RealStocking package add a new Lora in FLUX Dev world, dedicated to Tan (skin colour) backseam Fully Fashioned Stockings, worn with garter.
On the contrary if you need loras I trained for holdup stockings with lace top for FLUX, SDXL, or SD1.5 you can find them in my hosiery collection.
Concerning this Lora, RealTanFFStockings, is focused on vintage Fully Fashioned Skin colour Stockings, made for garter belt. The Lora is trained to give a detailed layout, including reinforcements on the heel and top and the iconic back seam.
Specific Lora is dedicated to Black FF Stocking, available for SDXL, PonyDiffusion and SD1.5. For Flux a specific Black FF Stocking lora is in progress as well.
If you like this Lora, please review it, use it and share pictures.
For a great output please read the following instructions.
Trigger word:
RealTanFFStocking
Activating Lora:
A strenght of 0.7 or 0.9 is highly recommended. Higher values can trigger artifacts.
So a good activation keyword for lora is:
<lora:RealTanFFStocking_FL:0.8>To ensure the visibility that garter add in the prompt the key:
garterA tipical sequence of keyword for this lora is:
RealTanFFStocking, garter, suspenders
The garter keyword trigger often to nude waist outputs. To render only a glimpse of clips and suspenders use only the following keywords:
RealTanFFStocking, suspendersTo render the backseam, typical of this stocking, trigger a back view of the subject the following specific keywords:
back view, RealTanFFStocking, garter, suspendersSampling steps recommended: 30-40
If prompt is very long and structured, a further inpainting step is suggested to obtain a more effective detialed result for stockings.
Use the sample image prompt as further examples for generating images.
Final statements:
1. You are solely responsible for any creative work using this model.
2. You cannot use this model to intentionally produce or share illegal images or content.
3. Furthermore can't use this model for malicious, hurtful, defamatory, fraudulent or political purposes.
Description
FAQ
Comments (2)
Thank you! Works well, looks amazing.
You might have some hands in your training data, making hands a bit harder to keep under control.
Offcourse in the training some hands are fatally included, but they are very well described in the training prompt. Maybe the very high details necessary to render this particular cloth needed to find a compromise with a slight overtraining of this Lora. One very easy solution to bypass this occasional problem could be adding a further step of inpainting only for hands in your workflow. In one of the demo images probably I used an inpaint step for hands, with a good output. Moreover one further solution is using this Lora with a little lower strength (0.7).