[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-71046":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":16,"stars7d":16,"stars30d":17,"stars90d":16,"forks30d":16,"starsTrendScore":16,"compositeScore":18,"rankGlobal":10,"rankLanguage":10,"license":19,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":22,"topics":23,"createdAt":10,"pushedAt":10,"updatedAt":24,"readmeContent":25,"aiSummary":26,"trendingCount":16,"starSnapshotCount":16,"syncStatus":27,"lastSyncTime":28,"discoverSource":29},71046,"sygil-webui","Sygil-Dev\u002Fsygil-webui","Sygil-Dev","Stable Diffusion web UI","",null,"Python",7882,864,70,77,0,1,64.91,"GNU Affero General Public License v3.0",false,"master",true,[],"2026-06-12 04:00:58","# \u003Ccenter>Web-based UI for Stable Diffusion\u003C\u002Fcenter>\r\n\r\n## Created by [Sygil.Dev](https:\u002F\u002Fgithub.com\u002Fsygil-dev)\r\n\r\n## Join us at Sygil.Dev's Discord Server [![Generic badge](https:\u002F\u002Fflat.badgen.net\u002Fdiscord\u002Fmembers\u002FttM8Tm6wge?icon=discord)](https:\u002F\u002Fdiscord.gg\u002FttM8Tm6wge)\r\n\r\n## Installation instructions for:\r\n\r\n- **[Windows](https:\u002F\u002Fsygil-dev.github.io\u002Fsygil-webui\u002Fdocs\u002FInstallation\u002Fwindows-installation)**\r\n- **[Linux](https:\u002F\u002Fsygil-dev.github.io\u002Fsygil-webui\u002Fdocs\u002FInstallation\u002Flinux-installation)**\r\n\r\n### Want to ask a question or request a feature?\r\n\r\nCome to our [Discord Server](https:\u002F\u002Fdiscord.gg\u002FgyXNe4NySY) or use [Discussions](https:\u002F\u002Fgithub.com\u002Fsygil-dev\u002Fsygil-webui\u002Fdiscussions).\r\n\r\n## Documentation\r\n\r\n[Documentation is located here](https:\u002F\u002Fsygil-dev.github.io\u002Fsygil-webui\u002F)\r\n\r\n## Want to contribute?\r\n\r\nCheck the [Contribution Guide](CONTRIBUTING.md)\r\n\r\n[Sygil-Dev](https:\u002F\u002Fgithub.com\u002FSygil-Dev) main devs:\r\n\r\n* ![ZeroCool940711's avatar](https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F5977640?s=40&v=4)[ZeroCool940711](https:\u002F\u002Fgithub.com\u002FZeroCool940711)\r\n* ![Kasiya13's avatar](https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F26075839?s=40&v=4)[Kasiya13](https:\u002F\u002Fgithub.com\u002FKasiya13)\r\n\r\n### Project Features:\r\n\r\n* Built-in image enhancers and upscalers, including GFPGAN and realESRGAN\r\n\r\n* Generator Preview: See your image as its being made\r\n\r\n* Run additional upscaling models on CPU to save VRAM\r\n\r\n* Textual inversion: [Reaserch Paper](https:\u002F\u002Ftextual-inversion.github.io\u002F)\r\n\r\n* K-Diffusion Samplers: A great collection of samplers to use, including:\r\n\r\n  - `k_euler`\r\n  - `k_lms`\r\n  - `k_euler_a`\r\n  - `k_dpm_2`\r\n  - `k_dpm_2_a`\r\n  - `k_heun`\r\n  - `PLMS`\r\n  - `DDIM`\r\n\r\n* Loopback: Automatically feed the last generated sample back into img2img\r\n\r\n* Prompt Weighting & Negative Prompts: Gain more control over your creations\r\n\r\n* Selectable GPU usage from Settings tab\r\n\r\n* Word Seeds: Use words instead of seed numbers\r\n\r\n* Automated Launcher: Activate conda and run Stable Diffusion with a single command\r\n\r\n* Lighter on VRAM: 512x512 Text2Image & Image2Image tested working on 4GB (with *optimized* mode enabled in Settings)\r\n\r\n* Prompt validation: If your prompt is too long, you will get a warning in the text output field\r\n\r\n* Sequential seeds for batches: If you use a seed of 1000 to generate two batches of two images each, four generated images will have seeds: `1000, 1001, 1002, 1003`.