This LoRA was trained exclusively using only recent high-quality illustrations. I also added 5 images of the artist’s own 3D work to help smooth the quality.Unfortunately, a few watermarks slipped into the dataset sorry about that.trigger word: lifi_rebisRecommended strength: 0.8/1.0negative prompt: comic, empty speech bubble, english text

Files with SHA256: 379f353b28ca6383bb95f5c761d94e489cee3fe771f927c2d9b2d4fa506e7c5a