<h3 id="presetmodel-for-promptless-generation">Presetmodel for Promptless Generation</h3>Here is the next fresh-out-the-Box Model for my Promptless Workflow. Its a Model working with 100 Userprompts from CivitAI RED. Category: AnimeThats why, not fully nsfw...but not only sfw. A colorful Mixture.Results depending on the used Basemodel from Krea2 Familiy. Have Fun with that, and success while discover the Posibillities. Workflows are inside the Image below.Once left... this Model is based on Tokendense from 512 Token each Weightbomb a little bitchy with compiling the Theme and Charactermodels.Krea2 is wortking better, the more Token/Vectors are incomming. But full is full....But i working hard on it - dev goes on.Best you merge (mux) this Preset with: <a target="_blank" rel="ugc" href="https://civarchive.com/models/2827569/presetkreacivitaifromgiminic80">Presets trained Prompts from </a>@giminic Weights to 0.5 in Workflow/ Switch ForFromAA_ to true. Set Themes/Char Weights to 0.4.Anime is a well trained and weightful Part in Krea Models. Its nor easy to overweight theVectors - but like i said - its Prototyping yet. Dev will solve soon. Note: This Model can only used with my Workflow and the special Customnodes in ComfyUI. It is a transformed Promptcollection, what makes it possible to shorten Pipeline and Generatingtimes.With this System, i aimed to change the Generatingprocess in a different Way. Mainly Prompts can merged or selected by view on Results. Basemodel are all Krea2 Versions and its weights. This is the first Stage and like i told - experimental. Generating Outputs, you can see on my Profile.Instructions to use, you find on my Article Generatingguide krea2. Including the Workflow and the needed Customnodes.Drop Model in Path: ...\ComfyUI_windows_portable\preset_storageDetailed Instructions you find <a target="_blank" rel="ugc" href="https://civarchive.com/articles/33385/krea2-beyond-prompts-a-modular-preset-architecture-for-generative-image-pipelines">here and in my Workflows from Images.</a>

Files with SHA256: 46412f6ea48934ef91e58ae6e9dde57782f0a19adf3cc5fc6dec1e0d0a8f6a35