GPT2-Style Anime 17 is a cinematic anime LoRA focused on fantasy anime women, magical atmosphere, luminous eyes, detailed outfits, clean lineart, dramatic lighting, and rich key visual backgrounds.
The dataset was curated from high-quality anime images, with duplicates, low-quality images, and contact sheets . The style works best for fantasy, gothic, cyberpunk, elegant character portraits, full-body anime characters, and detailed illustration scenes.
Trigger word: gpt2style
Recommended LoRA weight: 0.6 - 1.8
Recommended prompt style:
gpt2style, masterpiece, best quality, score_7, anime fantasy girl, detailed eyes, cinematic lighting, clean lineart, detailed background
Negative prompt:
low quality, worst quality, blurry, bad , extra fingers, deformed hands, text, watermark, logo
Description
Detail-boosted version of GPT2-Style Anime 17.
This version adds stronger outfit details, sharper lineart, richer fantasy backgrounds, more luminous eyes, and a more polished cinematic anime look.
Trigger word:
gpt2style
Recommended strength:
0.5 - 2.0 full details
Use this version when you want more decorative details, stronger lighting, and a richer key visual style. If the image becomes too busy, lower the LoRA strength.