v1.0: 17k images, 6k tags. Added rating-separated versions. Deduped.<a target="_blank" rel="ugc" href="https://huggingface.co/spaces/gustproof/sd_prompts">Try the random prompt generator with better visual experience here</a>.This is a set of wildcards based on real tags from images on Danbooru. The images were randomly chosen from artworks created by artists in a curated list. Each image matches at least the tags <code>1girl solo original rating:g,s</code>.The general tags are broken down into 4 rough categories:<ul><li>Character-specific: These are attributes that would remain consistent even if the depicted character were engaged in a different activity or wearing a different outfit. This category includes body attributes and hairstyles, but it excludes accessories.</li><li>Outfit-specific: This pertains to what the character is wearing or not wearing. It encompasses anything that indirectly describes the outfit or implies the presence or absence of a particular outfit.</li><li>Action-specific: This refers to the actions that the character is performing. It covers poses, expressions, and other actions, as long as they do not directly reference the outfit.</li><li>Scene-specific: This relates to the image's setting, encompassing the location, background elements, composition, and essentially everything that does not fall under the other three categories.</li></ul>Since there are only 4 categories, this set of wildcards is more coarse-grained than most others. Note that some tags are inherently ambiguous and tags from different categories may contradict each other.The tags were first zero-shot categorized by GPT-3.5 Turbo with the definitions above, but the human ended up redoing ~15% of the tags. The Danbooru tag wiki is nice but sadly the tag groups aren't as comprehensive as desired.Made to make testing models easier.All sample images used the prompt template <code>__characters__ , __outfits__ , __actions__ , __scenes__ ,</code>(general variants for v1.0).

Files with SHA256: bd5fd0d4707776c5d52383a405da4dd88f2ecfbc82b47f662ce9c8e36910b26d