Anima model version of my gothic anime screencap lora

Trigger word - G4SS

P.S.: This version was trained with poor hand recognition results due to a low-quality dataset. I tried retraining it with a modified dataset, but I couldn't achieve the desired style (lucky random). Therefore, you definitely need "bad , bad hands, missing finger" in your negative prompts, maybe this should help

The trigger word also seems to have no effect, because I forgot about "Keep token" and "shuffle caption". You can use trigger word just to make sure it works

Why this version? Because it's pleasing to my eyes, and I couldn't achieve the same result by retraining, even with a good dataset and tweaking training options