RAG Application Documentation

Contents:

• Introduction
  • What exactly is RAG?
  • Key Technologies Used
    • Mistral & llama3 via Ollama
    • 🦜️🔗 LangChain
    • Streamlit
• Installation
  • Prerequisites
  • Steps to Install Dependencies
• Usage
• Application Structure
  • 1. LLM Setup
  • 2. Embedding and Vector Store
  • 3. Document Processing
  • 4. Retrieval-Augmented Generation (RAG) Chain
  • 5. Streamlit Interface
• LangChain RAG Workflow
  • Step 1: LLM Initialization
  • Step 2: PDF Upload and Chunking
  • Step 3: Vector Store Creation
  • Step 4: Query Workflow
  • Step 5: Output Display
• Testing
  • 1. Uploading a PDF and Asking Contextual Questions
  • 2. Asking Out-of-Context Questions
  • 3. Testing with AceGPT:7b (Fine-Tuned LLM for Arabic)
    • Overall Testing Results