{"id":21220856,"url":"https://github.com/rupeshs/diffusionmagic","last_synced_at":"2025-07-10T12:31:24.264Z","repository":{"id":67375193,"uuid":"597796131","full_name":"rupeshs/diffusionmagic","owner":"rupeshs","description":"Easy to use Stable diffusion workflows using diffusers ","archived":false,"fork":false,"pushed_at":"2023-10-31T18:13:31.000Z","size":23453,"stargazers_count":114,"open_issues_count":3,"forks_count":14,"subscribers_count":7,"default_branch":"main","last_synced_at":"2024-05-02T01:38:41.252Z","etag":null,"topics":["artificial-intelligence","deep-learning","diffusers","huggingface","illusiondiffusion","python","pytorch","stable-diffusion","stablediffuion","stablediffusionxl","texttoimage","windows","wuerstchen"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/rupeshs.png","metadata":{"files":{"readme":"Readme.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2023-02-05T17:04:47.000Z","updated_at":"2024-04-28T06:17:30.000Z","dependencies_parsed_at":"2023-05-26T13:45:29.066Z","dependency_job_id":"cde2d56e-82c1-4c44-b4c0-c4d45f076587","html_url":"https://github.com/rupeshs/diffusionmagic","commit_stats":null,"previous_names":[],"tags_count":4,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rupeshs%2Fdiffusionmagic","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rupeshs%2Fdiffusionmagic/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rupeshs%2Fdiffusionmagic/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/rupeshs%2Fdiffusionmagic/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/rupeshs","download_url":"https://codeload.github.com/rupeshs/diffusionmagic/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":225637111,"owners_count":17500365,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["artificial-intelligence","deep-learning","diffusers","huggingface","illusiondiffusion","python","pytorch","stable-diffusion","stablediffuion","stablediffusionxl","texttoimage","windows","wuerstchen"],"created_at":"2024-11-20T22:14:21.165Z","updated_at":"2024-11-20T22:14:21.638Z","avatar_url":"https://github.com/rupeshs.png","language":"Python","readme":"## DiffusionMagic \nDiffusionMagic is simple to use Stable Diffusion workflows using [diffusers](https://github.com/huggingface/diffusers). \nDiffusionMagic focused on the following areas:\n- Easy to use\n- Cross-platform (Windows/Linux/Mac)\n- Modular design, latest best optimizations for speed and memory\n\n ![ DiffusionMagic SDXL support](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic_3_sdxl.png)\n\n## Segmind Stable Diffusion 1B (SSD-1B) \nThe Segmind Stable Diffusion Model (SSD-1B) is a smaller version of the Stable Diffusion XL (SDXL) that is 50% smaller but maintains high-quality text-to-image generation. It offers a 60% speedup compared to SDXL.\n\nYou can run SSD-1B on Google Colab\n\n[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1z-3upaBQCqBw-QNeFYdNsLxbNE_pw1Rx?usp=sharing)\n\n## Stable diffusion XL Colab\nYou can run StableDiffusion XL 1.0 on Google Colab\n\n[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1eEZ_O-Fw87hoEsfSxUnGZhdqvMFEO5iV?usp=sharing)\n\n## Würstchen Colab\nYou can run Würstchen 2.0 on Google Colab\n\n[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1ib6W1CeK9V533Nc9MnoBe3TmU7Uaghtg?usp=sharing)\n\n\n## Illusion Diffusion Colab (beta)\nYou can run  Illusion Diffusion on Google Colab\n\n[![Open in Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/drive/1M3igyVklKkUh1Pgzy68JWms2KQy2z7s7?usp=sharing)\n\nIllusion diffusion supports following types of input images as illusion control :\n - Color images\n - Text images\n - Patterns\n\n  You need to adjust the illusion strength to get desired result.\n\n  ![  Color image](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusionmagic-illusion-diffusion-color-image.jpg)\n\n   ![ Text image](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusionmagic-illusion-diffusion-text.jpg)\n\n   ![ Patterns](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusionmagic-illusion-diffusion-pattern.jpg)\n\n## Features\n- Supports Würstchen\n- Supports Stable diffusion XL\n- Supports various Stable Diffusion workflows\n    - Text to Image \n    - Image to Image \n    - Image variations\n    - Image Inpainting\n    - Depth to Image\n    - Instruction based image editing\n- Supports Controlnet workflows\n   - Canny\n   - MLSD (Line control)\n   -  Normal\n   - HED\n   - Pose\n   - Depth\n   - Scribble\n   - Segmentation\n- Pytorch 2.0 support\n- Supports all stable diffusion Hugging Face models \n- Supports Stable diffusion v1 and v2 models, derived models\n- Works on Windows/Linux/Mac 64-bit\n- Works on CPU,GPU(Recent Nvidia GPU),Apple Silicon M1/M2 hardware\n- Supports DEIS scheduler for faster image generation (10 steps)\n- Supports 7 different samplers with latest DEIS sampler\n- LoRA(Low-Rank Adaptation of Large Language Models) models support (~3 MB size)\n- Easy to add new diffuser model by updating stable_diffusion_models.txt \n- Low VRAM mode supports GPU with RAM \u003c 4 GB \n- Fast model loading\n- Supports Attention slicing and VAE slicing\n- Simple installation using install.bat/install.sh\n\nPlease note that AMD GPUs are not supported.