If you want a portrait photo, try using a 2:3 or a 9:16 aspect ratio. Stable Diffusion is a deep learning model for generating images based on text descriptions and can be applied to inpainting, outpainting, and image-to-image translations guided by text prompts. Sensitive Content. Sensitive Content. This checkpoint includes a config file, download and place it along side the checkpoint. Stable Diffusion is a machine learning model that generates photo-realistic images given any text input using a latent text-to-image diffusion model. pt files in conjunction with the corresponding . This model is named Cinematic Diffusion. Additionally, the model requires minimal prompts, making it incredibly user-friendly and accessible. --English CoffeeBreak is a checkpoint merge model. Seeing my name rise on the leaderboard at CivitAI is pretty motivating, well, it was motivating, right up until I made the mistake of running my mouth at the wrong mod, didn't realize that was a ToS breach, or that bans were even a thing,. Checkpoint model (trained via Dreambooth or similar): another 4gb file that you load instead of the stable-diffusion-1. pth <. Some Stable Diffusion models have difficulty generating younger people. Model based on Star Wars Twi'lek race. Then, uncheck Ignore selected VAE for stable diffusion checkpoints that have their own . . 5 and 10 CFG Scale and between 25 and 30 Steps with DPM++ SDE Karras. Browse undefined Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAs Counterfeit-V3 (which has 2. code snippet example: !cd /. Browse anal Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsCivitai Helper. All Time. Stable Diffusion Webui Extension for Civitai, to handle your models much more easily. A spin off from Level4. This model is a checkpoint merge, meaning it is a product of other models to create a product that derives from the originals. It DOES NOT generate "AI face". 2. Civitai Helper . Based64 was made with the most basic of model mixing, from the checkpoint merger tab in the stablediffusion webui, I will upload all the Based mixes onto huggingface so they can be on one directory, Based64 and 65 will have separate pages because Civitai works like that with checkpoint uploads? I don't know first time I did this. Install the Civitai Extension: Begin by installing the Civitai extension for the Automatic 1111 Stable Diffusion Web UI. 打了一个月王国之泪后重操旧业。 新版本算是对2. I guess? I don't know how to classify it, I just know I really like it, and everybody I've let use it really likes it too, and it's unique enough and easy enough to use that I figured I'd share it with. Supported parameters. Some tips Discussion: I warmly welcome you to share your creations made using this model in the discussion section. This model has been archived and is not available for download. Additionally, if you find this too overpowering, use it with weight, like (FastNegativeEmbedding:0. Browse pose Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsBrowse kemono Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsUse the negative prompt: "grid" to improve some maps, or use the gridless version. 2. . Happy generati. No results found. xのLoRAなどは使用できません。. in any case, if your are using automatic1111 web gui, in the main folder, there should be a "extensions" folder, drop the extracted extension folder in there. . How to use models Justin Maier edited this page on Sep 11 · 9 revisions How you use the various types of assets available on the site depends on the tool that you're using to. Expect a 30 second video at 720p to take multiple hours to complete with a powerful GPU. This Stable diffusion checkpoint allows you to generate pixel art sprite sheets from four different angles. No one has a better way to get you started with Stable Diffusion in the cloud. Usage: Put the file inside stable-diffusion-webui\models\VAE. 45 | Upscale x 2. Illuminati Diffusion v1. The model is based on a particular type of diffusion model called Latent Diffusion, which reduces the memory and compute complexity by applying. Stable Diffusion . Paste it into the textbox below the webui script "Prompts from file or textbox". lora weight : 0. So, it is better to make comparison by yourself. I'm just collecting these. From here结合 civitai. 