Clothing - Aetherium Couture: Grand Gowns

If you are using ChatGPT, Gemini, Claude and so on for your prompt generation, you can feed it with this paste so the model knows directly how to use the LoRA - example prompts included.

Aetherium Couture: Grand Gowns is a Flux LoRA focused on monumental couture gowns with a strong fantasy-luxury identity. It is designed to generate dramatic, high-fashion looks built around sculptural silhouettes, sweeping trains, ornate embroidery, wing-like backpieces, crowns, filigree ornamentation, and elegant couture structure.

This LoRA is not aimed at simple generic evening dresses. Instead, it specializes in grand, editorial, highly stylized couture with a clear Aetherium design language: regal, theatrical, ornamental, and visually striking.

What it does well

• grand couture gowns

• sweeping trains and elongated silhouettes

• sculptural bodices and fitted couture structure

• ornate embroidery and filigree detailing

• wing-like backpieces, butterfly wings, bat-like wings, angelic shapes

• crowns, headpieces, ear ornaments, couture accessories

• black, ivory, gold, red, blue, and other elegant color variations

• studio, gallery, runway, palace, and editorial fashion settings

Style identity

This LoRA tends to produce body-centered, sculptural grand couture rather than classic heavy princess ballgowns. Many results combine fitted gown cores with dramatic trains, extensions, and ornamental backpieces, creating a strong fantasy couture look.

Trigger

aetherium_couture_grandgown

Recommended use

Works especially well for:

• full-body fashion portraits

• editorial couture concepts

• regal / dark fantasy / angelic / gothic looks

• highly ornamented ceremonial fashion

• accessory and detail shots such as crowns, collars, and ear ornaments

This is, just like my other clothing LoRAs, not a “single outfit LoRA”. Instead, I trained the design language, giving you the freedom to create similar looking outfits and garments.

Would love to see what you create with it.

The LoRA was trained on the default flux1-dev model for maximum compatibility.

Dataset and training data:

• 114 images

• 5 steps per image

• 10 epochs

• bucks with 1152px max size, no auto scaling

• cosine scheduler with 0.2 warmup and 0.7 decay

• Captioning was done with OpenAI

• alpha == dim: 32

If you use it with other loras, for example a character lora, try to lower the weight - I had good results at around 0.4 - 0.5.

As always: I did not polish any of the images in the showcase, they all came out like this. I did 2 images per prompt and picked the better looking one.