← sources

I really like the Gaussian Splash LoRA for 2511 and really wanted it on Qwen-Image 2.1. So I trained one for it!

source: reddit (reddit:StableDiffusion)published 2026-10-26 15:39ingested 2026-10-27 02:38

retained text

https://preview.redd.it/3g8v08bixtxh1.jpeg?width=3740&format=pjpg&auto=webp&s=466ed907730b2fcc2bc42662ec7a7203ece1310a

I really liked and missed this lora from 2511

So I trained my own, It turned quite well! 

Old LoRA (2511): [https://huggingface.co/dx8152/Qwen-Image-Edit-2511-Gaussian-Splash](https://huggingface.co/dx8152/Qwen-Image-Edit-2511-Gaussian-Splash)

My LoRA (Qwen-Image 2.1): [https://github.com/aaljman/splat-repair-lora](https://github.com/aaljman/splat-repair-lora)

(This version is trained to be used with prompting, since that's the way I used to use the 2511 one. If you liked how plug and play the 2511 LoRA was, you might be a bit disappointed, but this gives more control.)

document fetched 2026-10-27 00:32 · canonical https://reddit.com/r/StableDiffusion/comments/1xg0ann/i_really_like_the_gaussian_splash_lora_for_2511