STYLIZARD - 3D Stylized character prototyping
A merge of two LoRa. One trained with CivitAi training using captions for style and prompt understanding, the other trained on my 8gb gpu without captions for style manipulation.
After using them both in prompts I try to find a good balance, then I merge the two LoRa in supermerger (wich is the longest part. It's hard to find the correct weights and options to reproduce the same effect as the two LoRa used togheter).
This required a lot of trial and error and took most of the time.
Use it a 0.9 ~ 1.2 weight (depending on how much stylization you need and how much you wanna risk artifacts)
Struggles with backgrounds, I use it mainly for character concept to reproduce in digital sculpting softwares.
Training LoRa necessitate a lot of coffee. If you think that what I do is useful and want to help me keep doing this you can buy me a coffe here: https://ko-fi.com/clumsy_trainer