A LoRA model for Dongari's style and characters. This model has been trained on 186 images for about 2000 steps.Tags used to train this model:Dongari: the style tag that was included for all images, trained on a total of 186 imagesDayeong: the character Dayeong (다영), trained on a total of 16 imagesEunhee: the character Eunhee (은희), trained on a total of 6 imagesGaeul: the character Gaeul (가을), trained on a total of 32 imagesGahin: the character Gahin (가인), trained on a total of 40 imagesNari: the character Nari (나리), trained on a total of 43 imagesSuyeon: the character Suyeon (수연), trained on a total of 8 imagesYerin: the character Yerin (예린), trained on a total of 37 imagesThe character tags, per se, don't seem to be sufficient for invoking a character, so you are likely to need to resort to adding descriptions of the character yourself. For example, for the image of Gahin, I used the following tags in addition to the character tag: black hair, purple eyes, hair clip, tattoo.I have no explicit training license to the images used to train this model. However, it seems that no legal restrictions are stopping me from doing so at the moment. Feel free to let me know if the case changes. I will remove the model as soon as possible.
