[Digianko (Ankoman)] style

This LoRa was trained using noobaiXLNAIXL_epsilonPred11 as the base model.prompt:ankoman,