Supports simultaneous use with other character models!
The trigger word is "Super perspective", and you can also use words such as "fisheye, wide shot, foreshorten, vague, lighting stance" and perspective words such as "from above, from below" to achieve better results.
The training set of this Lora model includes 100 exaggerated perspective and combat pose character images, supporting both white and simple backgrounds, as well as 2D and 3D. You can use control net to achieve better results.
Bottom mold: meinamixV11 768 * 768
支持同时和其它人物模型共同使用!
触发词是“Super perspective”,你也可以同时使用“fisheye,wide shot,foreshorten,vague,fighting stance”等词语以及“from above,from below”等视角词获得更好的效果。该lora模型的训练集包含了100张夸张透视和战斗姿态的人物图片,支持白色背景和简单背景,同时支持2D和3D。你可以使用control net获得更好的效果。
底模:meinamixV11 ,768*768
Description
FAQ
Comments (7)
會更改到大模型的風格
不過拿來找角度上的構圖還滿好用的
您好楼主大大,可以请教一下您这个透视lora是如何训练的么,好厉害,方便加您联系方式么,
没啥别的好办法,这一版是训练集堆出来的,还有改进空间,我的Q:401708040
you should train a pony or illustrious Loras or share the dataset if possible
I'd love to see a Illustrious or Pony version, I can manage to get this to somewhat work on certain Illustrious checkpoints.
Yet CMD pops up that an error occurs, sometimes I can just ignore it.
[LORA] LoRA version mismatch for KModel: D:\*Redacted*\webui\models\Lora\superperspectivev1-000006.safetensors
Other times the perspective doesn't properly function & need to temporarily unload it.






