This model is based on the famous manhwa Omniscient Reader’s Viewpoint.<h3 id="a-total-of-38-images-with-8-passes-and-4000-steps-were-selected-for-training.">A total of 38 images with 8 passes and 4000 steps were selected for training.</h3>To create this model, several chapters from the beginning, middle, and end of the manhwa were used to obtain an average result.Also I recommend using a model strength of 0.8 to 1. 0.8 to reduce artifacts, and 1 for a strong style identity.This Lora have problems with hands, it's always together. To solve this problem, you can reduce the strength of the model, or the following tags in the negative prompt can help "(holding hands:1.4),(hands close:1.4)"

Files with SHA256: bea35f426c738f58511a51382c81a2c31c7800c7f814f479b96faba2b924756d