About

A simple tool designed to make efficient use of LLM's by leveraging Retrieval Augmented Generation in order to obtain relevant information from a provided corpus of text quickly, and reduce chances of hallucination.

Steps

Virtual Environment

• Create a virtual environment in a directory
• mkdir <name> && cd <name>
• python3 -m venv .venv/<name>
• activate with source .venv/bin/activate

Dependencies

• Dependencies have beel listed in requirements.txt
• install them with pip install -r requirements.txt

Config

• Configure, the prompt, corpus, model to be used in config.toml

Models

• Generate API keys to either OpenAI or Gemini
• Populate .env with OPENAI_API_KEY or GEMINI_API_KEY
• perform source .env before running the program

Run

• run python3 main.py
• output written to response.txt