Will remove if requested
Description
FAQ
Comments (2)
Awesome style!
The LoRA itself is solid, but the frequent watermarks and text artifacts in the outputs suggest the original training data wasn’t carefully cleaned. Have you considered retraining with a refined dataset?