Advanced RAG Example

Advanced RAG Example: query markdown documents.

Installing dependencies

Install uv and execute the command uv sync in the project root to install dependencies.

Also install ollama and pull the models llama3.1:8b and nomic-embed-text-v2-moe

Setting up your environment variables

Create an .env file in the project root with your OpenAI and HuggingFace secret tokens

OPENAI_API_KEY=***
HF_TOKEN=***

Option A: RAG with Langchain

Ingesting markdown Ruby on Rails documentation in ror_kb directory

To ingest the markdown files and convert them to a vector (Chroma) database run:

python langchain_rag/ingest.py

Running the Gradio interface

To run the chatbot UI (Gradio) after having executed the ingestion:

python langchain_rag_app.py

Option B: RAG without Langchain

Ingesting markdown Ruby on Rails documentation in ror_kb directory

To ingest the markdown files without using the Langchain

python advanced_rag/ingest.py

Running the Gradio interface

To run the chatbot UI (Gradio) after having executed the advanced ingestion without Langchain:

python advanced_rag_app.py