DEPRECATED - I've created a MUCH better and more reliable model, available HERE

I just can't figure out how to easily put it into the versions on this model. Oh well.

This was an experiment targeted toward creating black gradient arms and legs.

Additional training for high heeled feet, which sometimes works and sometimes does not.

Once upon a time, I put out a bounty for this concept. Two people created LoRA's for the bounty, but neither gave consistent results.

I figured it was time to just buckle down and give it a try myself. And now I have one that produces ever-so-slightly more consistent results! Progress!

I'm going to use this LoRA to try to produce some high quality training images that clearly display the concept, combine that with the best of my dataset for training this model, and eventually hopefully make a second version that gives properly consistent output.

The V2 model is now available and works a million times better.

So any images you post will be greatly appreciated, and will go toward training v2 if they fit the bill!