This is my first LoRA trained on the style from the animated television series "Gravity Falls".

I have had good results with a weight around 0.5 although this may depend on what model you use. I have had the best results by using these in the prompt:

cartoon, in the style of gravityfalls,

Given that nearly every outdoor scene in the show has trees in the frame, they do tend to show up in the generated images. I plan to work on a v2 to help improve that as well as character faces.

Combining this with an anime model creates a unique style for characters as well that you can see in the examples.

Note for example images: some have been up-scaled or inpainted to improve details, so you may not get the exact same results with the seeds.

ALSO make sure you are using the correct LoRA name in your prompt if you are copying from the example image generation data. It needs to match the file name you downloaded, which may change as the version increases. The examples may have an outdated LoRA name in the prompt.

UPDATE: I have also just discovered that the seeds from the examples were overwritten with different ones when I up-scaled, so I will try to find the original seeds and get those updated.