3.5: Base model NoobAI VPred
A few fewer training images compared to 3.0
对比3.0减少了一点训练图片
3.0: Base model Obsession (Illustrious-XL) - v-pred_v1.1 | Illustrious Checkpoint | Civitai
This is a NoobAI+Illustrious fusion model, and perhaps Lora is also generic
这是一个NoobAI+Illustrious融合模型，或许Lora也具有通用性
Tips: cleats,black cleats,green cleats,......
提示词：cleats,black cleats,green cleats,......
Detail is missing in panoramas, so try increasing the number of pixels or partially redrawing
在全景图片中细节会缺失，可以试试提高像素数量或局部重绘
2.0：Reduce and use single-style training assets, but rely on large models to still generate shoes of different colors.
Trigger: sxzs \(shoes\),
optional: red shoe,green shoe,
1.0
Old canvas sneakers This is a lora without convergence, but it still works.
Change the color trigger word weight > 1 example: (white qxz:1.2),
weight:0.8-1