I created the 30 datasets images with AI (DAL-E) based on my own prompt. LORA was created with fluxgym.

Captions: strong cinematic lighting, stylized cinematic lighting, bold contrast, pronounced contrast, glowing light sources, radiant highlights, deep shadow depth, deep shadowing, volumetric haze effects, atmospheric haze, warm fiery tones, warm orange gold colors, dramatic ambience, epic mood, detailed surfaces, high detail