\r\n\r\n* Prompt matrix: Separate multiple prompts using the `|` character, and the system will produce an image for every combination of them.\r\n\r\n* [Gradio] Advanced img2img editor with Mask and crop capabilities\r\n\r\n* [Gradio] Mask painting 🖌️: Powerful tool for re-generating only specific parts of an image you want to change (currently Gradio only)\r\n\r\n# SD WebUI\r\n\r\nAn easy way to work with Stable Diffusion right from your browser.\r\n\r\n## Streamlit\r\n\r\n![](images\u002Fstreamlit\u002Fstreamlit-t2i.png)\r\n\r\n**Features:**\r\n\r\n- Clean UI with an easy to use design, with support for widescreen displays\r\n- *Dynamic live preview* of your generations\r\n- Easily customizable defaults, right from the WebUI's Settings tab\r\n- An integrated gallery to show the generations for a prompt\r\n- *Optimized VRAM* usage for bigger generations or usage on lower end GPUs\r\n- *Text to Video:* Generate video clips from text prompts right from the WebUI (WIP)\r\n- Image to Text: Use [CLIP Interrogator](https:\u002F\u002Fgithub.com\u002Fpharmapsychotic\u002Fclip-interrogator) to interrogate an image and get a prompt that you can use to generate a similar image using Stable Diffusion.\r\n- *Concepts Library:* Run custom embeddings others have made via textual inversion.\r\n- Textual Inversion training: Train your own embeddings on any photo you want and use it on your prompt.\r\n- **Currently in development: [Stable Horde](https:\u002F\u002Fstablehorde.net\u002F) integration; ImgLab, batch inputs, & mask editor from Gradio\r\n\r\n**Prompt Weights & Negative Prompts:**\r\n\r\nTo give a token (tag recognized by the AI) a specific or increased weight (emphasis), add `:0.##` to the prompt, where `0.##` is a decimal that will specify the weight of all tokens before the colon.\r\nEx: `cat:0.30, dog:0.70` or `guy riding a bicycle :0.7, incoming car :0.30`\r\n\r\nNegative prompts can be added by using  `###` , after which any tokens will be seen as negative.\r\nEx: `cat playing with string ### yarn` will negate `yarn` from the generated image.\r\n\r\nNegatives are a very powerful tool to get rid of contextually similar or related topics, but **be careful when adding them since the AI might see connections you can't**, and end up outputting gibberish\r\n\r\n**Tip:* Try using the same seed with different prompt configurations or weight values see how the AI understands them, it can lead to prompts that are more well-tuned and less prone to error.\r\n\r\nPlease see the [Streamlit Documentation](docs\u002F4.streamlit-interface.md) to learn more.\r\n\r\n## Gradio [Legacy]\r\n\r\n![](images\u002Fgradio\u002Fgradio-t2i.png)\r\n\r\n**Features:**\r\n\r\n- Older UI that is functional and feature complete.\r\n- Has access to all upscaling models, including LSDR.\r\n- Dynamic prompt entry automatically changes your generation settings based on `--params` in a prompt.