I trained the Lora with own prompt results with the current sdXL_v1.0VAEFIX.Saftensors Model and create with the 1024x1024 resuts a SDL1.5 Lora 512x512.

The aim was to create a nice mixture of styles. a bit of pixel art, a bit of Aquarell

It works just with lora, no extra prompt is needed. Lora weight between 0.7-1 brings in my opinion nice results.

The example pictures are made with easy prompt (e.g. a portrait of a red cat), the LoRa and as base model the amazing Dreamshaper https://civarchive.com/models/4384/dreamshaper

it is my first LoRa so I am happy for every input/feedback from you. Enjoy and have fun with it