Welcome to epiCPhotoGasm
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.
Whats special?
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.
How to use
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
Additional Ressources
Style Negatives: colorful Photo | soft Photo
Useful Extensions
!After Detailer | ControlNet | Agent Scheduler | Ultimate SD Upscale
⭐ Feel free to leave Reviews and Samples - and always have fun creating ❤
Description
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.
FAQ
Comments (28)
Here we go again😂
Not a lot of letters left in the alphabet now...
And not much improvement either 😋
I created some images. The Y version provides so much improved results than the X series. Congratulations. I will prepare and upload a few images comparing X, Y.
This model SD 1.5 also works wonders. I guess there will be an SDXL version one day.
I'll be watching with great interest.
It is downright impressive how this was uploaded almost 6 hours ago and it's about to hit 1k downloads already. If that isn't a testament of how much we look forward to your work, I don't know what is.
Thank you for yet another version Epi <3
Cross fingers for the Final “Z” version… 🤞🏻 make it very perfect.
welp there goes my whole day, had a whole bunch of stuff planned but now im just gonna make AI pr0n all day
Hey, telling it like it is. I like that. People are so afraid that AI is going to put artists out of business, but I think it's going to put pr0n sites out of business, because now we can make our own, made especially how we like it. LOL!
@Leee1414 Generating AI art is an art form in itself, there's just a new breed of artists now.
How do I get normal sized breasts? Everything generated, regardless of how simple the prompt is, generates massive chests. "Small Boobs" and NP "Busty" does not have any affect.
Do not use the word breast in the positive prompt at all. Put the word breasts in negative prompt and adjust its weights.
@reinerwinkla I do not want a flat chest or a boy chest, but I'll try that and see what the results are. Thanks.
@Demosthine if you are looking for a Lora solution, I can recommend Breast Size Slider
specify age and play with its weight
naturally 18yo tend to have smaller breasts
@mahghhbb I hadn't thought about looking for that. Thank you.
Chest size typically works better with all models. The mention of breast and the models go straight for size and ignore, tiny, small or medium but chest like "medium chest" or "small chest" typically yield more favorable results.
This is great if you're not trying to give your girls blue eyes.
Bruh. When a new epicphotogasm model drops I feel like a kid on Christmas morning again.
it's so damn realistic
he can't keep getting away with this
<downloads immediately>
It's a shame that it generates (almost) only from Cup-E to Cup-K without external help, excluding the regular cup sizes, good model tho
Use busty in negative 🤓
Will we get an Inpaint version of the newer Models?
While I do think Y has more details, I find it less consistent and harder to control than X
Do you have comparisons or detailed use cases which does not fit correctly?
@epinikion Hey, first I want to say thank you for your work.
I'll try to describe what I'm doing.
I'm trying to create a consistent character. I'm using the same structure of a prompt which includes 2 loras and dynamic prompting. The dynamic prompts are of different backgrounds and different clothing (which sometimes themselves include clothing lora).
While using X, or even V4, I can make 100 images with about 5 being accurate to my prompt or what I consider good results.
Using Y, I made about 280 images and not a single one I found which listened to my prompt.
Sometimes it focuses too much on a specific lora and ignoring others, sometimes it focuses too much on a clothing color, coloring the other parts in the image.
Again though, the details in Y are visibly and easily better
Love your models and the frequency of the updates! So much appreciated. Are you planning a SDXL version?
I‘m planning to take my time for other stuff in my life for a few weeks. Probably I will return with a new SDXL model anytime but not soon.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.