LoRA for Generating My Little Pony–Style Feral Characters
While the Anima model is capable of generating such characters on its own, the results are often inconsistent and require finesse in natural language prompting to achieve the desired style and anatomical accuracy.
This LoRA simplifies the process, enabling you to reliably produce images with your preferred artistic style and body type.
Recommended Settings:
Version 1.1
Apply a LoRA weight between 0.7 and 1, depending on how much influence you want from other styles or LoRAs.
Turbo recommended for more flat colored styles.
Version 1.0
Apply a LoRA weight between 0.4 and 0.6, depending on how much influence you want from other styles or LoRAs. A trigger word is not required, but you may use NW-MLP to enhance consistency.
Explicitness Control:
Leverage Anima’s native settings — safe, nsfw, or explicit — to adjust the level of explicitness as needed.
Description
Trained with bias towards lower steps to keep the anatomy correct but keep the desired style
Lower learning rate of network to keep the model from overfitting, now making possible to use it for full potential on 0.8-1 strengths.