\r\n- Includes quick and easy ways to send generations to Image2Image or the Image Lab for upscaling.\r\n\r\n**Note: the Gradio interface is no longer being actively developed by Sygil.Dev and is only receiving bug fixes.**\r\n\r\nPlease see the [Gradio Documentation](https:\u002F\u002Fsygil-dev.github.io\u002Fsygil-webui\u002Fdocs\u002FGradio\u002Fgradio-interface\u002F) to learn more.\r\n\r\n## Image Upscalers\r\n\r\n---\r\n\r\n### GFPGAN\r\n\r\n![](images\u002FGFPGAN.png)\r\n\r\nLets you improve faces in pictures using the GFPGAN model. There is a checkbox in every tab to use GFPGAN at 100%, and also a separate tab that just allows you to use GFPGAN on any picture, with a slider that controls how strong the effect is.\r\n\r\nIf you want to use GFPGAN to improve generated faces, you need to install it separately.\r\nDownload [GFPGANv1.4.pth](https:\u002F\u002Fgithub.com\u002FTencentARC\u002FGFPGAN\u002Freleases\u002Fdownload\u002Fv1.3.4\u002FGFPGANv1.4.pth) and put it\r\ninto the `\u002Fsygil-webui\u002Fmodels\u002Fgfpgan` directory.\r\n\r\n### RealESRGAN\r\n\r\n![](images\u002FRealESRGAN.png)\r\n\r\nLets you double the resolution of generated images. There is a checkbox in every tab to use RealESRGAN, and you can choose between the regular upscaler and the anime version.\r\nThere is also a separate tab for using RealESRGAN on any picture.\r\n\r\nDownload [RealESRGAN_x4plus.pth](https:\u002F\u002Fgithub.com\u002Fxinntao\u002FReal-ESRGAN\u002Freleases\u002Fdownload\u002Fv0.1.0\u002FRealESRGAN_x4plus.pth) and [RealESRGAN_x4plus_anime_6B.pth](https:\u002F\u002Fgithub.com\u002Fxinntao\u002FReal-ESRGAN\u002Freleases\u002Fdownload\u002Fv0.2.2.4\u002FRealESRGAN_x4plus_anime_6B.pth).\r\nPut them into the `sygil-webui\u002Fmodels\u002Frealesrgan` directory.\r\n\r\n### LSDR\r\n\r\nDownload **LDSR** [project.yaml](https:\u002F\u002Fheibox.uni-heidelberg.de\u002Ff\u002F31a76b13ea27482981b4\u002F?dl=1) and [model last.cpkt](https:\u002F\u002Fheibox.uni-heidelberg.de\u002Ff\u002F578df07c8fc04ffbadf3\u002F?dl=1). Rename `last.ckpt` to `model.ckpt` and place both under `sygil-webui\u002Fmodels\u002Fldsr\u002F`\r\n\r\n### GoBig, and GoLatent *(Currently on the Gradio version Only)*\r\n\r\nMore powerful upscalers that uses a separate Latent Diffusion model to more cleanly upscale images.\r\n\r\nPlease see the [Post-Processing Documentation](https:\u002F\u002Fsygil-dev.github.io\u002Fsygil-webui\u002Fdocs\u002Fpost-processing) to learn more.\r\n\r\n-----\r\n\r\n### *Original Information From The Stable Diffusion Repo:*\r\n\r\n# Stable Diffusion\r\n\r\n*Stable Diffusion was made possible thanks to a collaboration with [Stability AI](https:\u002F\u002Fstability.ai\u002F) and [Runway](https:\u002F\u002Frunwayml.com\u002F) and builds upon our previous work:*\r\n\r\n[**High-Resolution Image Synthesis with Latent Diffusion Models**](https:\u002F\u002Fommer-lab.com\u002Fresearch\u002Flatent-diffusion-models\u002F)\r\n[Robin Rombach](https:\u002F\u002Fgithub.com\u002Frromb)\\*,\r\n[Andreas Blattmann](https:\u002F\u002Fgithub.com\u002Fablattmann)\\*,\r\n[Dominik Lorenz](https:\u002F\u002Fgithub.com\u002Fqp-qp)\\,\r\n[Patrick Esser](https:\u002F\u002Fgithub.com\u002Fpesser),\r\n[Björn Ommer](https:\u002F\u002Fhci.iwr.uni-heidelberg.de\u002FStaff\u002Fbommer)\r\n\r\n**CVPR '22 Oral**\r\n\r\nwhich is available on [GitHub](https:\u002F\u002Fgithub.com\u002FCompVis\u002Flatent-diffusion). PDF at [arXiv](https:\u002F\u002Farxiv.org\u002Fabs\u002F2112.10752). Please also visit our [Project page](https:\u002F\u002Fommer-lab.com\u002Fresearch\u002Flatent-diffusion-models\u002F).