75, Hires upscale: 2, Hires steps: 40, Hires upscaler: Latent (bicubic antialiased) Most of the sample images are generated with hires. You can also upload your own model to the site. Highest Rated. MeinaMix and the other of Meinas will ALWAYS be FREE. No dependencies or technical knowledge needed. This model would not have come out without XpucT's help, which made Deliberate. Stable Diffusion creator Stability AI has announced that users can now test a new generative AI that animates a single image generated from a. Instead, use the "Tiled Diffusion" mode to enlarge the generated image and achieve a more realistic skin texture. Please do mind that I'm not very active on HuggingFace. The output is kind of like stylized rendered anime-ish. stable-diffusion-webui-docker - Easy Docker setup for Stable Diffusion with user-friendly UI. high quality anime style model. ( Maybe some day when Automatic1111 or. This is the fine-tuned Stable Diffusion model trained on images from modern anime feature films from Studio Ghibli. Check out the Quick Start Guide if you are new to Stable Diffusion. 5 and "Juggernaut Aftermath"? I actually announced that I would not release another version. Stable Diffusion in particular is trained competely from scratch which is why it has the most interesting and broard models like the text-to-depth and text-to-upscale models. Dreamlike Photoreal 2. SilasAI6609 ③Civitai | Stable Diffusion 从入门到卸载 【中文教程】 前言. It has a lot of potential and wanted to share it with others to see what others can. Animagine XL is a high-resolution, latent text-to-image diffusion model. 3: Illuminati Diffusion v1. A preview of each frame is generated and outputted to stable-diffusion-webuioutputsmov2mov-images<date> if you interrupt the generation, a video is created with the current progress. REST API Reference. yaml file with name of a model (vector-art. Cetus-Mix. Civitai: Civitai Url. Here is a Form you can request me Lora there (for Free too) As it is model based on 2. For instance: On certain image-sharing sites, many anime character LORAs are overfitted. Note: these versions of the ControlNet models have associated Yaml files which are. 日本人を始めとするアジア系の再現ができるように調整しています。. It is advisable to use additional prompts and negative prompts. Steps and upscale denoise depend on your samplers and upscaler. 0 is another stable diffusion model that is available on Civitai. Universal Prompt Will no longer have update because i switched to Comfy-UI. VAE recommended: sd-vae-ft-mse-original. You can disable this in Notebook settingsBrowse breast Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsBrowse feral Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsOriginally posted to HuggingFace by PublicPrompts. Try adjusting your search or filters to find what you're looking for. このモデルは3D系のマージモデルです。. Browse nsfw Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsCivitai . Use Stable Diffusion img2img to generate the initial background image. Supported parameters. ckpt to use the v1. Try to experiment with the CFG scale, 10 can create some amazing results but to each their own. . Dreamlike Diffusion 1. There are recurring quality prompts. py. Kind of generations: Fantasy. . Get some forest and stone image materials, and composite them in Photoshop, add light, roughly process them into the desired composition and perspective angle. New to AI image generation in the last 24 hours--installed Automatic1111/Stable Diffusion yesterday and don't even know if I'm saying that right. Patreon. i just finetune it with 12GB in 1 hour. MeinaMix and the other of Meinas will ALWAYS be FREE. No longer a merge, but additional training added to supplement some things I feel are missing in current models. Latent upscaler is the best setting for me since it retains or enhances the pastel style. Try adjusting your search or filters to find what you're looking for. I will show you in this Civitai Tutorial how to use civitai Models! Civitai can be used in stable diffusion or Automatic111. SDXLをベースにした複数のモデルをマージしています。. Originally posted to HuggingFace by ArtistsJourney. If you'd like for this to become the official fork let me know and we can circle the wagons here. But you must ensure putting the checkpoint, LoRA, and textual inversion models in the right folders. Fast ~18 steps, 2 seconds images, with Full Workflow Included! No ControlNet, No ADetailer, No LoRAs, No inpainting, No editing, No face restoring, Not Even Hires Fix!! (and obviously no spaghetti nightmare). Raising from the ashes of ArtDiffusionXL-alpha, this is the first anime oriented model I make for the XL architecture. Updated: Dec 30, 2022. Use it with the Stable Diffusion