This is my second LoRA i ever created but it seems to work wery nice so I decided to share it.LoRA strenght 0.4 - 0.6 CFG Scale from 7 to 20Sampling steps 30+best to start with prompt " A photo of (urszula woman) "

Urszula Debska

Updated version is train on differnt checkpoint witch is analog maddnes. I play a beat with the source pics so it should work better than previous one. Weight: 0.6 - 0.7 works best while using trigger word.

Files with SHA256: 9c18d938ec8f4a51245e398fbfc8215f460a0c1b3b40ff67192595d86e4feb79