Trained on images featuring characters with their back to the viewer , facing a large structure or landscape: https://imgur.com/gallery/immersuve-worlds-lora-training-data-1wEXmSX
Example prompts:https://pastebin.com/ERm5uYGb
Example prompts 2 : https://pastebin.com/URLqwjkw
FLUX Denell works well at 10 steps , 3.5 CFG , using Euler with 'Simple' Scheduler