WeShop-UI
Welcome to Weshop_UI_1.0.0! This software is an open-source solution derived from WeShop.AI, which specialize in generating images using AI. We've extracted some business/professional functionalities, removed proprietary elements, and are gradually releasing parts of our services to the public. Our goal is to foster innovation and collaboration in AI-powered image creation while maintaining our core business offerings. WeShop UI features multi-task management, and a simple and user-friendly interface, perfect for beginners.
Introduction
Weshop_UI_1.0.0 is an open-source solution that brings some functionalities of weshop.ai to your local operation. Weshop_UI_1.0.0 leverages Stable Diffusion Webui as its core inference engine, enabling you to deploy WeShop.AI functionalities locally and generate high-quality AI images without internet or cloud reliance.
News
- 2024.07 We officially launched WeShop UI
Open-source Plan
Currently, we've open sourced web and Electron frontend, changes made for Automatic1111 Stable Diffusion Webui. Our plan regarding key functionalities is listed below:
Keep upgrading Webui with the latest one on https://github.com/AUTOMATIC1111/stable-diffusion-webui
Open source the front end part of this project
Refactor and open source the backend part of this project. This project uses compiled executables currently.
Support ComfyUI simultaneously.
Support multiple-GPU inference scheduling.
Gradually adapt workflows/features from WeShop.AI
More functions to come and feel free to leave you feedbacks!
Features
Seamlessly integrated SD WebUI functionalities, supporting SDXL model and SDXL's Controlnet
Enable shared graphics card utilization for enhanced performance
Harness the power of robust task management functionalities
Unlock batch task execution, interrupt capabilities, and execute single tasks on a large scale for unparalleled efficiency
How to use WeShop AI Locally
Click the "Start" button on the bottom right to launch the tool. Depending on the computer's specification, it takes some time to launch.
This website will pop up after the software is launched. Select the type of task desired to be executed. Currently, functionalities include text-to-image (txt2img) and image-to-image (img2img).
Select desired parameters for image generations.
Further refine generations by selecting desired ControlNet and AfterDetailer parameters. Click the "Generate" button on bottom right to start generation process.
Resulting images will shortly be displayed after generation process is finished. Click "Edit again" button on the top right to adjust parameters or generate again.
For any bugs encountered, click the "BUG Report" button, then bug info will be sent. Alternatively, you can send feedback from our Github page https://github.com/weshopai/WeShop-UI
Requirements
Windows 7 or above system
8GB VRAM Nvidia GPU
16GB RAM
Download
You can download WeShop UI 1.0.0 from the following sources:
Huggingface: https://huggingface.co/WeShopAI/WeShop_UI_1.0.0
Steam: in progress
Installation & Usage
Unzip weshop-1.0.0.zip and weshop-model-pack.zip
Replace weshop/models with weshop-model-pack/models
Replace weshop/extensions/sd-webui-controlnet/models with weshop-model-pack/extensions/sd-webui-controlnet/models
Replace weshop/extensions/sd-webui-controlnet/annotator/downloads with weshop-model-pack/extensions/sd-webui-controlnet/annotator/downloads
Click weshop.exe to launch
For Steam users, find Weshop_AI in the market, then directly launch the software from Steam.
Gallery
Acknowledgements
This project makes use of the following open-source projects:
https://github.com/sgreben/http-file-server: A file server to store necessary local files created
https://github.com/pierrre/imageserver: A file server dedicated to store image files created
https://github.com/facebookresearch/segment-anything: Segment uploaded images to create masks
https://github.com/sczhou/CodeFormer: Upscale images