I created a viewfinder.

This model is trained only on the minimum necessary hierarchy. There's no need to specify hierarchy with "LoRA Block Weight".

The impact on the checkpoint model is minimized. When mixing with other models, there's a possibility that the model may not function properly. In that case, try increasing the weight of the "viewfinder" or "recording" prompt.

This is my first attempt, so I would appreciate it if you could share your comments and the generated images.

The aspect ratio of the generated image may cause the frame to be cut off. Please fix one side and adjust accordingly.