<h1 id="hello-this-is-ether-blu-mix-ragtu1he4">🔹☁️ Hello, this is Ether Blu Mix ☁️🔹</h1>Ether Blu Mix is a stylized anime model specialized in creating character illustrations with vibrant colors, expressive backgrounds and a painterly feel. Based on Silicon29 and Counterfeit 3, I consider this my interpretation of a pastel-mix style of model.✨ Please Share Your Cool Creations Below! ✨<h2 id="*-version-7-update:-p4hiq6i3j">* Version 7 Update:</h2>Ether Blu 7, while being largely a weights focused change, looks to further enhance the lighting focused, slightly niji feel of EBM 6 as well as its painterly origins.Might be a little overcooked. Some prompts and generations, particularly those with heavier quality related keywords and negative prompts, are more likely to experience this. Please use the recommended negative prompts to help alleviate it.VAE is baked, same one I've always preferred.Thanks everyone, have fun. I'll be looking forward to your feedback!<h3 id="-vivk5jkf8"></h3><h3 id="*-huggingface-archive-48zjt7zvk"><a target="_blank" rel="ugc" href="https://huggingface.co/gamerdan69/EtherMix">* Huggingface Archive</a></h3><h3 id="-3aoj10e1s"></h3><h2 id="generation-recommendations-1uy2kb66b">☄️ Generation Recommendations</h2>Use whatever sampler, steps, CFG, clip-skip you prefer.Sample images were generated with<ul><li>Sampler: DPM++ 2M Karras</li><li>Steps: 20 - 28</li><li>CFG: 7 - 11</li><li>Clip Skip: 1</li><li>ENSD: 31337</li></ul>Increasing CFG can lead to more vibrant results.Negative Prompt Example:"NSFW, (worst quality, low quality:1.3)"I recommend this simple negative prompt if you wish to achieve a similar style to the preview images. You can further alter positive and negative prompts based on the type of style you'd like to achieve.Style focused negative prompt examples:* Photorealism feel:<ul><li>"sketch, drawing, anime, cartoon, illustration, NSFW, (worst quality, low quality:1.3)"</li></ul>* Flat Anime/Cartoon:<ul><li>"realistic, photorealistic, photograph, 3D, NSFW, (worst quality, low quality:1.3)"</li></ul><h3 id="recommended-vae:-885yegsew">Recommended VAE:</h3>The VAE is baked:<a target="_blank" rel="ugc" href="https://huggingface.co/hakurei/waifu-diffusion-v1-4/blob/main/vae/kl-f8-anime2.ckpt">Waifu Diffusion VAE kl-f8-anime2</a>Strongly recommended as the model was tuned with this VAE's color and saturation in mind.Sample images were further upscaled using AUTO1111 Hi-res fix. Please utilize it if you wish to achieve similar results.<ul><li>Latent (nearest-exact)</li><li>Upscale by: 2</li><li>Denoising strength: 0.55</li></ul>Are your results changing too much after Hi-res fix? Please try <a target="_blank" rel="ugc" href="https://openmodeldb.info/models/4x-Remacri">4x-foolhardy-Remarci.</a>(Also recommended for simple style illustrations, anime, cartoon, etc.)<ul><li>4x-foolhardy-Remarci</li><li>Upscale by: 2</li><li>Denoising strength: 0.45 - 0.55</li></ul><h2 id="-p6l412297"></h2><h3 id="prompting-advice-wn746qbr4">🖊️ Prompting Advice</h3>The model was mixed and tested using these prefixes and will give you the most consistent style results. Of course you're free to prompt however you'd like as booru-style models tend to be quite flexible in prompting.If you look at the prompts included with the sample images they demonstrate that even with a very loose and vague style of prompting you can achieve fun, high quality images.Experiment and give the AI some freedom to interpret your ideas!General guidelines:<ul><li>Prompt in the style of anime/booru-style models, "masterpiece, best quality, ..... "<ul><li>More anime style images</li></ul></li><li>Prompt in the style of photography/non-anime models, "professional photograph of ..... "<ul><li>More realistic images, probably higher style variety</li></ul></li></ul><h2 id="-yzz9wbwzl"></h2><h2 id="model-merging-info-v7-5r1wi9baj">📘 Model Merging Info - v7</h2>I have no training or real knowledge of machine learning. I merged the model using the information provided on <a target="_blank" rel="ugc" href="https://huggingface.co/WarriorMama777/OrangeMixs">WarriorMama777's AOM model page</a> as well as the info from <a target="_blank" rel="ugc" href="https://rentry.org/BlockMergeExplained">these</a> <a target="_blank" rel="ugc" href="https://rentry.org/Merge_Block_Weight_-china-_v1_Beta">Rentrys.