Trying another approach with this newest version, and it feels like it paid off. Most of the images used in this dataset were ai generated based off Ghibli's style and some mixture of retro aesthetic aswell with themes based off; dark fantasy, sci-fi, post-apocalypse, general fantasy, etc. It rewards having a creative prompt, trust me.
Recommended settings:
Strength 0.9, lower for more flexibility. Same goes for your cfg, anything below 4.5 is golden. Also try not to use too many LoRAs together as it cam deminish the quality of your image fairly quickly.
A few trigger words that will completely improve not only the quality of the image, but the overall end result:
scenery, anime_coloring, anime_screencap, detailed lighting, framevault, movie still
Thank you again for using my LoRA, you can support me through either donating Buzz, or you give a thumbs up, follow me or you can also use one of the links provided beneath my profile, anything helps really.
Reworked the dataset again, and trained it on my new settings. Hopefully this version is a bit more flexible and consistent.
1. The rights to reposted models belong to original creators.
2. Original creators should contact SeaArt.AI staff through official channels to claim their models. We are committed to protecting every creator's rights. Click to Claim
