This model was made to recreate what the pyramids of Giza may have looked like thousands of years ago, cased with polished limestone and capped with gold. It's pretty fiddly and you might have to experiment with different checkpoints, strengths and prompts to get desired results. I've used the Juggernaut model for the test images, and between 0.4 and 0.8 strength seems to be optimal depending on the prompt. The same trigger words may not always work, but "wh1t3ng0ld style pyramid", "white pyramid", "pyramid with a golden top" and simply "pyramid" all work at times, you may just have to play around to find what's best for your prompt.