Basic RAG setup

[getnamo]

Enable feeding in data into a memory DB and vector embed it. Use a small and simple vector search to support RAG lookup of info. Potentially structure various chunking and search strategies. Likely a secondary model link to embedding model for vectorization.

[getnamo]

Resources:

• memory db c lib: https://github.com/facebookresearch/faiss/blob/main/c_api/INSTALL.md
• using embedding model to generate vectors: https://github.com/ggerganov/llama.cpp/blob/ceb/bert/examples/embedding/embedding.cpp already built in support for Bert style embedders.
• NB best results are achieved if we combine with something like a normal search (matches exact text which is a weakness of vector searches) e.g. using bm25 algorithm: https://github.com/scorta/BM25