Webui. This model was trained to generate illustration styles! Join our Discord for any questions or feedback!. Leveraging Stable Diffusion 2. This one's goal is to produce a more "realistic" look in the backgrounds and people. Should work well around 8-10 cfg scale and I suggest you don't use the SDXL refiner, but instead do a i2i step on the upscaled. Submit your Part 2 Fusion images here, for a chance to win $5,000 in prizes!It’s GitHub for AI. そのままでも使用に問題はありませんが、Civitaiのデータをより使いやすくしてくれる拡張機能が「Civitai Helper」です。. No animals, objects or backgrounds. I know it's a bit of an old post but I've made an updated fork with a lot of new features which I'll be maintaining and improving! :) Civitai là một nền tảng cho phép người dùng tải xuống và tải lên các hình ảnh do AI Stable Diffusion tạo ra. still requires a. This model is derived from Stable Diffusion XL 1. AnimeIllustDiffusion is a pre-trained, non-commercial and multi-styled anime illustration model. Clip Skip: It was trained on 2, so use 2. To find the Agent Scheduler settings, navigate to the ‘Settings’ tab in your A1111 instance, and scroll down until you see the Agent Scheduler section. To utilize it, you must include the keyword " syberart " at the beginning of your prompt. A repository of models, textual inversions, and more - Home ·. Bad Dream + Unrealistic Dream (Negative Embeddings, make sure to grab BOTH) Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee ☕. The purpose of DreamShaper has always been to make "a better Stable Diffusion", a model capable of doing everything on its own, to weave dreams. This model is available on Mage. Installation: As it is model based on 2. In the hypernetworks folder, create another folder for you subject and name it accordingly. Browse beautiful detailed eyes Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAs. For some workflow examples and see what ComfyUI can do you can check out: ComfyUI Examples Installing ComfyUI FeaturesDownload the TungstenDispo. Settings Overview. Cetus-Mix is a checkpoint merge model, with no clear idea of how many models were merged together to create this checkpoint model. Created by u/-Olorin. Settings Overview. This is a fine-tuned Stable Diffusion model designed for cutting machines. 1168 models. The only thing V5 doesn't do well most of the time are eyes, if you don't get decent eyes try adding perfect eyes or round eyes to the prompt and increase the weight till you are happy. 推荐参数Recommended Parameters for V7: Sampler: Euler a, Euler, restart Steps: 20~40. 5D, so i simply call it 2. The origins of this are unknowniCoMix - Comic style Mix! Thank you for all Reviews, Great Model/Lora Creator, and Prompt Crafter!!! See on Huggingface iCoMix Free Generate iCoMix. . Hello my friends, are you ready for one last ride with Stable Diffusion 1. 6/0. 0 Model character. g. . Am i Real - Photo Realistic Mix Thank you for all Reviews, Great Trained Model/Great Merge Model/Lora Creator, and Prompt Crafter!!! Size: 512x768 or 768x512. The model merge has many costs besides electricity. com) in auto1111 to load the LoRA model. Space (main sponsor) and Smugo. ControlNet will need to be used with a Stable Diffusion model. Trigger words have only been tested using them at the beggining of the prompt. Browse 1. Sensitive Content. Provide more and clearer detail than most of the VAE on the market. The effect isn't quite the tungsten photo effect I was going for, but creates. 1 to make it work you need to use . With SDXL (and, of course, DreamShaper XL 😉) just released, I think the " swiss knife " type of model is closer then ever. Browse pee Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsBrowse toilet Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsWhat Is Stable Diffusion and How It Works. Before delving into the intricacies of After Detailer, let's first understand the traditional approach to addressing problems like distorted faces in images generated using lower-resolution models. 