This LORA was created via Civitai LORA training services and intended to replace the existing Nurse Luka model which I created with obsolete SD1.5 models. Works well with most of SDXL models including anime ones. Minimum Strength level: 0.7 for realistic and anime. Clip skip settings may not bother much for SDXL. Also, please enable these settings on Auto1111's SD WebUI for best performance (both are found on Settings&gt;Optimizations section):<ol><li>FP8 Weight (Drastically reducing VRAM usage, allowing GPU cards with 6GB VRAM can be used.</li><li>Cache FP16 Weight for LORA</li></ol>Note: If you want to use this LORA with another LORA, you may need to use SD Webui Supermerger to merge LORAs into checkpoints in order to produce an exact correct images when 2 or more LORAs activated by trigger words.

Nurse Luka XL

Early version of this LORA. Based from SDXL base model. Should be working well with most of SDXL models despite having some issues, thus adjusting its strength to lower ones is required.

Files with SHA256: b292eeeb85a3a75e1fe42591d5d4929be7a49e6b8c7eb5378f3be8599a120893