Hi guys. I need some help understanding something. I donwloaded this model: strangerzonehf/Flux-Midjourney-Mix2-LoRA · Hugging Face on my computer. I run the same set of keywords but I get a completly different results. Online works really well. Instead, from my local machine running stable diffusion I dont get the same results regarding quality, and overall accuracy. Can someone help me with this? What am i doing wrong?
If you are able to generate images and the content and quality are slightly different, then you probably haven’t made a mistake.
The software used to generate images on the Hugging Face Hub, called Diffusers, and the software used in local GUIs, such as A1111 WebUI and ComfyUI, are different software with the same purpose. It’s a bit like the difference between Windows and Mac. The results will also be slightly different. In most cases, it won’t be a problem…
Also, you can use Diffusers locally, but even in that case, I don’t think it will often have exactly the same status as the Hub. (I think it will have almost the same status if you load it with bfloat16, but the Hub’s detailed settings are a mystery.)
However, I think it is fairly easy to make the output similar, and in theory, it should be possible to reproduce exactly the same output in some cases. (Same settings & Seed, same model, same accuracy, same prompt…)