trained at 896x1232 resolution so cross-shaped pupils show up very often when prompted
better dataset and tagging makes it easier to prompt different cloths, perspectives, and poses
LoRA produces fantastic results at weight = 1.0
trained at 896x1232 resolution so cross-shaped pupils show up very often when prompted
better dataset and tagging makes it easier to prompt different cloths, perspectives, and poses
LoRA produces fantastic results at weight = 1.0