Running Models Locally

Set up Ollama and llama.cpp. Run open-source models for development and testing.