This LoRa was trained using noobaiXLNAIXL_epsilonPred11 as the base model.

prompt:

rikume,

If you encounter any issues—such as frequent watermarks or signatures—feel free to leave a comment. I'll do my best to fix them and release an updated version.