It’s an interesting model. It seems to have some imprinted idea about the lightning and the pose.

So far my biggest difficulties in the 20-30 attempts were

- get a plausible light for my outdoor setting

- tendency to use waist chains

- „creative“ and unexpected shape of the cuffs

- the metallic effect of the cuffs is dominant with other objects in the scene

- a lot of examples with cuffs but no linking chain between them

- the humans in the scene tend to get drawn in comic style, hair is the biggest issue here

- control of the pose

If you go for an enhanced LoRA, I suggest to train it more on the physics and give it more variety to learn the difference between the object (cuffs) and other stuff in the scene.

But I don’t want to sound too critic here, 😇 I enjoy this model very much to generate even more variations of my „handcuffed woman in a park“-theme

I was in search for ways to get better hands and thought "handcuffs" have a focus on hands - I will try that. And yes with this lora ... there are improvements. If you prompt the cuffs it will also get you those. I posted some pictures in the gallery. Its a great lora to have; thx for sharing.