Like all my LoRAs this probably exists in multiple different forms. Regardless, just thought I'd do it for fun. Think this works without trigger word, but included just in case.
I will share both 1750 / 2000 step variations of the LoRA. I think 1750 at around 1.2 strength may be best.
Comparison shots are generally a very basic prompt such as "girl on couch taking selfie" - then compared with LoRA and without. In some cases, I combined Realistic Snapshot or real sliders on top of LoRA. I think this gives best results.