</a> I very arbitrarily mixed the UBM weights of each model, modifying the values with trial and error until I was visually satisfied with it.I utilized the <a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger">SuperMerger</a> extension for AUTO1111 by hako-mikan.<h3 id="please-refer-to-the-'about-this-version'-for-more-detailed-information.-xticxffkg">Please refer to the 'About This Version' for more detailed information.</h3><h3 id="-i9ebjofa5"></h3><h2 id="credits-8kc79aht5">💙 Credits</h2>Thank you to WarriorMama777 for providing <a target="_blank" rel="ugc" href="https://civarchive.com/models/9942/abyssorangemix3-aom3">AOM</a> and other various mixes as well as detailing your workflow and inspiring me to try mixing my own models.Thank you to Luna for sharing your process on creating <a target="_blank" rel="ugc" href="https://civarchive.com/models/3850/kenshi">Kenshi</a> and <a target="_blank" rel="ugc" href="https://civarchive.com/models/40199/aurora">Aurora</a>. It deepened my understanding of model making as well as inspired me to mix LORA's.Thank you to hako-mikan for creating the <a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger">SuperMerger</a> extension allowing a quicker workflow for merging.Thank you to gsdf, rqdwdw for <a target="_blank" rel="ugc" href="https://civarchive.com/models/4468/counterfeit-v25">Counterfeit.</a>Thank you to Xerxemi and Xynon for <a target="_blank" rel="ugc" href="https://huggingface.co/Xynon/SD-Silicon">SD-Silicon.</a>Thank you to YozoRaAru for <a target="_blank" rel="ugc" href="https://civarchive.com/models/12262/night-sky-yozora-style-model">Night Sky YOZORA</a> and <a target="_blank" rel="ugc" href="https://civarchive.com/models/21200/color-box-model">Color Box.</a>Thank you to DynamicWang for <a target="_blank" rel="ugc" href="https://civarchive.com/models/84476/awpainting">AWPainting.</a>Thank you to Jemnite for <a target="_blank" rel="ugc" href="https://civarchive.com/models/89145/lunar-radiance-mix">Lunar Radiance.</a>Thank you to CTD for <a target="_blank" rel="ugc" href="https://huggingface.co/closertodeath/ctdlora">Dpep4, Xlimo and lightingv2.</a>Thank you to dolphinz for <a target="_blank" rel="ugc" href="https://huggingface.co/dolphinz/stlora/blob/main/loras/sttabi_v1.4-04.safetensors">Sttabi.</a>Thank you to SatyamSSJ10 for <a target="_blank" rel="ugc" href="https://huggingface.co/SatyamSSJ10/Satyam_Public_LORA/blob/main/Jordan_3.safetensors">Jordan 3.</a>Thank you to ashen-sensored for <a target="_blank" rel="ugc" href="https://huggingface.co/ashen-sensored/lora-isolation-collection">breakA lighting isolation.</a>Thank you to Terry9993 for <a target="_blank" rel="ugc" href="https://civarchive.com/models/95007/niji-default-style-niji">Niji Default style.</a>Thank you to those at the <a target="_blank" rel="ugc" href="https://huggingface.co/LMFResearchSociety">LMFResearch Society</a> for their archive of SD LORA's and checkpoints.Thank you to AUTO1111 for creating the web UI.Thank you to StabilityAI for starting everything with Stable Diffusion.Thank you to CivitAI and Huggingface for providing places where open source AI can flourish.Thank you to the entire SD community for continuing to openly share and create.

