Simple telegram bot that generates images based on users text inputs by harvestingm00n 🍃
Currently, the backend is hooked up with Stable Diffusion v2.0 🎨 . It works comfortably on a RTX 3060 with 6gb VRAM.
It is built using the ptb package as well as the huggingface 🤗 diffusers library
I will be continuously improving and working on this model more so as a small project :
Currently, the board can only send photos that are up to 5mb big due to the limitations of ptb (as well as other post processing factors like limiting pixel to only 10000px). However, I am working on a feature where the bot will send a document which would bypass this issue. Any suggestions do send a pull request or let me know! 😄
v1.0: Original release of model
v1.1: Added support for sending document for higher quality images!
Instructions on how to install and set up the project.
Step 1: Clone the repo using
git clone https://github.com/harvestingmoon/StableVisionBot.git
Step 2: Install the requirements needed for the bot to run
pip install requirements.txt
Step 3: Insert your model (Any Stable Diffusion Model from Huggingface) and telegram bot API_KEY on on config.yaml
=> in the future I will open it up to other models
Step 4: Run the model either using cmd prompt using the following code (Before that, ensure that your command prompt is in the same directory as the file):
python bot.py
or run it on VSCode by clicking the top right triangle
/start: To Start The Bot
/info: Technical Information about the bot
Model currently supports the following features:
- Text to image AI generation
- Image to Image AI generation (Image editing)
- Higher quality images now sent via PDF
Future features:
- Support other pipelines (currently support stable diffusion only)
- Image manipulation via text
This project is licensed under the terms of the MIT License.