Lora trained from one of my creations to maintain the character's consistency; it was made in the flux model, and I'm still testing it.
2419890_training_data.zip
Train_ana-000003.safetensors