When creating a style through the API, images could be automatically generated in a few different categories, in the spirit of amiantos/AI-Horde-Styles-Previews.
For example, it could generate an image of a person, place, and thing. In addition, the preview images would also have some additional metadata that supplies the prompt. InputStyleExamplePost would have an additional (optional?) parameter, prompt, which contains the prompt used to generate the image.
This also ensures that the preview images accurately represent the style.
If the style was changed, the previous images could be deleted and new ones generated.
When creating a style through the API, images could be automatically generated in a few different categories, in the spirit of amiantos/AI-Horde-Styles-Previews. For example, it could generate an image of a person, place, and thing. In addition, the preview images would also have some additional metadata that supplies the prompt.
InputStyleExamplePost
would have an additional (optional?) parameter,prompt
, which contains the prompt used to generate the image. This also ensures that the preview images accurately represent the style. If the style was changed, the previous images could be deleted and new ones generated.