Hi! So I’ve been having a persistent type of error for the last two weeks. Basically I have a RAG based LLM query engine which uses a dataset of Dungeons and Dragons PDFs to answer questions about character classes and game rules.
I’m currently trying to speed it up and one thing I am trying to do is store the VectorStoreIndex in a pickle file which will then be called when I run the LLM query engine rather than generating the VectorStoreIndex everytime. However when I try to do this I run into one of two errors: either 'AttributeError: HuggingFaceEmbedding has no attribute ‘_model’ OR 'AttributeError: LangchainEmbedding object has no attribute ‘_langchain_embedding’. I’ve linked the repository where this code is so you can take a look at it for yourself: https://github.com/vishaljoshi24/oracle. Please look at test.py as that is what is leading to the second of the errors I’ve highlighted here.