My HuggingFace: <a target="_blank" rel="ugc" href="https://huggingface.co/jspsoli">https://huggingface.co/jspsoli</a>The starting point of this model was a Anima realistic finetune merge.I baked this lora into that finetune: <a target="_blank" rel="ugc" href="https://civarchive.com/models/2650553/anima-edit">https://civarchive.com/models/2650553/anima-edit</a>Then trained a lora exclusively for uncensoring on top of that.Finally, I merged the resulting lora into the model it was trained on and this is the result.(lora diff version: <a target="_blank" rel="ugc" href="https://civarchive.com/models/2804197/anima-uncensor-v1-merge-finetune-lora-diff">https://civarchive.com/models/2804197/anima-uncensor-v1-merge-finetune-lora-diff</a>)REQUIRES: <a target="_blank" rel="ugc" href="https://github.com/Mirumo0u0/ComfyUI-Cosmos-Reference">ComfyUI-Cosmos-Reference</a> custom nodeBASE PROMPT (ALWAYS USE AS PREFIX): remove bar censor and mosaic censoring. nsfw, uncensoredNegative prompt: censored, bar censor, mosaic censoringBasic Workflow with masking support included - check 'Optional Files'.EDIT: forgot to mention in the workflow instructions - so you might miss it - but the 'base prompt' is automatically added as prefix to your prompt. the text prompt area of the Main Subgraph is simply a suffix to thatTraining data:<ul><li>Image Types:</li><li>hentai-like images (~70%)</li><li>doujinshi (~10%)</li><li>realistic content (~20%)</li></ul><ul><li>Censor Types:</li><li>gaussian blur (~5%)</li><li>bar censors (~70%)</li><li>mosaic censoring (~25%)</li></ul><ul><li>Censored Anatomy:</li><li>pussy (~40%)</li><li>anus (~5%)</li><li>penis (~50%)</li><li>nipples (~5%)</li></ul>Its a very limited test model but I'm sharing in case someone finds it useful.

Files with SHA256: 861077d98bda3aef26aa8618f247097b396fb5a4065426ac3d715d5b1ee956e1