This model is used to generate middle finger gestures to make the video more interesting.

This model is trained using a 512*512 dataset. Since it is just a gesture, there will not be any large movements.

At the same time, because it is just a gesture, there is a lot of room for creativity because it will not pollute other loras.

Because other related words may be confused with existing concepts in the checkpoint, "FKY" is used as the trigger word.

It is very suitable for the character Lora.

The key prompt words are as follows：

She is raising the middle finger giving a "fky" gesture

She is holding up both hands with the middle fingers extended, giving a double "fky" gesture

I am used to generate 640*480 (4:3) and 640*368 (16:9) sizes.

I only tested the strength of 0.8, because I usually use this strength for all loras.

I'd love to see your posts and reviews.