Attention: You need to get your own VAE to use this model. It does NOT work without a VAE. I recommend you try this one out.
This model is intended to be a cartoonish, colorful, high quality 2D model. Please give it a spin. I intend to keep this model as work in progress with regular updates.
See example images for examples of how to prompt the model and a sample negative prompt. I recommend always having infant in negative prompt for obvious reasons. Check "about this version" on the sidebar for the full list of models used.
Description
Models used:
Elysium V3
Counterfeit V3.0
Mistoon Sapphire
Loras:
InCase style
MAGIC Lora
Pear
Lit
As usual, thanks goes to everyone involved in making these models.
FAQ
Comments (5)
incase style?
Yes, InCase style. Searching on civitai will get you that lora. The only lora in this model that you can't find in civitai is the magic lora, unfortunately.
I really like the style / goal of this model.. but currently it's not quite usable for me.
Some feedback
- Feels overtrained?, getting very similar facial expressions / looks that require a lot of prompting to move off of for certain tags / concepts
- Doesn't quite generate the aesthetic I'm looking for.. looks almost 'clay-like'
- Could be pruned down to 2 gigs
Please continue working and improving this. I overall love the aesthetic but I think it needs to be cleaned up a little bit.
Thanks for reading.
Hey thanks for the feedback. Will keep it all in mind going forward.
对词条的理解很精准并且在多数情况下能够生成正确的肢体，非常喜欢这个模型！
Have a very accurate understanding of the prompt and can generate the correct limbs in most cases. I really like this model!