Trigger word: @mbharus

This LoRA is trained to capture the original art style while preserving most of the base model's characteristics.

For the closest possible reproduction of the original style, I recommend using the Anima base model.

P.S. I spent days trying different training configurations to get the best possible results for this style while minimizing overfitting and preventing the LoRA from reproducing specific images from the dataset. As a reward, could you share some of your generations with me? I'd love to see what you create with it, lol.

If something is wrong with the model or you have any concerns, please contact me. I will remove the model if necessary.