models/SameFace Fix [Flux Lora] - v0.1

SameFace Fix [Flux Lora] - v0.1

7/22/2025

11:24:01 AM

Related Keywords & Tags

character,diversify look of women,flux.1 d,flux model,generative ai faces,lora,negative weight,sameface fix [flux lora],sameface lora,v0.1,vadimfedenko,woman

Portrait of a woman with black hair and blue eyes, smiling confidently under cool blue lighting with a sleeveless black shirt and minimalistic background.

A large serene dandelion dragon lounging on a rocky hilltop surrounded by children playing with floating dandelion seeds under a golden setting sun.

Highly detailed CGI illustration of a woman with long braided hair wearing a black dress, standing on a dirt path surrounded by green foliage and dry grass, with a hand lightly grasping her skirt.

A fantasy woman dressed in a red hooded cloak sits beside a large dark-furred wolf in a misty, dark forest illuminated by glowing mushrooms.

Realistic CGI render of a young woman wearing a blue ribbed turtleneck sweater and a flowing golden skirt standing in an autumn forest with tall trees and a wooden fence, wearing a black cat ear headband.

A Black woman with dark skin and short curly hair stands indoors under a glowing amber fabric art installation. She wears an avant-garde ochre outfit with arm wraps and a layered skirt, illuminated by warm golden light.

A hyper-realistic portrait of a woman with silver-white hair, wearing an intricate metallic silver headdress adorned with glowing blue gemstones, dressed in a navy cloak with celestial embroidery.

Cinematic portrait photograph of a woman wearing sunglasses and a beige blazer, sitting with a softly focused bokeh background and warm golden hour lighting.

A goth girl with long black hair and black lipstick stands inside an ornate ancient mirror frame in a ruined dungeon, wearing a translucent white night gown, with dark and misty dungeon background.

Close-up portrait of an aurora borealis cyborg featuring shimmering panels and glowing cosmic pathways, with an ethereal and serene cosmic background.

Close-up portrait of a stone cyborg with dark cracked rock and glowing mechanical implants, set against a misty mountain background.

Tips

Use the Lora with a NEGATIVE weight to act as a negative prompt without explicitly using one.

This Lora diversifies appearances by instructing the model what NOT to generate, avoiding typical same-face issues.

The model is highly experimental and feedback is encouraged to improve future versions.

Version Highlights

First Version!

Important: This Lora should be used with a NEGATIVE weight.

Also: The model is aimed at diversifying the look of women only. I plan to do a separate model for men later (and men with beards especially)

Also cool fact: the size of the Lora is only 4mb (it is limited to 2 blocks with a high rank)

So, What's a "Sameface"?

It's a common problem in generative AI, in which the model generates people with identical appearances. Unlike with SD1.5/SDXL, this is not just a result of the diverse dataset converging into a single token like "woman". in Flux, Dev/Schnell were distilled from the Pro version, resulting in a loss of diversity.

Why use Negative weight?

Lora was trained on the most generic photos of women that Flux generates.

If you then apply this Lora with a NEGATIVE weight, it will work as a sort of negative prompt, but without negative prompt!

By using this clever trick, Lora will basically tell Flux to generate anything except a typical Flux face. And as result, we'll get beautiful unique faces every time!

Why Lora and not tricks like lower Guidance?

People want diverse faces with high guidance in order to maintain high prompt adherence and image quality, that's why interventions at the latent space level are optimal.

How does it differ from other character Loras?

While other Loras introduce their own "samefaces", this Lora, only tells Flux what should NOT be in the generation, not what to add. As a result, this Lora diversifies people's appearance, rather than narrowing it even further.

How was the Lora actually made?

At first I was disappointed because nothing worked out. I was about to give up this project, but then tried to train Lora with just 2 layers, and suddenly during testing, I was met with beautiful faces, and with authentic, distant from Flux, appearance.

What's next?

I have a lot to experiment with. Different learning rates, layers and ranks have to be tested in order to improve quality. Merging different Lora variants might fix the problems with artifacts at higher values. I'll make sure to make a better version, If this Lora has enough interest

I'm really forward to your feedback, as this model is highly experimental. Any feedback is appreciated and will hopefully help me to improve the model.

special thanks to Ostris for AI-Toolkit