This is without doubt the most ridiculous Wildcard I've ever made. I stumbled across a database of metadata for two of the largest and most popular adult movie websites. It turns out that between 2007 and 2013 nearly 2 million movies were uploaded to these two sites. So I did the obvious thing, I took the titles of every one of those movies and used them to create a prompt that would summarise the movie.The full prompt that I used was:(score_9,score_8_up,score_7_up:0.5), photo (medium), (@wildcard:1.1), medium shot, lens flare, warm colors, dramatic lighting, (film grain:1.3).I used my TMBMix SDXL model with no LORAs. I obviously haven't checked all 2 million prompts, and the ones that I have seen come in lots of different languages so I don't know what a lot of them even mean. But they pump out some pretty interesting images!

Files with SHA256: cec3b854cdb134f3e9b5cde53b90dab7256b32330fcbd783fe52173e0163fa66