This Model is highly tuned for Photorealism with the tiniest amount of exessive prompting needed to shine.
All Showcase images are generated without Negatives (V1) to show what is possible on the bare prompt.
The model has highly knowledge of what a photo is, so if u promt u can avoid using photo. If the prompt tends to be fantasy like the model will turn away from photo and u have to tweak by prompting and/or negatives.
The model can do various ethnicities well, so try them out.
Age is also well trained and known by the model, so try them out too.
use simple prompts without "fake" enhancers like "masterpiece, photorealistic, 4k, 8k, super realistic, realism" etc.
don't use a ton of negative embeddings, focus on few tokens or single embeddings
you can still use atmospheric enhances like "cinematic, dark, moody light" etc.
start sampling at 20 Steps
no extra noise-offset needed
Style Negatives: colorful Photo | soft Photo
!After Detailer | ControlNet | Agent Scheduler | Ultimate SD Upscale
ā Feel free to leave Reviews and Samples - and always have fun creating ā¤
have to rush, but belive me - it's insane :D
Samples will follow next week if i have the time for that.
Private Model Release
This is my finetuned private Model which is based on the Z Model with a merge of finetuned Photo Lora to improve photorealistic outcomes, dont expect too much change to the Z Version
This is finally the last model update to come i swear
get that crappy look of reallife Photos somehow into 1.5, understanding and variety is very low
This should be most versatile model of this epiCPhotoGasm model and probably it will be the last.
Have fun trying it out!
Did restart and a bunch of tests over the days. Seems to be mostly stable at hands now and following the prompts even better with a bit of more details.
Did some merges with V1 and fintuned Versions after V4, think i ruined the Hands but the realism is more crisp i hope š
Have another Baking here and can't decide, i choosed that one first.
Tried to refine the overall output consistency (Arms/Hands/Legs) and also improving overall photo realness.
Let's see if this works out since i had spare time for in-depth testing
Can't decide, here's the second one
Tried to refine the overall output consistency (Arms/Hands/Legs) and also improving overall photo realness.
Let's see if this works out since i had spare time for in-depth testing
Okay okay, can we even get more realism? Whats Different? I added a trained Lora which was meant to be a Dodge and Burn Filter, turns out it boosts reality pretty much, so i merged it into my model š Let's see what you guys say.