The Lora is trained on MidJourney Images that use my "Personalised Style" + Rococo Steampunk.

Works without the trigger words ArsMJStyle, Rococo Steampunk

Works great in the 0.2 - 1.5 range.

Start ~1 and adjust according to your preference.

I experimented with the way I create the datasets for the LoRAs with this one. If you prefer it please hit the "Like" on this model - this is how I will judge whether to continue using this approach or return to my "normal" one.