RAG Agent with LangChain and LLaMa3

Fully local RAG agent powered by LLaMa3 and LangChain

Installation

Python

Create a virtual environment and install the requirements:

python -m venv .venv
pip install -r requirements.txt

LLM

If you don't have ollama installed, follow this doc.

And then run:

ollama pull llama3

LangSmith (Optional)

If you want to use LangSmith, copy the .env.example to .env and fill the LANGSMITH_API_KEY with your API key.

Running the project

Go to Run and Debug in VSCode and select Debug App.

References

• YouTube
• Jupyter Notebook
• LLaMa3
• Ollama

Troubleshooting

Ollama

If you have any issues with ollama running infinetely, try to run the following command:

sudo systemctl restart ollama

Or:

pgrep ollama # returns the pid
kill -9 <pid>
sudo systemctl start ollama