3 on Civitai for download . With SDXL (and, of course, DreamShaper XL 😉) just released, I think the " swiss knife " type of model is closer then ever. Get early access to build and test build, be able to try all epochs and test them by yourself on Patreon or contact me for support on Discord. It took me 2 weeks+ to get the art and crop it. Worse samplers might need more steps. V6. このマージモデルを公開するにあたり、使用したモデルの製作者の皆様に感謝申し. Step 2: Create a Hypernetworks Sub-Folder. The yaml file is included here as well to download. Civitai Helper 2 also has status news, check github for more. For next models, those values could change. This model performs best in the 16:9 aspect ratio, although it can also produce good results in a square format. Characters rendered with the model: Cars and. Civitai stands as the singular model-sharing hub within the AI art generation community. ChatGPT Prompter. Use the tokens ghibli style in your prompts for the effect. Automatic1111. I will show you in this Civitai Tutorial how to use civitai Models! Civitai can be used in stable diffusion or Automatic111. Steps and CFG: It is recommended to use Steps from “20-40” and CFG scale from “6-9”, the ideal is: steps 30, CFG 8. You can use DynamicPrompt Extantion with prompt like: {1-15$$__all__} to get completely random results. One of the model's key strengths lies in its ability to effectively process textual inversions and LORA, providing accurate and detailed outputs. 43 GB) Verified: 10 months ago. It proudly offers a platform that is both free of charge and open source, perpetually advancing to enhance the user experience. It's VAE that, makes every colors lively and it's good for models that create some sort of a mist on a picture, it's good with kotosabbysphoto mode. Although this solution is not perfect. Final Video Render. Works only with people. Browse japanese Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsHere is the Lora for ahegao! The trigger words is ahegao You can also add the following prompt to strengthen the effect: blush, rolling eyes, tongu. Space (main sponsor) and Smugo. Verson2. (avoid using negative embeddings unless absolutely necessary) From this initial point, experiment by adding positive and negative tags and adjusting the settings. Additionally, the model requires minimal prompts, making it incredibly user-friendly and accessible. Vaguely inspired by Gorillaz, FLCL, and Yoji Shin. For the Stable Diffusion community folks that study the near-instant delivery of naked humans on demand, you'll be happy to learn that Uber Realistic Porn Merge has been updated to 1. Sensitive Content. The change in quality is less than 1 percent, and we went from 7 GB to 2 GB. This model is fantastic for discovering your characters, and it was fine-tuned to learn the D&D races that aren't in stock SD. You can customize your coloring pages with intricate details and crisp lines. Usually this is the models/Stable-diffusion one. Try adjusting your search or filters to find what you're looking for. --> (Model-EX N-Embedding) Copy the file in C:Users***DocumentsAIStable-Diffusion automatic. Size: 512x768 or 768x512. This model is very capable of generating anime girls with thick linearts. This model has been republished and its ownership transferred to Civitai with the full permissions of the model creator. That model architecture is big and heavy enough to accomplish that the. Updated - SECO: SECO = Second-stage Engine Cutoff (I watch too many SpaceX launches!!) - am cutting this model off now, and there may be an ICBINP XL release, but will see what happens. Sensitive Content. You can still share your creations with the community. I wanna thank everyone for supporting me so far, and for those that support the creation. The platform currently has 1,700 uploaded models from 250+ creators. Civitai is the go-to place for downloading models. 1 or SD2. The model is the result of various iterations of merge pack combined with. Afterburn seemed to forget to turn the lights up in a lot of renders, so have. Built to produce high quality photos. Title: Train Stable Diffusion Loras with Image Boards: A Comprehensive Tutorial. Even animals and fantasy creatures. Option 1: Direct download. pt file and put in embeddings/. Browse photorealistic Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAs. Civitai is the ultimate hub for. Trang web cũng cung cấp một cộng đồng cho người dùng chia sẻ các hình ảnh của họ và học hỏi về AI Stable Diffusion. After scanning finished, Open SD webui's build-in "Extra Network" tab, to show model cards. Aptly called Stable Video Diffusion, it consists of two AI models (known as SVD and SVD-XT) and is capable of creating clips at a 576 x 1,024 pixel resolution. Stable Diffusion model to create images in Synthwave/outrun style, trained using DreamBooth. Joined Nov 20, 2023. Use between 4. I know it's a bit of an old post but I've made an updated fork with a lot of new features which I'll. The first version I'm uploading is a fp16-pruned with no baked vae, which is less than 2 GB, meaning you can get up to 6 epochs in the same batch on a colab. 