Caellwyn/long-memory-character-chat
📦 Open Source ProjectCaellwyn
An AI character chat framework featuring integrated medium and long-term memory for persistent, context-aware conversations.
The long-memory-character-chat repository provides a sophisticated implementation of memory management for LLM-based chatbots. Unlike standard chat interfaces that rely solely on short-term context windows, this framework utilizes a multi-tiered memory system. It employs medium-term memory for immediate session context and long-term memory via vector databases to store and recall historical user interactions. Built with Python and LangChain, the project integrates seamlessly with various LLMs, including Mistral-7B and OpenAI models. Key technical features include automated document retrieval, semantic search capabilities for past memories, and a Streamlit-based web interface for rapid prototyping. The architecture is modular, allowing developers to swap out embedding models or vector stores, making it highly adaptable for custom character-driven applications, roleplay bots, or personalized AI assistants that require deep, evolving context.
💡Highlights
- ├─Multi-tiered memory architecture
- ├─LangChain-powered RAG integration
- └─Supports Mistral-7B and OpenAI
🎯For
- ├─AI Developers
- ├─Chatbot Enthusiasts
- └─Narrative Designers