Infinity Memory implementation

Hi @haithemyk0707

Welcome to the OpenAI community.

You can use embeddings to find out semantically relevant messages from from the conversation and pass them to answer questions.

I also write a tutorial about it: Use embeddings to retrieve relevant context for AI assistant

While it may not literally give infinite context, it should help you go well beyond the context length of the chat completion models and also save tokens.

2 Likes