Training Information

Trained in Anima-TrainFlow

• This LoRA is trained using Anima Base with the total image curated dataset of 170+ of Delga from Monster Ancient Cline/Mugen.

Prompt/Parameter Settings:

Lora Strength: .7 - .9

Base

Activation Tag:

1boy, delga, furry 

Reinforcement Tags:

dragon, green eyes, blond hair, arm tattoo, tail, 

Negative Tags:

black penis, purple background

Outfit

Delga Outfit:

clothes,  vest, clothes around waist,white tank top, fingerless gloves,  gauntlets,  pants, 

After thoughts

• This LoRA version was trained to a dataset that I loosely curated that leans in more a beefier and buff version of Delga, hence the name "XXL Delga".

• When generating background, the dataset lacks variety within the background, so there will be moments the background will be purple.

• Despite having a default style, it has the tendencies in generating black or latex like material on his genitals. Also, some occasion, generating a latex like clothes despite not being added within the prompt. So I highly recommended to add them within your negative prompts (or use NegPip which allows negative prompts within your main prompt) and specified a specific color for his genital like beige color.