2.0版本使用300图训练，更好的泛化。
缺点是手稍微有点崩
--
SD1.5版本记得开高清修复
SDXL版本使用A3.1训练
男人，就是要快（）
TAG:
mavuika_(genshin_impact)Training model: Pony
Recommend PONY model: Movable figure model Pony/T-ponynai3/AutismMix SDXL
Recommend SDXL model: Kohaku-XL Epsilon
Training resolution: 1024*1024
Recommended output resolution: 832*1216
换装建议lora的权重：0.6~0.8
When it comes to changing outfits, I would suggest assigning a weight of 0.6 to 0.8 for Lora.
特别鸣谢：算力赞助 nieta.art / GPU Sponsor: Neta
Discord : https://discord.com/invite/CFpyHceF52
Recommended dataset collection and processing tools: https://github.com/deepghs/waifuc
Possibly the most user-friendly dataset processing UI currently in use： https://github.com/Eugeoter/waifuset