BadQual
The model is trained to produce natural noise, camera grain, motion blur, and a realistic photo look. Please set the adapter/LoRA weight between 0.6 and 0.8, as this range yields the best results. The trigger word to place at the start of your prompt is badqual.
Keep in mind that the model focuses on style, so for peak quality you might need additional detailers (such as Face Detailer / Hand Detailer).