FoxxMD Posted April 15 Share Posted April 15 (edited) Application Name: aiyabot Application Site: https://github.com/Kilvoctu/aiyabot Github Repo: https://github.com/Kilvoctu/aiyabot Docker Hub: https://github.com/Kilvoctu/aiyabot/pkgs/container/aiyabot Template Repo: https://github.com/FoxxMD/unraid-docker-templates Overview aiyabot is a self-hosted discord bot for interacting with a Stable Diffusion API to create AI generated images. It is compatible with AUTOMATIC1111's Stable Diffusion web UI or SD.Next APIs and provides these features through Discord commands: text2image image2image prompt from image generate prompts from text upscale image image identification (metadata query) Generation settings are also configurable globally or per channel: Model to generate with Width/Height Extra nets (lora, hypernet) Prompt/Negative prompt Batch generation settings highres fix/face restore/VAE strength/guidance(cfg)/clip skip/styles mark as spoiler Be aware that aiyabot will help you craft valid generation commands but it does not magically "make good images" -- the bot will only generate with the settings you specifically configure. You will need a basic understanding of Stable Diffusion in order to achieve good result. Requirements Discord Bot - You need to create a Discord bot in a Server you have appropriate permissions in and have a token ready to use with Aiyabot. This is covered in Setting up a Discord bot in the project wiki. Compatible Stable Diffusion instance - The bot needs access to a SD API. Recommend to use stable-diffusion in CA Apps with (04 SD.Next) or (02 Automatic1111) Usage Map the SD container's API port: * Container Port 7860 -> Host Port 7860 Fill in the CA app template with your discord bot token and URL to the SD API and the start the app to get the bot running. All commands and configuration is done through discord slash commands in the server the bot is running on. Generating Prompts Prompt generation (using /generate command) is disabled by default because it requires downloading and installing an ML model into the container which can be a large download and take a long time to process. If you wish to use this command then modify the CA template variable for USE_GENERATE (Download and install optional prompt generation command) from false to true. Edited April 15 by FoxxMD Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.