Ether Blu Mix

* Version 2 Update:With version two, problems such as unprompted NSFW generations as well as poor quality hands and fingers have been reduced. The tradeoff seems to be a slight loss in realism.If you prefer the realistic look of v1, please add "realistic" to the start of your prompts for a similar look in v2.Placing "NSFW" into the negative prompt is still recommended if you're concerned about occasional nudity or more suggestive generations as the mix is still biased towards suggestive and NSFW imagery.Since "NSFW" can greatly alter the output, I recommend experimenting with adding or removing it.From my short amount of experimentation: including NSFW tends to produce more stylized generations, while excluding it tends to introduce more realistic elements. (?)📘 Model Merging Info - V2I have no training or real knowledge of machine learning. I merged the model using the information provided on <a target="_blank" rel="ugc" href="https://huggingface.co/WarriorMama777/OrangeMixs">WarriorMama777's AOM model page.</a> I very arbitrarily used the UBM weights given to merge and create AOM3 and its derivatives and modified the values with trial and error until I was visually happy with it.I utilized the <a target="_blank" rel="ugc" href="https://github.com/hako-mikan/sd-webui-supermerger">SuperMerger</a> extension for AUTO1111 by hako-mikan.▼ Merging<ul><li>MODEL A: AOM3.safetensors</li><li>MODEL B: pastelmix-better-vae-fp16.safetensors</li></ul>Weight Sum merge - Use MBW<ul><li>Alpha: 1</li><li>0,0.7,0.7,0.7,0.7,0.7,0,0,0,0,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0.7,0,0<ul><li>Model: AOM3-PastelSumS2-07</li></ul></li></ul>▼<ul><li>MODEL A: AOM3-PastelSumS2-07</li><li>MODEL B: deliberate_v2.ckpt</li></ul>Weight Sum merge - Use MBW<ul><li>Alpha: 1</li><li>0,0.3,0.3,0.3,0.3,0.3,0,0,0,0,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0.3,0,0<ul><li>Model: AOM3-PastelSumS2-07-DelibV2SumS2-03</li></ul></li></ul>▼<ul><li>MODEL A: AOM3-PastelSumS2-07-DelibV2SumS2-03</li><li>MODEL B: chilloutmix_NiPrunedFp16Fix.safetensors</li></ul>Weight Sum merge - Use MBW<ul><li>Alpha: 1</li><li>0,0.2,0.2,0.2,0.2,0.2,0,0,0,0,0.2,0.1,0.2,0.2,0.2,0.2,0.2,0.1,0.1,0.1,0.2,0.2,0.1,0.2,0.2,0.2<ul><li>Final Model: EBM-v2</li></ul></li></ul>Models Used:<ul><li>AOM3.safetensorsD124FC18F0232D7F0A2A70358CDB1288AF9E1EE8596200F50F0936BE59514F6D</li><li>chilloutmix_NiPrunedFp16Fix.safetensors59FFE2243A25C9FE137D590EB3C5C3D3273F1B4C86252DA11BBDC9568773DA0C</li><li>deliberate_v2.ckpt9ABA26ABDFCD46073E0A1D42027A3A3BCC969F562D58A03637BF0A0DED6586C9</li><li>pastelmix-better-vae-fp16.safetensorsD01A68AE76F97506363F387F5F28BB564AD9E20924844FD5945E600B72D39E79</li></ul>

Files with SHA256: dc8c2409c851d060610e3b979817032e68923c474a652e4538f594019c1dd11f