# Model to use with Ollama — change this to any Ollama-compatible model
# Examples: llama3, llama3.1, llama3.2, gemma2, mistral, phi3, qwen2, codellama
OLLAMA_MODEL=llama3