This is a 5G expert system. It uses retrieval augmented generation to augment the Q&A capabilities of an LLM.
Make sure you have a settings.yml file in the same directory as settings.py.
It should have these vars:
API_KEYDOC_DIRMODEL_NAMENUM_EXTRA_DOCS-> the number of additional docs to retrieve per run. NOT depth, closer to top kCHROMA_DIR-> The directory the chromadb sqlite db will be storedSPEC_COLL_NAME: "context" -> These should be set to these values. They represent what the collections in the chromadb database will be named. There are two collections, one to store spec content and the other to store tdoc content.TDOC_COLL_NAME: "reason"IS_SMART_RETRIEVAL: (boolean) indicates whether or not smart retrieval/ deep context is turned on.ARE_ENDPOINTS_GETTABLE: (boolean) indicates whether the endpoints you provide to the autofetcher can be immediately queried with a get request to download a file or not.NUM_DOCS_INITIAL_RETRIEVAL: (int) the number of documents retrieved by the first round (non deep context) of retrieval
Do python frontend.py
gradio langchain_openai langchain_community chromadb bs4 docx2txt MarkItDown
Initializing the db only happens when the resync button is hit. The DB is persistent so between runs you access the same database.