I trained this loRa on all sorts of fantasy stuff, some scifi stuff too but mainly fantasy,
I also noticed it adds nice detail when using it with junk juice Velvet V3 (I use it around 0.6-0.7)
(I did not use hi-res fix on the sample images, the lora adds a nice touch of magic to it 😊)
My advice use this as a prompt Template:
masterpiece, best quality, absurdres, highres, amazing quality, very aesthetic, concept painting, ultrahigh definition, detailed, intricate details, semi-realistic,
(dynamic angle, dynamic pose, blurry background, depth of field, dark lighting, low key lighting, dust particles, embers), Fbb,
Trigger word: Fbb
(you don't always have to use the trigger word for the loRa to take effect, but it helps)
WARNING!!!
Make sure when using LoRas you lower CFG a bit
Adjust weights when stacking
Depending on model, you might need to higher it or lower the weights, so be mindful of how you set your weights, (some illustrious models need loRas set to 2 while others need it at like 0.5)
(If you don't listen, you are going to have a bad time.)