特にjapanese doll likenessとの親和性を意識しています。. Note that there is no need to pay attention to any details of the image at this time. 本插件需要最新版SD webui，使用前请更新你的SD webui版本。All of the Civitai models inside Automatic 1111 Stable Diffusion Web UI Python 2,006 MIT 372 70 9 Updated Nov 21, 2023. 2-0. AI Community! | 296291 members. This model works best with the Euler sampler (NOT Euler_a). It has been trained using Stable Diffusion 2. Browse weapons Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsStable Difussion Web UIを使っている方は、Civitaiからモデルをダウンロードして利用している方が多いと思います。. This model is a 3D merge model. Inspired by Fictiverse's PaperCut model and txt2vector script. You can use some trigger words (see Appendix A) to generate specific styles of images. . Given the broad range of concepts encompassed in WD 1. Ryokan have existed since the eighth century A. Extract the zip file. 0, but you can increase or decrease depending on desired effect,. Model Description: This is a model that can be used to generate and modify images based on text prompts. It proudly offers a platform that is both free of charge and open source. NED) This is a dream that you will never want to wake up from. 3 here: RPG User Guide v4. See example picture for prompt. Keep in mind that some adjustments to the prompt have been made and are necessary to make certain models work. Originally uploaded to HuggingFace by NitrosockeBrowse civitai Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsThey can be used alone or in combination and will give an special mood (or mix) to the image. 5 using +124000 images, 12400 steps, 4 epochs +3. Life Like Diffusion V2: This model’s a pro at creating lifelike images of people. D. Experience - Experience v10 | Stable Diffusion Checkpoint | Civitai. Use the same prompts as you would for SD 1. . . In your stable-diffusion-webui folder, create a sub-folder called hypernetworks. This is just a merge of the following two checkpoints. This is the fine-tuned Stable Diffusion model trained on screenshots from a popular animation studio. LORA: For anime character LORA, the ideal weight is 1. Anime Style Mergemodel All sample images using highrexfix + ddetailer Put the upscaler in the your "ESRGAN" folder ddetailer 4x-UltraSharp. Verson2. Simply copy paste to the same folder as selected model file. Scans all models to download model information and preview images from Civitai. Copy image prompt and setting in a format that can be read by Prompts from file or textbox. Current list of available settings: Disable queue auto-processing → Checking this option prevents the queue from executing automatically when you start up A1111. How to use models. Use the negative prompt: "grid" to improve some maps, or use the gridless version. Trigger words have only been tested using them at the beggining of the prompt. Click Generate, give it a few seconds, and congratulations, you have generated your first image using Stable Diffusion! (you can track the progress of the image generation under the Run Stable Diffusion cell at the bottom of the collab notebook as well!) Click on the image, and you can right-click save it. :) Last but not least, I'd like to thank a few people without whom Juggernaut XL probably wouldn't have come to fruition: ThinkDiffusion. Link local model to a civitai model by civitai model's urlCherry Picker XL. Try adjusting your search or filters to find what you're looking for. 9). May it be through trigger words, or prompt adjustments between. Side by side comparison with the original. I found that training from the photorealistic model gave results closer to what I wanted than the anime model. Through this process, I hope not only to gain a deeper. 起名废玩烂梗系列，事后想想起的不错。. Explore thousands of high-quality Stable Diffusion models, share your AI-generated art, and engage with a vibrant community of creators Browse from thousands of free Stable Diffusion models, spanning unique anime art styles, immersive 3D renders, stunning photorealism, and more. Silhouette/Cricut style. 