This is a LORA of VI from Arcane. This LORA is for Z-ImageTurbo. LORA Strength: If 0.9 is too much try lower it to 0.6. It always depends on: How much trigger words/sentences you used to describe a scene, tho more you use the stronger it needs to be. If you use less tags/words/sentences then lower the LORA too. This counts for all LORAS usually. Since i train my Loras with more Head Pictures than Body, it should work because it will still generate her Head if the body is not what you want. So if you lower the LORA strength you will recieve her head but less her body so you can change the body like you want while her head remains. (maybe strength 0.6-0.7). Each Image uploaded and Video shows the workflow. No tricks, no cheating! - My Images of this LORA are the exactly same i used to generate those pictures, i hope it will help you peoples to see and understand how this works. Maybe it also helps someone generate his first video or image right out from scratch. The videos need to get reviewed by moderators before they show up, so check out the page here later to download the videos i uploaded. How it works? To get this VI Look, use this LORA and type in things like "Red-Black Hair, Spiky Hairs, She has dark hair and she looks like VI from Arcane, her overall mood is dark and she wears a black biker jacket with some white wraps or bandages around her upper chest and breasts" describe her and you will get her Look. Workflow is included inside the Pictures AND Videos (Video Workflow is WAN2.2, if i uploaded videos here.) NOTE: I trained this LORA only with her head, wich means it is highly compatible with everything you want. BUT, i also trained it on 2-3 Body Pictures of her, so the LORA also knows her Body. If the generation will show stuff you did not want, then change the lora strength from 0.8 to 0.6, as i trained more head pictures than body, it will keep her head and face.

Files with SHA256: 946f5a2cfe1afae6c0dac9fab17916ac7136098361cbbfa41e787bb79552545c