This is to assist image generation models apply random keywords within a prompt; in this instance the color of the subjects eyes. Base prompt used: "Extreme close up shot, focus on eyes,beautiful 18 y.o. woman with __naturalEyeColor__ eyes,highly detailed skin with hairs"Here are some examples:<pre><code>Extreme close up shot, focus on eyes,beautiful 18 y.o. woman with Brown eyes,highly detailed skin with hairs</code></pre><pre><code>Extreme close up shot, focus on eyes,beautiful 18 y.o. woman with Blue eyes,highly detailed skin with hairs</code></pre><pre><code>Extreme close up shot, focus on eyes,beautiful 18 y.o. woman with Green eyes,highly detailed skin with hairs</code></pre><pre><code>These would be randomly generated by the following prompt: Extreme close up shot, focus on eyes,beautiful 18 y.o. woman with __naturalEyeColor__ eyes,highly detailed skin with hairs</code></pre>It should be saved as a .txt file in the wildcards directory if generating images locally in Fooocus, and wildcards_official for Ruined Fooocus.For A1111: This is a Wildcard collection, it requires an <a target="_blank" rel="ugc" href="https://github.com/AUTOMATIC1111/stable-diffusion-webui-wildcards">additional extension in Automatic 1111</a> to work on that platform. (See the guidance to the right on this page).Usage: Simply put two underscores around the filename prefix. i.e.(__naturalEyeColor__).Keep in mind that your checkpoint or lora may not recognize the color and will produce a 'default' response, and images generated are highly dependent on the checkpoint and lora training material; larger scale checkpoints are likely to produce more hits than the more specific ones.Comments, suggestions, amendments and inclusions most welcome in the comments.(Note that 'red' is included as it is present in people with Albinism (albinos). Feel free to remove it, but it does look cool. Also you can of course simple use one of the other color based wildcard lists to get even more variety.

Files with SHA256: 4e55ad08858eb16408470fb18170df162b42e1f5d2c9a1020b52a0cbf67166b9