paper-qa-agent

A RAG-based QA agent for the paper Retrieval-Augmented Generation for Large Language Models: A Survey.

Setup

.env file setup

Create a .env file from this template. Fill in your own OpenAI API Key.

WCS_URL=https://tobys-sandbox-oo0vfo50.weaviate.network
WCS_API_KEY=HaDolSevqI9HxhvwMnl9DssgKMObfc0eXfC7
OPENAI_API_KEY=sk-proj-xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Install dependencies

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Usage

Use query_qa_rag_agent.ipynb.

Architecture

Tech Debt

• retrievers.py and vector_dbs.py have some duplicate code
• Weaviate client management needs to be refactored to use the weaviate context manager [1]
• preprocessors.py
  • formats tables weird
  • doesn't handle images
  • doesn't format subheadings correctly
• vector_dbs.py: vectorization strategy is declared here, and I think it makes more sense for it to be declared in indexer.py
• indexers.py: calling syntax doesn't make it clear what side effects the constructor has
• dspy-rag-pipeline.ipynb:
  • teleprompter parameters are unoptimized
  • both a global and contextualized language model are used, and this is harder to read. Change to contextualized LMs only

[1] https://weaviate.io/developers/weaviate/client-libraries/python#connection-termination