This is a Telegram bot designed to interact with various language models. It offers support for system prompts and image processing capabilities.
- Integration with the Telegram platform
- Support for any number of language models (compatible with OpenAI API format)
- Ability to handle system prompts, generation parameters, and characters
- Image processing and generation (with gpt-image-1)
- Function calling with CodeAct (from smolagents)
To install and run the bot, follow these steps:
- Clone the repository:
git clone https://github.com/IlyaGusev/saiga_bot
- Navigate to the project directory:
cd saiga_bot
- Install the required dependencies:
uv venv
make install
- Add a token from @BotFather and an admin's user name and ID to configs/bot.json.
- Add URLs and tokens to configs/providers.json. Modify all other configs if needed.
- Run the bot with
make serve
make serve
Or with an explicit command:
uv run python -m src.bot --bot-config-path configs/bot_prod.json --providers-config-path configs/providers.json --db-path db.sqlite --characters-path configs/characters.json --yookassa-config-path configs/yookassa.json --localization-config-path configs/localization.json --tools-config-path configs/tools_prod.json
For any questions or feedback, please open an issue on the GitHub repository or contact @YallenGusev.
