llm_chat is a robust backend system built with Python and FastAPI, designed to facilitate dynamic interactions with large language models (LLMs). The system supports both user-driven conversations and automated LLM-to-LLM dialogues, allowing users to engage or intervene at any point.
- Start a new conversation with an LLM
- Continue a conversation by sending messages
- Generate messages from an LLM within an ongoing conversation
- View conversation history to see past messages
- List all available LLM models for interaction
- OpenAPI documentation for seamless exploration of endpoints
- Start Conversation:
/start_conversation(POST) - Send Message:
/send_message(POST) - Generate Message:
/generate_message(POST) - Session History:
/session_history/{session_id}(GET) - Models:
/models(GET) - API Documentation:
/docs(GET)
To set up and run the project, execute the following command:
make upThen, open your browser and navigate to http://localhost:8000/docs to view the API documentation.
python cli.py
We welcome contributions! Please feel free to open an issue or submit a pull request for any enhancements or bug fixes.
This project is licensed under the MIT License. See the LICENSE file for details.