New versions: P_v3 (Pony) and IL_v2 (Illustrious)

Used a new dataset of mostly new 20 images:

Used images of very high resolution. Picked images that were more clear when viewing them at 100% zoom so that they would still be clear once they got resized for training by Civitai. Also eliminated images where the character's face is not clearly visible.
Cropped the images so they would only contain the character.
Simplified tags a little.
Dataset mostly contained images with the character's face in focus

Activation tag: dx_Adam_Jensen

Eyes: sunglasses / cybernetic eyes (even though I used the tag "cybernetic eyes" in training, it might be better not to use it for generation - the results are too wild)

Type of image: face portrait / upper body

Outfit: black coat / black coat with golden pattern / red prisoner's uniform / black combat armor / shirtless

Arm: cybernetic arms

Weapon: holding rifle

Background: indoor background

Will have to do some tests, but the character's face seems to get busted if you try to make a full-body image. I'll see what I can do.

Old versions: P_V1 (pony); P_V2 (pony); IL (illustrious)

Activation tag: adam_jensen

Eyes: sunglasses / retracted shades next to eyes

Outfit: black coat / black coat with golden pattern / red prisoner's uniform / black combat armor / shirtless

Arm: cybernetic arms

Weapon: cybernetic blade on arm / holding gun

You might need to put "logo, watermark" in the negative tags because the screenshots I used had the youtuber's watermark on them.

P_V2 is mostly the same as P_V1. Only changes are as such:

1 image with the cybernetic blade replaced (so use V1 if you want to try to create an image with it)
"1boy" tag changed to "1man"

Search AI Models & Addons

very few results