A training and content contest with over 2 Million Buzz in prizes and 50% off H3 generation for the duration of the contest.

The Civitai H3 Contest: 2 Million Buzz Up For Grabs

Lora for Medusamon from digimon, specifically digimon liberator. Trained mostly on images from the webcomic. I recommend checking the notes for each version to see what they do and what settings are recommended

Anima version of the model trained on the same dataset as prior versions for 9 epochs. Seems to reliably produce the character with minimal extra hassle. The trigger is "Medusamon", it helps to add some of the other example keywords depending on how much of the character is visible. It seems to work with a lora strength of 0.8-1.0 and doesnt mess up other characters too much when used alongside them.

Anima V1

Locally trained and improved (hopefully) version of the model utilizing 32 total images from across the internet and digimon trading card game. This version demonstrates an improved ability to reproduce her snake hair. It seems to function across both a variety of Illustrious and N00bAI models at a strength of 1.0, though that may need to be adjusted based on the model. The base trigger word is just medusamon, but you will need additional descriptors depending on the style and model you are using, which I will include in the trigger words section.

Illustrious V2

Improved version of my Medusamon Lora, trained locally on Illustrious 0.1 for about 50 epochs on 25 images. This version seems to get the character much more consistently and works on a variety of styles and models. The base trigger word is "medusamon", but "1girl,solo,red skin,snake hair,blue hair,pointy ears" seems to help with the consistency of the look. I also recommend "platform footwear,detached sleeves,fingerless gloves,low wings,tail" for full body shots. You can try "twintails" to get her hair look properly, but its not very consistent. I recommend a lora weight of around 0.9-0.97.

Illustrious V1

Chroma version of the model trained on AI generated joycaptions of the already existing dataset I gathered. I might still need to tweak some of the parameters to get optimal generations, but just describing the character and including the "Medusamon" tag seems to get consistent results. Refer to the example images for some prompts to trigger the character.

Chromav1

Beta pony version trained mostly on card art and the webcomic using civitai's training service. This version seems to have some issues with the snake hair, but with the right prompting you can get good results. The trigger word is just "medusamon", but you may need to use some additional trigger words depending on what style and or other loras you are using. Works best with a strength around 0.95, but you can go anywhere from 0.8 to 1.0 with it.

PonyXL beta

Beta version of the model trained on 15 images from the webcomic and cards using civitai's training service. This version is a bit inconsistent, especially with her snake hair, as it doesnt generate them properly as twintails. Using a lot of trigger words helps with the look, but the base trigger word is "medusamon". I recommend a strength of around 0.9-1.0.

Illustrious Beta

Beta version trained on civitai's training service for FLUX for 8 epochs. This version was trained on sentence captions partially generated by joycaption. The basic trigger word is "medusamon", but you will need to describe her appearance to get accurate generations. I recommend looking at the example images to see good prompts to use. This version was mostly trained on close-ups of her face and upper body,so it has some issues generating her snake hair and tail. It also tends toward a more anime style, so i recommend lowering the strength to around 0.8 if you want to do different styles.

Flux Beta

Trained on a similar dataset as my prior two versions using civitai's training service. This model isnt as consistent as the others and seems to have issues generating realistic style images, but it works in multiple checkpoints and only requires a single keyword to consistently get the character. The trigger word is simply "medusamon", though adding other descriptors can help. I recommend a strength of around 0.95.

SDXL Beta Version

Medusamon (Digimon)

physical violence

weapon violence

wide hips

revealing clothes

downblouse

convenient censoring

pg-13

corpses

suggestive

oral invitation

pg13

sexy

huge breasts

thick thighs

sexual situations

male nudity

disturbing

male swimwear or underwear

female swimwear or underwear

partial nudity

undressed

female nudity

breasts out

exposed female nipple

breast out

male underwear

hair over breasts

female swimwear

gigantic breasts

no panties

graphic violence or gore

covered nipples

huge butt

strapless leotard

sitting on face

emaciated bodies

one breast out

female underwear

nude

nsfw

lingerie

graphic male nudity

adult toys

illustrated explicit nudity

nudity

graphic female nudity

hentai

futanari

porn

sexual intent

genitals

peeing

vore

oral

sexual activity

anal

blowjob

dildo riding

incest

hanging

hate symbols

nazi party

white supremacy

diapers

scat

self injury

hate speech

urine

extremist

child on child

latex clothing

swimwear

bukkake

fellatio

cumshot

implied fellatio

eat_cum

cumdrip

cum in pussy

cum on face

after fellatio

cum on hair

cum on body

cum on tongue

cum on hands

cum in mouth

triple fellatio

autofellatio

fucked silly

cum on pussy