sexual situations

physical violence

disturbing

male nudity

hanging

hate symbols

nazi party

revealing clothes

weapon violence

female swimwear or underwear

male swimwear or underwear

partial nudity

white supremacy

adult toys

graphic male nudity

illustrated explicit nudity

nudity

graphic violence or gore

graphic female nudity

corpses

wide hips

convenient censoring

peeing

emaciated bodies

exposed female nipple

sexual activity

sexual intent

undressed

male underwear

female swimwear

genitals

female underwear

thick thighs

breasts out

strapless leotard

vore

breast out

one breast out

huge breasts

gigantic breasts

huge butt

covered nipples

hair over breasts

no panties

sitting on face

anal

dildo riding

downblouse

oral

porn

futanari

hentai

nude

lingerie

sexy

nsfw

child on child

incest

oral invitation

blowjob

suggestive

bestiality

self injury

extremist

hate speech

diapers

urine

menstruation

scat

female nudity

bukkake

cumshot

cumdrip

cum in pussy

cum on face

cum on hair

cum on body

cum on tongue

cum on hands

cum in mouth

cum on pussy

eat_cum

fellatio

implied fellatio

after fellatio

triple fellatio

autofellatio

pov fellatio

fucked silly

presenting ass

These are my rough attempts at a character LoRA for Emmrich Volkarin from Dragon Age: The Veilguard.<hr />V3 I tried to further refine the training tags, being more selective with training images, and trying to improve the background options by including scenery shots in the training.<ul><li>I tried training for hazel eyes (instead of brown), as I noticed I sometimes see green in-game, so I wanted it to be more accurate.</li><li>I tried to training for both gray AND black hair (as well as the multicolored hair tag) since his hair definitely still has patches of black/dark gray so i felt just having a gray hair tag wasn't accurate.</li><li>The likeness is consistent and the bracelets/glove are still random at times but not too terrible.</li><li>The mage staff...is a nightmare to work with on the more realistic checkpoints (like homosimile) but it seems "okay" on 2D checkpoints (like AutismMix and the Pony V6 checkpoints).</li><li></li></ul>I recommend keeping the LoRA strength around 0.7-0.8 and CFG between 4-7.An adetailer is also recommended for the more zoomed out shots.V3 main tags are as follows:<pre><code>e_volkarin, mature male, hazel eyes, mustache, grey hair, black hair, multicolored hair, very short hair</code></pre>optional - add "muscular" to negative prompt if he looks too bulky (especially in NSFW or shirtless shots).V3 outfit tags are as follows:<pre><code>green vest, yellow shirt, bracelets, multiple rings, single red glove, red waist sash, tan pants, red boots</code></pre>V3 optional tags:<pre><code>staff, garden, cemetery</code></pre>The settings I used for my V1 generation testing were 4-7 CFG, 20 steps, Euler A - Automatic using Forge.<hr /><hr />V2 (clip skip 1) I tried adjusting the training tags and also added more images to the training to make the LoRA more versatile when used with 2D and 2.5D checkpoints. I recommend keeping the LoRA strength around 0.8 and CFG between 3-5.An adetailer is also recommended for the more zoomed out shots.V2 main tags are as follows:<pre><code>e_volkarin, mature male, hazel eyes, mustache, grey hair, short hair</code></pre>optional - add "muscular" to negative prompt if he looks too bulky (especially in NSFW or shirtless shots).V2 outfit tags are as follows:<pre><code>green vest, yellow shirt, bracelets, multiple rings, single red glove, red waist sash, tan pants, red boots</code></pre>V2 optional tag<pre><code>staff (weapon)</code></pre>The settings I used for my V1 generation testing were 3-5 CFG, 20 steps, DPM++ 2M - Karras using Forge.<hr /><hr />V1 (clip skip 1) was my first attempt for this LoRA and generates decent results but not necessarily the greatest (the mustache sometimes comes out thicker than it should be). I recommend keeping the LoRA strength around 0.8 and CFG between 5-7.An adetailer is recommended and you may also need to add "fingerless gloves" to negative prompts and "closed mouth" to positive prompts.V1 main tags are as follows:<ul><li>e_volkarin - (subject tag)</li><li>mature male</li><li>hazel eyes</li><li>mustache</li><li>grey hair</li><li>short hair</li><li>optional tag - "closed mouth" - you may need this if he keeps generating with an open mouth as if he's speaking.</li></ul>V1 outfit tags are as follows:<ul><li>green vest</li><li>yellow shirt</li><li>collar chain (jewelry)</li><li>gloves - (you may need to add "fingerless gloves" to negative prompt if the red glove keeps rendering oddly)</li><li>red waist sash</li><li>tan pants</li><li>red boots</li><li>optional tag - "staff"</li></ul>The settings I used for my V1 generation testing were 5-7 CFG, 30 steps, DPM++ 2M - Karras using Forge.

V3 I tried to further refine the training tags, being more selective with training images, and trying to improve the background options by including scenery shots in the training. <ul><li>I tried training for hazel eyes (instead of brown), as I noticed I sometimes see green in-game, so I wanted it to be more accurate.</li><li>I tried to training for both gray AND black hair (as well as the multicolored hair tag) since his hair definitely still has patches of black/dark gray so i felt just having a gray hair tag wasn't accurate.</li><li>The likeness is consistent and the bracelets/glove are still random at times but not too terrible. </li><li>The mage staff...is a nightmare to work with on the more realistic checkpoints (like homosimile) but it seems "okay" on 2D checkpoints (like AutismMix and the Pony V6 checkpoints).</li></ul>I recommend keeping the LoRA strength around 0.7-0.8 and CFG between 4-7.An adetailer is also recommended for the more zoomed out shots.

V2 (clip skip 1) I tried adjusting the training tags and also added more images to the training to make the LoRA more versatile when used with 2D and 2.5D checkpoints. I recommend keeping the LoRA strength around 0.8 and CFG between 3-5.An adetailer is also recommended for the more zoomed out shots.

V1 (clip skip 1) was my first attempt for this LoRA and generates decent results but not necessarily the greatest (the mustache sometimes comes out thicker than it should be). I recommend keeping the LoRA strength around 0.8 and CFG between 5-7. An adetailer is recommended for this LoRA. You may also need to add "fingerless gloves" to negative prompts and "closed mouth" to positive prompts.

Dragon Age: The Veilguard - Emmrich Volkarin - Pony

00080-1086752184.png

Collection for my Dragon Age models trained for Pony.

My Dragon Age (Pony) models