\r\n\r\n[Stable Diffusion](#stable-diffusion-v1) is a latent text-to-image diffusion\r\nmodel.\r\nThanks to a generous compute donation from [Stability AI](https:\u002F\u002Fstability.ai\u002F) and support from [LAION](https:\u002F\u002Flaion.ai\u002F), we were able to train a Latent Diffusion Model on 512x512 images from a subset of the [LAION-5B](https:\u002F\u002Flaion.ai\u002Fblog\u002Flaion-5b\u002F) database.\r\nSimilar to Google's [Imagen](https:\u002F\u002Farxiv.org\u002Fabs\u002F2205.11487),\r\nthis model uses a frozen CLIP ViT-L\u002F14 text encoder to condition the model on text prompts.\r\nWith its 860M UNet and 123M text encoder, the model is relatively lightweight and runs on a GPU with at least 10GB VRAM.\r\nSee [this section](#stable-diffusion-v1) below and the [model card](https:\u002F\u002Fhuggingface.co\u002FCompVis\u002Fstable-diffusion).\r\n\r\n## Stable Diffusion v1\r\n\r\nStable Diffusion v1 refers to a specific configuration of the model\r\narchitecture that uses a downsampling-factor 8 autoencoder with an 860M UNet\r\nand CLIP ViT-L\u002F14 text encoder for the diffusion model. The model was pretrained on 256x256 images and\r\nthen finetuned on 512x512 images.\r\n\r\n*Note: Stable Diffusion v1 is a general text-to-image diffusion model and therefore mirrors biases and (mis-)conceptions that are present\r\nin its training data.\r\nDetails on the training procedure and data, as well as the intended use of the model can be found in the corresponding [model card](https:\u002F\u002Fhuggingface.co\u002FCompVis\u002Fstable-diffusion).\r\n\r\n## Comments\r\n\r\n- Our code base for the diffusion models builds heavily on [OpenAI's ADM codebase](https:\u002F\u002Fgithub.com\u002Fopenai\u002Fguided-diffusion)\r\n  and [https:\u002F\u002Fgithub.com\u002Flucidrains\u002Fdenoising-diffusion-pytorch](https:\u002F\u002Fgithub.com\u002Flucidrains\u002Fdenoising-diffusion-pytorch).\r\n  Thanks for open-sourcing!\r\n\r\n- The implementation of the transformer encoder is from [x-transformers](https:\u002F\u002Fgithub.com\u002Flucidrains\u002Fx-transformers) by [lucidrains](https:\u002F\u002Fgithub.com\u002Flucidrains?tab=repositories).\r\n\r\n## BibTeX\r\n\r\n```\r\n@misc{rombach2021highresolution,\r\n      title={High-Resolution Image Synthesis with Latent Diffusion Models},\r\n      author={Robin Rombach and Andreas Blattmann and Dominik Lorenz and Patrick Esser and Björn Ommer},\r\n      year={2021},\r\n      eprint={2112.10752},\r\n      archivePrefix={arXiv},\r\n      primaryClass={cs.CV}\r\n}\r\n```\r\n","Sygil-WebUI 是一个基于浏览器的用户界面，用于操作Stable Diffusion模型生成图像。它使用Python开发，集成了多种先进的图像增强和放大工具如GFPGAN和realESRGAN，并支持多种采样器（例如k_euler、k_lms等），以提供更高质量的图像输出。此外，该项目还提供了文本反转、循环反馈等功能来增加创作灵活性，并且通过优化模式在有限显存条件下也能运行。适用于需要快速原型设计或实验性艺术创作的个人开发者及小型团队，在不需要深入编程知识的情况下探索AI生成图像的可能性。",2,"2026-06-11 03:35:37","high_star"]