ammaradam/rag-toolbox

Compilation of all relevant tools for Retrieval-Augmented Generations

RAG Toolbox

Let's chat with your documents!

Tech Stacks

• Frontend: Streamlit
• VectorDB: Qdrant
• Orchestration: LlamaIndex
• LLM: OpenAI GPT3.5
• Embedding model: Fastembed

User Stories

• User wants to upload own PDF document
• User wants to preview uploaded PDF document
• User wants to ask question related to PDF document
• User wants Bot to only reply within context
• User wants to chat with previous conversation context

Architecture

Knowledge base --> Chunks --> Embedding model --> Vector DB

Vector DB --> Prompt template --> LLM --> Chat UI

Quickstart

Step 1: Install dependencies

pip install -r requirements.txt

Step 2: Run application

streamlit run app.py

Application will be opened in a browser.

Done!

Credit to Akshay for inspiration!