I'm going back and re using my dataset to create lycoris versions of them. So no other adjustment other then creating a dreambooth version and exporting it out as lycoris. The base model is the RealisticVision so it wont be as flexible in some other models.
Let me know what you think, and any suggestions for the future.
Description
FAQ
Comments (2)
Which do you find more accurate? Which do you find easier to train?
Depends. This one is based off the RealisticVision model so its way more focused on being "real". But that means it might be tougher to use in other models. The TI was trained under the based SD 1.5 so its more flexible to be used in different models. Its going to depend on what you want to look for.
The dataset is the same between the two.