It’s locally run stable diffusion 1.5 with a few models and loras, nothing too special. Using an old 1080ti it’s more than enough.
Most of the work here is being done by the model
dax-nxx
- 21 Posts
- 24 Comments
Joined 1 year ago
Cake day: June 16th, 2023
You are not logged in. If you use a Fediverse account that is able to follow users, you can follow this user.
12·10 months agoExtras -
Nice, like how that turned out!
Which model did you use?
Warrior version, original prompt with less focus on boobs and using img2img based on a game screenshot.
Extra Generations, Same prompt.
Extras / Bonus Content
Bonus Content
Bonus Content:
Theme Suggestion: Water (in it, near it, under it etc).
spoiler
parameters
1woman, (assassin). (armour), (armour with breast cutout). (long black hair), (irish), (deep green eyes). (skin cutout). (Medieval), (High Fantasy:1.2). , (ultra quality, highly detailed, realistic, high quality, 4k, 8k, Meticulous design). (Drawn), (exposed flesh). (sexy), (fantasy)., (sword on back).
Negative prompt: FastNegativeV2, painting, plastic, airbrushed, shine, shiny, photoshop, doll, bad anatomy, bad hands, extra fingers, deformed hands, watermark, text, ugly, child, immature,
Steps: 40, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 1388854282, Face restoration: CodeFormer, Size: 512x512, Model hash: 05f1465502, Model: realcartoon3d_v9, Version: 1.6.0 postprocessing Postprocess upscale by: 4, Postprocess upscaler: R-ESRGAN 4x+
Nice! Broken down truck - Check, Trashy Tattoos - Check, Cans of cheep beer - Check. Ready to record a country song.
The grey haired one with her arms up? Not sure what you mean exactly but nothing like that intended. I’ll remove it to be safe.
Very cool!
Bonus content, mummies through time / unexpected places.
Okay went with Egyptian mummies (sort of).
Theme suggestion: magic, clerics, sorceresses, powerful magic woman etc,
Entry:
Prompt
spoiler
parameters (cgi), 1woman, (ancient Egyptian mummy), (mummy wrapped in bandages), (dark tanned skin), (bathing by a river), (bandaged arms, bandaged leg, bandaged torso, bandaged stomach), (natural breasts, sexy pose, relaxed), (sunset ), (alluring, smile, sexy, slutty, inviting),
[ultra quality, highly detailed, realistic, high quality, 4k, 8k, Meticulous design], , [perfect face, perfect hands, perfect hand, detailed hand]. Negative prompt: EasyNegative, painting, plastic, airbrushed, shine, shiny, photoshop, doll, bad anatomy, bad hands, extra fingers, deformed hands, watermark, text, ugly, child, immature, cartoon. Steps: 40, Sampler: DPM++ 2M Karras, CFG scale: 7, Seed: 3122720251, Face restoration: CodeFormer, Size: 512x512, Model hash: d05db48865, Model: Cust_Rev122_SFEB11_v01, Lora hashes: “mummy_costume_v0.2: 1fe09ffc6744”, Version: 1.6.0
postprocessing Postprocess upscale by: 4, Postprocess upscaler: R-ESRGAN 4x+
I like the idea of a weekly challenge, sound like a fun way to generate content and engagement as well.
Suggestion for next time: Medieval Heroes/Villains/Adventurers.
Normal Clowns
Bizarre, Creepy, Unsettling Clowns
** My Personal Favorite **
Thanks. First time posting a link on Lemmy. The () was empty so I think it was linking to itself. Should be fixed now.
Hey, I know you got downarrowed to all hell, but I tried to think about what “Woke Nukem” would really look like. Pretty, no muscle almost scrawny, someone cast for looks not action chops. Basically the opposite of what’s been shown in the games.
I find the best chance of having a recognisable face is to have the subject facing the viewer. With subject looking away or at a weird angle it has a hard time for sure. Have part of your prompt reference eyes or face “gold eyes” or “portrait” may help.
Turning on face restore helps.
Embeddings and Loras also help to some extent.
embedding: ng_deepnegative_v1_75t / EasyNegative / badhandv4,
lora : epiCRealismHelper
Models: I am finding the “realcartoon” set to be pretty reliable.
You can also try adetailer, which after the base image is generated will re-generate in more detail either the face or body (your choice). Note: I am not using this at the moment as it slows down generation considerably.
What model(s) are you using? Does the same happen if you copy this prompt for example? Happy to try one of your prompts out to see what the result is.
Edit: final thought, some models struggle with resolutions they weren’t trained in, so stick to 512x512 unless you’re sure then upscale your choice.