Normal v0.2

I created this LoRA back in October 2024 and did not release it because it wouldn't come close to anything i expected. It is possible to create good images with it but it needs the right prompting and patience. Only tested with the full flux dev fp32 model and ComfyUI. Will probably not work with the Civtai generator.

______________________________________________________________________________________________

Reverse v1

Don't expect miracles, we are still talking about base FLUX.

It seems to be pretty flexible and works with other character LoRAs, at least for me.

No trigger words, grab some captions from the showcase and edit it.

The deis sampler worked pretty good with it.

Doesn't seem to work good on Civitai generation

_______________________________________________________________________________________________

Normal v0.1

My very first Concept/Pose LoRA

It's probably way too large for it's purpose, 64/64 - fp32

The training dataset was only 40 images for test purposes, so it's not very flexible.

Trigger words: "woman on top of a man" and "cowgirl position"