\n## Screenshots\n### Image variations\n ![  Image variations SDXL](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic_image_variations_sdxl.PNG)\n\n### Image Inpainting\n ![ Image Inpainting](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic_inpainting.PNG)\n### Depth to Image\n ![ Depth To Image](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic_depth_image.PNG)\n ### Instruction based image editing\n ![ Depth To Image](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic_instruct_to_pix.PNG\n)\n## System Requirements:\n- Works on Windows/Linux/Mac 64-bit\n- Works on CPU,GPU,Apple Silicon M1/M2 hardware\n- 12 GB System RAM\n- ~11 GB disk space after installation (on SSD for best performance)\n\n## Low VRAM mode \u003c 4GB\nDiffusionMagic runs on low VRAM GPUs.\n[Here](https://nolowiz.com/easy-way-to-run-stable-diffusion-xl-on-low-vram-gpus/) is our guide to run StableDiffusion XL on low VRAM GPUs.\n\n## Download Release\nDownload release from the github DiffusionMagic [releases](https://github.com/rupeshs/diffusionmagic/releases/).\n## How to install and run on Windows\nFollow the steps to install and run the Diffusion magic on Windows.\n- First we need to run(double click) the `install.bat` batch file it will install the necessary dependencies for DiffusionMagic.\n(It will take some time to install,depends on your internet speed)\n- Run the  `install.bat` script.\n- To start DiffusionMagic double click `start.bat`\n\n\n ![ DiffusionMagic started on Windows](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic_windows.PNG)\n## How to install and run on Linux\nFollow the steps to install and run the Diffusion magic on Linux.\n\n - Run the following command:\n  `chmod +x install.sh`\n- Run the  `install.sh` script.\n ` ./install.sh`\n- To start DiffusionMagic run:\n` ./start.sh`\n\n## How to install and run on Mac (Not tested)\n*Testers needed - If you have MacOS feel free to test and contribute*\n\n#### prerequisites \n- Mac computer with Apple silicon (M1/M2) hardware.\n- macOS 12.6 or later (13.0 or later recommended).\n\nFollow the steps to install and run the Diffusion magic on Mac (Apple Silicon M1/M2).\n - Run the following command:\n  `chmod +x install-mac.sh`\n- Run the  `install-mac.sh` script.\n`./install-mac.sh`\n- To start DiffusionMagic run:\n` ./start.sh`\n\n Open the browser `http://localhost:7860/`\n##  Dark mode \nTo get dark theme :\n\n `http://localhost:7860/?__theme=dark`\n\n## How to switch models\nDiffusion magic will change UI based on the model selected.\nFollow the steps to switch the models() inpainting,depth to image or instruct pix to pix or any other hugging face stable diffusion model)\n- Start the Diffusion Magic app, open the settings tab and change the model\n ![ DiffusionMagic settings page](https://raw.githubusercontent.com/rupeshs/diffusionmagic/main/docs/images/diffusion_magic%20setting.PNG)\n- Save the settings\n- Close the app and start using start.bat/start.sh\n ## How to add new model\nWe can add any Hugging Face stable diffusion model to DiffusionMagic by \n- Adding Hugging Face models  id or local folder path to the configs/stable_diffusion_models.txt file\nE.g `https://huggingface.co/dreamlike-art/dreamlike-diffusion-1.0`\nHere model id is `dreamlike-art/dreamlike-diffusion-1.0`\nOr we can clone the model use the local folder path as model id.\n- Adding locally copied model path to configs/stable_diffusion_models.txt file\n## Linting (Development)\nRun the following commands from src folder\n\n`mypy --ignore-missing-imports --explicit-package-bases .`\n\n`flake8 --max-line-length=100 .`\n## Contribute\nContributions are welcomed.\n\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frupeshs%2Fdiffusionmagic","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Frupeshs%2Fdiffusionmagic","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Frupeshs%2Fdiffusionmagic/lists"}