Since AI image generators tend to default to more "natural" poses, it can actually be quite difficult to get them to produce a jiangshi pose—which is why I trained this LoRA.

(If you are already proficient with tools like ControlNet, you likely won’t need this LoRA.)

Additionally, because including the term "jiangshi" in a prompt often forces the character’s outfit to change into a traditional jiangshi costume, I’ve set a different trigger word instead.

( not "jiangshi pose" ---> "kangsi_pose" )

When using the built-in Anima character, a LoRA weight of 1.0 or less is sufficient; however, if you are using it alongside other style LoRAs, try increasing the weight to between 1.5 and 2.0. Please note that it may occasionally fail to produce the desired result.