Hi CivitAI members!
I've created my first style LoRA model, making it easier for you to generate wicker look. It's simple to use — just add the trigger word at the beginning of your prompt, for example:
wkrstyle a dog, ...
For me, the optimal LoRA strength for FLUX was between 1.00 and 1.20 (for SDXL: 0.50 - 1.20), but feel free to experiment with other strengths as well.
You can also add additional keywords to enhance the style further, for example:
wkrstyle, wicker dog, ...
wkrstyle, wicker style, ...
If you would like to share your images in the gallery, please UPLOAD your images via [ + Add Post ] at the bottom of the page. It helps the creators.
The workflow is available too!
Just download the model reference images and drag them into the ComfyUI.
If you enjoy my work, consider showing your support with a 👍 or ❤️ on the model or images — it really keeps me motivated! Thank you! ❤️
Have a nice day! 🤩
Description
Tips for Using LoRA:
• For simple prompts, the trigger word (wkrstyle) is often enough, or it might even work without it.
• For more complex prompts, I RECOMMEND also using the term "wicker," for example:
wkrstyle, a wicker ...
• Regarding the strength of the LoRA, I’ve noticed that if the character or object starts to look too much like a basket, it’s worth reducing the strength (e.g., to 0.7).
• I’ve tested the LoRA strength between 0.5 and 1.2, and it worked quite well within this range.
• For UpScaling, I recommend setting the denoise strength between 0.20 and 0.25 (it better preserves the woven texture)