This is a model based on ILLUSTRIOUS XL 2.0 that incorporates some lora art styles
This is my first attempt. It might not be very good and seems not very stable. However, some of the produced pictures still occasionally amaze me. So I want to share them with everyone. I hope you will like them.
I translated this passage using translation software, so the words might not be accurate.(ﾟ∀ﾟ)
Steps: 30-40
CFG: 4.5-
Scheduler Sampler: Euler A and DPM++ 2M
Description
FAQ
Comments (4)
Can it be used to train LoRA Illustrious models?
It's rare you'd want to use a finetune to train a LoRA
I think this model has a lot of potential. Illustrious 2.0 is pretty great but it lacks a decent base style. When this model works, it's quite nice. Hopefully you can find a way to stabilize it!
Actually after using it awhile:
1. One of my favorite models! great job.
2. Need to have a way to remove the facepaint it wants to apply. It has a great propensity towards putting stars and stuff on the face.
3. The overall style is really good and the prompt adherence is definitely acceptable.
4. Being based on Illu2 1536 x 1536 base res works fine!
Sick model! I like it so much! Thanks!