1 weight for generation / 1.2/1.5 weight for img2img

this model was trained with one image, so 2 weight will over fit and show the original image.

Keyword: motion blur, timelapse