If you feel like showing some love, awesome.

But the best kind of support?

Use the model. 🛠️

Post your results. 🖼️

Let the art speak. 🎯

Big thank you to @Uberocarina for the concept idea

FullRobeLady - Robe & Gown Codex

FullRobeLady is a long-form project dedicated to exploring the beauty, elegance, and power of women wearing robes across genres, cultures, and eras.
Where FullMetalLady mastered the craft of armor, FullRobeLady turns to flowing fabrics, ceremonial silhouettes, and layered garments that define iconic aesthetics throughout history and fantasy.

This model is not limited to one style. It is a codex of robe traditions, each trained as its own self-contained version, unified under the FullRobeLady banner. Every edition captures the feminine form with fit, elegance, and presence, spanning ceremonial to practical, romantic to martial.

The Vision

FullRobeLady is designed as a flexible library of robe archetypes. Each version is a deep dive into one robe culture or theme, but the shared philosophy remains:

• Naturally beautiful woman with a real effort to make em look realistic

• High-detail fabrics, lifelike drapery, and photoreal textile rendering

• Cinematic compositions with grounded realism

FORCE Mk.1

FORCE Mk.1 is the Star Wars–inspired branch of FullRobeLady, trained on practical Jedi-style robes without copying any specific characters. It focuses on layered tunics, tabards, sashes, trousers, boots, and the flow of fabric that makes Jedi silhouettes iconic. The goal is not cosplay but the feeling of a grounded, cinematic Jedi aesthetic brought to life in photoreal detail.

This LoRA was trained primarily on outfits, fabrics, and flowing cloth. A smaller portion of the dataset included lightsabers, so saber renders may require prompt tweaks or inpainting for consistency.

Usage Notes

• Tested with Euler + Beta

• Flux guidance between 2.5 and 3.5

• Around 40 steps is a good sweet spot (also gives good result at 25+)

• A LoRA strength of 1.0 is balanced for single-use, but may need to be reduced when combined with other LoRAs

What It Delivers

• Naturally beautiful women in photorealistic, Jedi-inspired robes

• Flowing fabrics, layered textiles, and cinematic silhouettes

• Flexible outputs ranging from serene portraits to dynamic action poses

PRINCESS Mk.1

PRINCESS Mk.1 is the fairytale-inspired branch of FullRobeLady, trained on extravagant gowns and voluminous dresses that echo the timeless silhouettes of Disney-style royalty. The focus is on flowing skirts, layered bodices, rich fabrics, embroidery, and ornate detailing that make each dress feel majestic and larger-than-life. It aims to bring the essence of a storybook princess into photoreal detail without imitating any single character.

This LoRA was trained primarily on outfits and fabrics, from high-volume ballroom gowns to layered two-tone designs. The dataset emphasizes intricate textile rendering, embroidery, and ornate construction, ensuring outputs that balance elegance and realism.

Usage Notes

• Tested with Euler + Beta

• Flux guidance between 2.5 and 3.5

• Around 40 steps is a good sweet spot (also gives good result at 25+)

• A LoRA strength of 1.0 is balanced for single-use, but may need to be reduced when combined with other LoRAs

Because of the voluminous scale of the dresses, full-body renders may occasionally produce distorted “lazy eyes.” Using Adetailer (or equivalent face refinement) is recommended in those cases. At mid-body and close-up ranges, women should appear gorgeous, photoreal, and charming — embodying the feel of adult fairytale princesses.

What It Delivers

• Naturally beautiful women in photorealistic, princess-inspired dresses

• Voluminous silhouettes, layered skirts, and intricate embroidery

• Photoreal textile detail with cinematic lighting and composition

• Flexible outputs ranging from regal portraits to full-body fairytale showcases

ARCANA Mk.1

ARCANA Mk.1 is the high-fantasy branch of FullRobeLady, trained on intricate mage, sorceress, and wizard attire inspired by Dungeons & Dragons–style adventurers and spellcasters. It focuses on layered fabrics, embroidered hems, ornate collars, cloaks, hoods, and form-fitting garments paired with flowing silhouettes. The aim is to capture the cinematic realism of powerful women immersed in the arcane — witches, mages, and sorceresses rendered in photoreal detail.

This LoRA was trained primarily on outfits and fabrics, featuring a wide variety of designs ranging from grounded adventurer looks to elaborate high-fantasy ensembles. Many images also included props such as staves, wands, tomes, scrolls, glowing orbs, and other magical items, adding further depth to the dataset and encouraging diverse outputs.

Usage Notes

• Tested with Euler + Beta

• Flux guidance between 2.5 and 3.5

• Around 40 steps is a good sweet spot (also gives good result at 25+)

• A LoRA strength of 0.50 to 0.70 gave the most balanced results, but it can be cranked up to 1.0 when more detail is desired but hands gets weird... fast

Because the dataset includes many images of women holding varied magical objects — such as books, wands, staves, orbs, and glowing spheres of energy — full-body renders may occasionally produce hand distortions. Changing seed or using inpainting usually resolves these issues, and a hands-focused LoRA can be helpful if problems persist. At mid-body and close-up ranges, expect naturally beautiful women with enchanting presence, defined by their magical attire.

What It Delivers

• Naturally beautiful women in photorealistic, mage-inspired attire

• Ornate textiles, embroidery, hoods, and layered fabrics with fantasy depth

• A balance between grounded adventurer outfits and elaborate high-fantasy spellcaster looks

• Flexible outputs ranging from mystical portraits to cinematic full-body showcases