This is just a demo of using llm (GPT) with stable diffusion and huggingface.

If you like this project, please give it a star ⭐️ to continue the development.

photo-GPT-telegram

is a telegram bot based on large language models (GPT-3) for image tasks like creating image, captioning, editing and general conversation like chatting about images based on stable diffusion and huggingface and langchain.

How to use (tested with macos only)

1- create a telegram bot using @BotFather and get the token in env variable TELEGRAM_BOT_TOKEN

2- Get OpenAI API key and put it in env variable OPENAI_API_KEY or Cohere API key and put it in env variable COHERE_API_KEY

3- bash git clone https://github.com/manskx/photo-GPT-telegram.git cd photo-GPT-telegram python -m venv venv source venv/bin/activate python -m pip install -r requirements.txt

4- bash export PYTHONPATH=$PWD cd photo_gpt python main.py

5- send /start to your bot in telegram

example

https://user-images.githubusercontent.com/7218339/220134549-1bb94ab8-646e-477d-919d-617b9cb78cb5.mp4

authors

• @manskx and copilot