This LoRa is created for semi-realistic and realistic models, it contains 65% real images, it is not recommended to use it with Anime models.The weight of this LoRa can be set from 0.1 to maximum 1.3, if you want a more precise characterization.Generally 0.7 to 1.0 gives good results. Above 1.3 you can get caricatures, exaggerated features and deformities.If you want to use it with anime models, put realism requests in the prompt and avoid completely flat models. Some 2.5D anime respond quite well to LoRa.
