"FluxFromBehind" is a LoRa model designed for the FLUX model, specializing in generating highly detailed and realistic images of female vagina, ass, anus, in a "bent over from behind" pose.
This LoRa is trained on images specifically matching the positions showcased
If you're having trouble getting subjects to face the camera, that's because the algorithm includes the trigger word 'facing away,' ensuring that generated images depict people from behind with their heads turned away.
For the best results, be sure to pair this with the 'Noisify' LoRa and its trigger words to create realistic, low-light scenes.
Description
This LoRa is trained on images specifically matching the positions showcased
If you're having trouble getting subjects to face the camera, that's because the algorithm includes the trigger word 'facing away,' ensuring that generated images depict people from behind with their heads turned away.
For the best results, be sure to pair this with the 'Noisify' LoRa and its trigger words to create realistic, low-light scenes.
FAQ
Comments (4)
This waste of training carbon emissions is not only way too high in dimensions, but is worse than SD15 quality, there is no point running Flux if you might as well go back to stable diffusion and get massively better results.
Any future LoRas shall be posted on Tensor, so do ensure you follow me there as well, as I do not endorse censorship on this platform. It is rather too snowflaky for my taste!
To get this to work in Forge you need to add <lora:lora-000014.TA_trained:1> and any activation words manually. Usually Forge add those when you select the LORA, but it does not for this one.
can you make LoRa like this with two angles? so the generated image will be a grid image with the first grid being the frontal view, and the second grid being the rear view, i think that would be a great LoRa
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.