This lora is made to make blindfolds. The type that was trained is a bound style that goes under their hair rather than wrapped around the head, not trained on the eye mask type blindfolds. Suggested to be used at 0.6 (0.7 if being stubborn). This model was trained on NAI anime-full-pruned and generated in Based68, and angel breed. Made with the Kohya tool for AI Training through google colab. Any suggestions to improve or recommended things for the data set would be appreciated.
Description
This was trained on 69 images (not on purpose) of blindfolds. As usual a mix of real and anime were used, only about 10 were realistic though. This is trained for a wrapped/bound style of blindfold, it wasn't trained to make eye mask style. Made because some checkpoints I was messing with were horrible at blindfolds.