9. Improves details, like faces and hands. Trigger word: 2d dnd battlemap. Insutrctions. If you have your Stable Diffusion. Civitai is a great place to hunt for all sorts of stable diffusion models trained by the community. Since its debut, it has been a fan favorite of many creators and developers working with stable diffusion. Highres fix with either a general upscaler and low denoise or Latent with high denoise (see examples) Be sure to use Auto as vae for baked vae versions and a good vae for the no vae ones. pruned. Civitai. if you like my stuff consider supporting me on Kofi Bad Dream + Unrealistic Dream (Negative Embeddings, make sure to grab BOTH) Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free. This tutorial is a detailed explanation of a workflow, mainly about how to use Stable Diffusion for image generation, image fusion, adding details, and upscaling. You can now run this model on RandomSeed and SinkIn . 1 model from civitai. We would like to thank the creators of the models we used. . Im currently preparing and collecting dataset for SDXL, Its gonna be huge and a monumental task. Try it out here! Join the discord for updates, share generated-images, just want to chat or if you want to contribute to helpin. The yaml file is included here as well to download. 5 model. 0 is based on new and improved training and mixing. 5 models available, check the blue tabs above the images up top: Stable Diffusion 1. 0. Browse textual inversion Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and. This notebook is open with private outputs. Create stunning and unique coloring pages with the Coloring Page Diffusion model! Designed for artists and enthusiasts alike, this easy-to-use model generates high-quality coloring pages from any text prompt. a. I'm happy to take pull requests. This extension allows you to seamlessly manage and interact with your Automatic 1111 SD instance directly from Civitai. Add export_model_dir option to specify the directory where the model is exported. 1 and V6. Comfyui need use. high quality anime style model. Enter our Style Capture & Fusion Contest! Join Part 1 of our two-part Style Capture & Fusion Contest! Running NOW until November 3rd, train and submit any artist's style as a LoRA for a chance to win $5,000 in prizes! Read the rules on how to enter here! Babes 2. It tends to lean a bit towards BoTW, but it's very flexible and allows for most Zelda versions. . Browse controlnet Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAs If you liked the model, please leave a review. trigger word : gigachad Lora strength closer to 1 will give the ultimate gigachad, for more flexibility consider lowering the value. SDXL-Anime, XL model for replacing NAI. FFUSION AI is a state-of-the-art image generation and transformation tool, developed around the leading Latent Diffusion Model. Browse snake Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsSynthwavePunk - V2 | Stable Diffusion Checkpoint | Civitai. Hires upscaler: ESRGAN 4x or 4x-UltraSharp or 8x_NMKD-Superscale_150000_G Hires upscale: 2+ Hires steps: 15+This is a fine-tuned Stable Diffusion model (based on v1. Use ninja to build xformers much faster ( Followed by Official README) stable_diffusion_1_5_webui. Sensitive Content. " (mostly for v1 examples) Browse chibi Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAs CivitAI: list: This is DynaVision, a new merge based off a private model mix I've been using for the past few months. Type. art. pixelart: The most generic one. All models, including Realistic Vision (VAE. ckpt file but since this is a checkpoint I'm still not sure if this should be loaded as a standalone model or a new. To find the Agent Scheduler settings, navigate to the ‘Settings’ tab in your A1111 instance, and scroll down until you see the Agent Scheduler section. v1 update: 1. Browse kiss Stable Diffusion models, checkpoints, hypernetworks, textual inversions, embeddings, Aesthetic Gradients, and LORAsOriginal Model Dpepteahand3. Step 2: Background drawing. Resources for more information: GitHub. And it contains enough information to cover various usage scenarios. Stylized RPG game icons. 5. pixelart-soft: The softer version of an. Western Comic book styles are almost non existent on Stable Diffusion. Add dreamlikeart if the artstyle is too weak. 50+ Pre-Loaded Models. 0 (B1) Status (Updated: Nov 18, 2023): - Training Images: +2620 - Training Steps: +524k - Approximate percentage of completion: ~65%. Downloading a Lycoris model. Civitai is a website where you can browse and download lots of Stable Diffusion models and embeddings. CoffeeNSFW Maier edited this page Dec 2, 2022 · 3 revisions. Remember to use a good vae when generating, or images wil look desaturated.