This was trained on the base SDXL model with a dataset of images I generated using Midjourney. It's far from perfect and requires more work, but it does output much better images than generations without the LORA so I'll upload this version.
To use this LORA I recommend using the phrases cheese sculpture of... or ...made of cheese, or both, there's probably other phrases that will fine work too.
You may have luck in specifying different types of cheeses, but no types were specified in the captioning.
If you enjoy this work, I'd really appreciate it if you could leave a like and a review.
I'd also love to see your generated images too! Thank you ❤