UmeAiRT

This project provides a suite of PowerShell scripts to fully automate the installation and configuration of ComfyUI on Windows. The approach uses a clean installation based on `git` and a Python virtual environment (`venv`), ensuring an isolated, easy-to-update, and maintainable setup. - Clean Installation: Clones the latest version of ComfyUI from the official repository and installs it in a dedicated Anaconda Python virtual environment. - Dependency Management: Automatically checks for and installs necessary tools: - Anaconda Python 3.12 (if not present on the system) - Git - 7-Zip - Aria2 (for accelerated downloads) - CSV-Managed Custom Nodes: Installs a comprehensive list of custom nodes defined in an external `customnodes.csv` file, making it simple to add new nodes. - Interactive Model Downloaders: Dedicated scripts guide you with menus to download the model packs you want (FLUX, WAN, HIDREAM, LTXV), with recommendations based on your graphics card's VRAM. - Dedicated Update Script: A specific `UmeAiRT-Updater.ps1` script allows you to update ComfyUI, all custom nodes, and workflows with a single command. - Automated Launchers: The project generates `.bat` files to run the installation, updates, and the final application, automatically handling administrator rights and PowerShell execution policies. - Supplementary modules: The script also installs some complex modules such as: Sageattention, Triton, Visual Studio Build Tools, ... - Workflow included: A large amount of workflows are pre-installed for each model. - Windows 10 or Windows 11 (64-bit). - An active internet connection. - An NVIDIA GPU is strongly recommended to use the models. The entire process is designed to be as simple as possible. 1. Download the Project: Download `UmeAiRT-Install-ComfyUI.bat` and put it to a folder of your choice (e.g., `C:\UmeAiRT-Installer`). 2. Run the Installer: - Run the file `UmeAiRT-Install-ComfyUI.bat`. - It will ask for administrator privileges. Please accept. - The script will first download the latest versions of all installation scripts from the repository to ensure you are using the most recent version. 3. Follow the Instructions: - The main installation script will then launch. It will install Python (if necessary), Git, 7-Zip, Aria2, and then ComfyUI. - Next, it will install all custom nodes and their Python dependencies into the virtual environment. - Finally, it will ask you a series of questions about which model packs you wish to download. Simply answer `Y` (yes) or `N` (no) to each question. At the end of the process, your ComfyUI installation will be complete and ready to use. Three main `.bat` files will be available in your folder to manage the application: - `UmeAiRT-Start-ComfyUI.bat` - This is the file you will use to launch ComfyUI. It activates the virtual environment and starts the server. - `UmeAiRT-Downloadmodels.bat` - Run this script if you want to add more model packs later without reinstalling everything. It will present you with the same selection menu as the initial installation. - `UmeAiRT-Update-ComfyUI.bat` - Execute this script to update your entire installation. It will update the code for ComfyUI, all custom nodes, and your workflows, and it will install any new Python dependencies if required. - `/` (your root folder) - `UmeAiRT-Installer-Updater.bat` (Main launcher that updates and installs) - `UmeAiRT-Start-ComfyUI.bat` (Created after installation to launch ComfyUI) - `UmeAiRT-Update-ComfyUI.bat` (Launcher for the update script) - `UmeAiRT-Downloadmodels.bat` (Menu to download more models later) - `scripts/` (Contains all PowerShell scripts) - `Install-ComfyUI.ps1` - `UmeAiRT-Updater.ps1` - `Download-FLUX-Models.ps1` (and other model downloaders) - `customnodes.csv` (The list of all custom nodes to install) - `ComfyUI/` (Created after installation, contains the application) - `logs/` (Created, contains installation/update logs) Suggestions and contributions are welcome. If you find a bug or have an idea for an improvement to the scripts, feel free to open an "Issue" on this GitHub repository. This project is under the MIT License. See the `LICENSE` file for more details. - To Comfyanonymous for creating the incredible ComfyUI. - To the authors of all the custom nodes that enrich the ecosystem.

This time I used more images for creation (30). I tried to give it a “painted” look that I really like. It was created from images I myself published on Deviant-Art : https://www.deviantart.com/umeairt/gallery/all Using too large a weight can result in poor image quality. Don't hesitate to send me feedback or suggestions! You should use `umesky` to trigger the image generation. Weights for this model are available in Safetensors format.

Classic series - Impressionism Hello this in my third LoRA. This time, I've tried to reproduce an important period of painting. Although Flux is capable of making this kind of image it remains very light and this Lora will allow you to give this unique touch of impressionism. Don't hesitate to send me feedback or suggestions! You should use `impressionist` to trigger the image generation. Weights for this model are available in Safetensors format.

Based on games I've personally really enjoyed from the modern 2d pixel art category. This is the first time I've worked with such a large dataset (100), so I'm not sure my settings are optimal. Don't hesitate to send me feedback or suggestions! You should use `umempart` to trigger the image generation. Weights for this model are available in Safetensors format.