
Flat style Illustration is now commonly used for marketing assets, such as banners, posters,etc.
This LoRA is trained on Anything base model via 11100 steps. It enables user to generate flat style art, especially characters with or without noise texture/facial details/exaggerated body proportions.
2D base models are recommended:
-
Anything/SD 1.5 : suggested when no facial details
-
Counterfeit/GhostMix : suggested when cute face required, as facial details will be mandatorily added if used
Recommended configuration:
-
Use trigger word flat illustration to trigger the style.
-
Full weight recommended
-
Use noise texture prompt to switch on noise texture effect
-
Use exaggerated body proportions prompt to switch on popular stretched body shapes
-
Use no facial details prompt and (eyes:2),(nose:2),(mouth:2) negative prompts if you want empty faces for your characters (not 100% success rate though, and only works with SD1.5/Anything base model)
Please kindly leave your comments and rating on this model. Thank you!
描述:
v1.0
Initial release to enable marketing flat style
训练词语: flat illustration,noise texture,exaggerated body proportions
名称: MarketingFlat-v1.safetensors
大小 (KB): 147577
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success