Women in data science

Build a very simple AI chatbot leveraging AI Agents and a locally hosted LLM.

Ollama set up

• We will be using Ollama to run a local LLM. Make sure you have at least 8GB RAM since we are using Gemma:2b. Download Ollama here - https://ollama.com/
• Run Ollama pull gemma:2b

Python env set up

• I like using Python venv to manage local Python environments.
• python3 -m venv <env-name>
• source <env-name>/bin/activate
• Install required package dependencies: pip install -r requirements.txt in backend to install backend files and also in frontend to install frontend files
• You need to get an Open AI Api key for the CrewAI agent we are using so create a free account and set the API key in your environment   export OPENAI_API_KEY=<openai api key>

Run the Uvicorn server in backend folder:  

• uvicorn api:app --reload

Run the streamlit server in front end folder:

• streamlit run app.py