Still the same ancient Chinese costumes - Hanfu. But used in winter
LoRA is trained from 280 evenly quality images, you can be quite comfortable in using and creating. But I still suggest the safe range when using LoRA is from 0.5 to 0.8
Trigger words: Hanfu, Winter_hanfu, cloak
Use DPM++ 2M Karras or Euler a
Use 40-ish sampling steps
CFG Scale: 10
Suggestion to use model: SDVN7-RealArtXL
Special thanks: buzimage https://civitai.com/user/buzimage