The recommended strength of the lora is around 1.0.
Description
FAQ
Comments (2)
Hello, I saw your excellent LoRA work. If possible, I’d like to request a custom LoRA.
Would it be possible to create a LoRA that can reliably generate a more top-down POV angle, like the example, where the POV can reach and grab the chest?
The LoRA I’m currently using tends to shift the scene into a standing doggy-style position, which causes various issues. What I want is a stable POV composition specifically for chest-grabbing from the POV perspective.
If this is feasible, I can provide reference images / pose references and details about the base model and training settings.
It's easier to understand if there are reference images.




