from langchain_community.vectorstores import FAISS def create_vector_store(documents, embedding_model): vectorstore = FAISS.from_documents(documents, embedding_model) return vectorstore def save_vector_store(vectorstore, path="faiss_index"): vectorstore.save_local(path) def load_vector_store(path, embedding_model): return FAISS.load_local(path, embedding_model)