If you make some cool images, please consider sharing them, and leaving a review. That is why I make these. :)
After the success retraining Dolph, I decided to try my luck with a new character, before I retrain Red-Nosed Raider. I really like Highwire this season, and wanted to see if I could make a character who has 3 different outfits in 1 LORA.
She came out pretty swell I think. Her glasses are kinda wonky, but it gets close enough I think. Maybe this can be smoothed out later. If you weigh them higher they'll turn out better, probably with inpainting they can turn out perfect.
Supports Wolf Version and Bunny version, with hood up and down.
You can get good results with 1 image, but I generate 5 and pick the best. That's pretty much what I do with all models, but just my habit. Can be clothed, nude, etc. LORA strength between 0.8 - 1.0 is ok from my tests.
You can add texting, holding cellphone, cellphone, for texting, or you can do holding cellphone, selfie, for taking a selfie.
Other tags, like Contrail, Skydiving, Glider, etc. You can specify different views with Front View, 3/4 View, Side View, 3/4 Rear View, or Rear View.
Lots of fun to be had. :)
If you like my models, please consider a tip, to help me train more models!
https://ko-fi.com/reindeerczar
Custom dataset of 132 images, no scraping, a lot of work. :P