Style trained on images of the AI artist "Twitwit"
Use HiresFix and Adetailer for better results
Description
FAQ
Comments (7)
You should do their pre-March 2025 style. It looks like a mix of Nyantcha/Thiccwithaq and BCC-Chan. (https://rule34.xxx/index.php?page=post&s=view&id=12737808&tags=twitwit)
Also, consider doing their style pre-June 2025 too: (https://rule34.xxx/index.php?page=post&s=view&id=13923956&tags=dug-lith-pls)
I think I'll do all that when I have enough free time.
@zaraza365d Just saw the new model. Good effort, but it has a really glossy look to it. The style is more sketchy and runny(?) and flat with its colors and shadows. What tutorial did you follow to create a LoRA?
@civitai1111111 Thanks for the feedback! I haven't really followed any tutorials.
@zaraza365d I just made a small pastebin link full of the art that should be made in the newly refined version: https://pastebin.com/cE7L9EaD
@civitai1111111 I updated the model, you can check the result
@zaraza365d (Please consider using ADetailer and Hires, Fix in your workflows) -- It still doesn't seem to have that sketchy, runny (with bold line art) look to it as seen here (might be an Illustrious thing): https://rule34.xxx/index.php?page=post&s=view&id=11933074&tags=twitwit





