model does not generate multiple types of images, no matter how much the seed is set, it is only a small variation of a single type of image. May I ask what is the reason for this? I would be very grateful for your answer!

nom57 commented 1 year ago

check your stats.jsonl

either diffaug doesnt work or you have a "NaN" value for your scores/fake , or you could have misconfigured batch size and learning rate.