I trained this on the film Ponyo. However I removed as many of the frames as possible with characters in them in an effort to get the style used in the backgrounds of that film. It kinda worked.The style applies by default, but you can increase it or apply it to prompts that are struggling with it with <code>illustration</code> and/or <code>detailed background</code>. The lora also does interesting things with water, because Ponyo.If you want to use this for landscapes, depending on what you are doing, you may want to turn the strength down a bit as the seaside nature of all the training data is a bit strong.

Files with SHA256: 9c32c35c03a8b65bd22c9f9e0ec6ad273c6bab0d72a7e0b0eae538c00d5e5433