Use V2, V1 has sometimes blurry outputs and exaggerated highlights I´ve improved the training for V2 with better images. Example prompt : cactusperson, cactus woman, samurai, black robe , blue eyes, red flower hair, holding katana with both hands This LoRA is designed to create cactus-themed characters for fantasy, everyday scenarios, or any other setting you like. It can also generate cactus-inspired animals for extra creativity. I´ve used strength 0.8-1 to create the example images. Mostly 0.8.<ul><li>Activation keyword: <code>cactuspeople</code></li><li>For humanoid characters: <code>cactus woman</code>, <code>cactus man</code> (use “male focus” for male characters)</li><li>They all have flower hair , prompt "color flower hair" , to change the color.</li><li>I´d use 0.8 strength if you want more detail from other prompts. for pure cactus use 1.0</li><li>I´ve noticed that you can cactify basically everything , using cactus x</li><li>For animals: Example prompts: <code>cactus rabbit</code>, <code>cactus dog</code>, <code>cactus x</code> </li></ul>I´ve used Novacartoon, WaiNsfw , ilustmix and my own merged models for the pictures.

Cactus People (and animals) concept

I´ve retrained the lora with better pictures , v1 was blurry sometimes and had exaggerated highlights.

Files with SHA256: 598309e384a31fcf83307045f5becf56fc203bff9e4ade600cc8efdee1b1082f