<p><span style="color:rgb(134, 142, 150)"><em>Daisies in flower language means "innocence and purity", while black in color theory is darkness, death, and power. With this in mind, a "black daisy" wouldn't be as simple as lost innocence, but a rebirth of ideals in the face of loss - "In darkness shall I bloom, in purity shall I bathe. An elusive beauty, dark and dangerous. Yet only a rare few can gaze." - Pandora</em></span></p><h1 id="what-is-this-model">What is this model?</h1><p>This is a Stable Diffusion 1.5 model born slightly out of hatred against "kawaii" anime models - chibi anatomy but with strong styles that can stand on its own. Initially, this was meant to be a merge that distills the essence behind those models and reverse it, but I decided to release the final results because it has surprising prompt adherence and uniqueness in it.</p><h1 id="how-to-use-this-model">How to use this model?</h1><p>First point: <strong>Please, try to exercise a little restraint in using this model. Due to its nature, chibis are much more likely to occur than usual.</strong></p><h3 id="prompting">Prompting</h3><p>This model uses booru tags like all other anime models. However, this model has somewhat decent natural language recognition as well. If you have a tag autocorrect extension, use <a target="_blank" rel="ugc" href="https://civarchive.com/models/950325?modelVersionId=1419692">Drac's tag list</a> for best results.</p><p>Model cutoff is ~2020, so don't expect too much from it.</p><p>Negative prompting is unnecessary, but could be used.</p><p>This model is meant mostly to be used as a 2nd pass in a workflow, but could be used separately as well.</p><h3 id="parameters:">Parameters:</h3><ul><li><p>Sampler (Basic): DPM++ 2M Karras or Euler ancestral</p></li><li><p>Sampler (Advanced): DPM++ 2M CFG++ SGM Uniform</p></li><li><p>Step count: 20-30 (50 for CFG++)</p></li><li><p>CFG: 7-10 (0.5-1 for CFG++)</p></li><li><p>CLIP Skip: 1-2</p></li><li><p>Resolution: 768 base resolution maximum, 512 base resolution recommended</p></li><li><p>Hi.res fix: recommended - 2x @ 0.5 denoise. If possible, do this using another model.</p></li></ul><h1 id="miscellaneous-stuff">Miscellaneous stuff</h1><h2 id="licensing">Licensing</h2><p>Check below profile card, but permission is similar to OpenSolera: No selling this model, ever. Other than that, everything else is fair game.</p><h2 id="merge-recipe">Merge recipe</h2><h3 id="v1-alpha">v1-Alpha</h3><p><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/83683c96-19f7-402b-9596-1b0ac42eb82d/width=525/83683c96-19f7-402b-9596-1b0ac42eb82d.jpeg" /></p><h3 id="v1-beta">v1-Beta</h3><p><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/c00e79bd-5838-4d85-aeb2-339b6ea0d051/width=525/c00e79bd-5838-4d85-aeb2-339b6ea0d051.jpeg" /></p><h3 id="v1-gamma">v1-Gamma</h3><p><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/9eed4ff7-e6c9-41d4-9d76-c7cb1ae59259/width=525/9eed4ff7-e6c9-41d4-9d76-c7cb1ae59259.jpeg" /></p><h2 id="cocktail-recipe">Cocktail recipe</h2><p>You read that right, I made a cocktail recipe for this as a joke. Thanks to all 4 of the big LLM service out there for helping with the recipe: ChatGPT Reason, Claude 3.5 Sonnet, DeepSeek-R1 Reason, Gemini 2.0 Flash Reasoning. I genuinely haven't tested this recipe, so if someone actually tested it, please tell me in Discord (@azeluslightvale).</p><p>"Black Daisy"</p><hr /><h3 id="portions">Portions</h3><p>Balanced Variation:</p><ul><li><p>2oz Dark Rum (recommend dry rum)</p></li><li><p>0.75oz Fresh lime juice</p></li><li><p>0.75oz Creme de Cassis</p></li><li><p>0.75oz Creme de Mure</p></li></ul><p>Blackstrap Variation:</p><ul><li><p>2oz Blackstrap Rum</p></li><li><p>1oz Fresh lime juice</p></li><li><p>0.5oz Creme de Cassis</p></li><li><p>0.5oz Creme de Mure</p></li></ul><h3 id="mixing-recipe">Mixing recipe</h3><ul><li><p>Prepare a pre-chilled rocks glass with an ice cube</p></li><li><p>Add the drinks in the given portion above with crushed ice into a shaker</p></li><li><p>Shake the mix and double strain it into the glass</p></li><li><p>Add a small pinch (35mg) of salt and stir it evenly</p></li><li><p>Garnished with slapped mint, raspberry and blackberry skewer</p></li><li><p>(Optional) For thematic, add a lawn daisy (Bellis perennis) as a final part of the garnish</p></li></ul><p></p>

[SD1.5] ⚫🌼BlackDaisyMix - "Kawaii" Anime Mix

<p>This is an attempt to "remake" a version of BlackDaisyMix with a more reproducible set of models after CocoaMix-v8 decided to disappear (with a few adjustments).<br />Checkpoints:</p><ul><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/88184/kaywaii">KayWaii - v9.0</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/28169?modelVersionId=265102">CuteYukiMix - kemiaomiao</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/210828/hotarubreed-cuteanimemix">HotaruBreed CuteAnimeMix - v7</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/600100?modelVersionId=670698">UsaChoMix</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/1060496?modelVersionId=1202628">ZenTypeEX - ex_WK</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/932103?modelVersionId=1242132">ZenTypeB - v2</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/291794?modelVersionId=1092816">ZenTypeD - v7</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/305758?modelVersionId=1035975">ZenTypeA - A2_v4</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/21138/project-k-kawai">Project-K (Kawai) - v5.1.0-Fabric</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/32355/dorayakimix">DoraYakiMix - v3.0-fp16</a></p></li></ul><p>LoRAs:</p><ul><li><p><a target="_blank" rel="ugc" href="https://huggingface.co/CluelessC/hll-test/blob/main/lora/hll6.3-a10-eps-resized-512.safetensors">CluelessC's Hololive LoRA (hll6.3-a10-eps-resized-512)</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/8765?modelVersionId=10638">Theovercomer8's Contrast Fix</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/242825?modelVersionId=273993">SD1.5 DPO LoRA</a></p></li><li><p><a target="_blank" rel="ugc" href="https://civitai.com/models/526379/spo-sd-v1-54k-p10eplorawebui">SPO SD1.5 LoRA</a></p></li><li><p>Initially, I meant to add HyperSD to this... before I realized it tanked prompt adherence so badly I decided not to bother. This is why there's a bypassed LoRA Loader.</p></li></ul><p>CLIP: <a target="_blank" rel="ugc" href="https://huggingface.co/zer0int/CLIP-KO-LITE-TypoAttack-Attn-Dropout-ViT-L-14">CLIP-KO-LITE-TypoAttack-Attn-Dropout-ViT-L-14</a></p><p>VAE (baked in): <a target="_blank" rel="ugc" href="https://huggingface.co/nubby/kl-f8-anime2-blessed/blob/main/WD1-4-kl-f8-anime2-bless09.safetensors">WD1.4-kl-f8-anime02-bless09</a></p><p>Merge recipe in model description and inside model metadata as a workflow.</p>

Files with SHA256: 62bb942aa7633611bda333ab8b5bf257a6bde1187a5b2cf47e56b014f6b4166c