Beeple style for Z-Image, trained on an expanded dataset as my previous beeple LoRAs with new captioning. V2 is much better and uses the trigger word B3PL3