If you like this model, please consider buying me a coffee Kofi: https://ko-fi.com/ericzhu
Here’s my first SDXL LoRA. I’ve wanted to do a SDXL Lora for quite a while. I was wondering whether I can use existing 1.5 LoRA to generate high-res images for training, since I already struggle to find high quality images even for 512×512 resolution. I heard that SDXL is more flexible, so this might be helpful for making more creative images.
I am not so familiar with SDXL prompts yet, you most likely can generate better images than showcased.
I recently get disturbed a bit by these horror themed models, so I will try other styles before I am ready. This is literally the worst model I’ve processed, in a disturbing way. I just hope this model can fulfill some of you guys’ fetishes.
All training images are generated using
Body Horror Creatures and Photon
then manually captioned by me.
Thanks to the Sibience for this very high quality SD1.5 LoRA.❤️ so I can gather training images a lot easier and include this concept in SDXL workflow.
Recommended Weight: around 1