A model that pursues character creation that is close to a real photo without a sense of discomfort
It doesn't matter if you don't have any default prompt settings.
You can also create realistic images by adding realistic-related prompts.
For information about the default prompt, please refer to the default image.
Description
A model that emphasizes a realistic feel over the existing version
FAQ
Comments (5)
I love these 2 is better then 1, but what is the diff from earlier in the week? D= more d?
Every generation comes out with a sepia-like filter on it, usually with weird sampling spots. Could it be that I use k diffusion? really like the outputs but I cant deal with the crazy filter
Same here. SD reForge and SwarmUI,
Sounds like you are not using a VAE. Add a VAE like sdxl_vae.safetensors and you should see good output.
I love this model! However my images are coming out with a grey wash in comfy UI. I dropped it into an image editor and adjusted the contrast. This made it 100% better. Is there a recommended VAE?
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.