Using Ollama

Simple Setup

The easiest way to use Ollama out of the box with coco is to add or update the service option in the coco config to be ollama.

Full Customization

If you want to go deeper than shorthand aliases of openai and ollama and tinker with the underlying model, temperature, endpoint, or other settings available in the OpenAILLMService or OllamaLLMService interfaces.

Example of Service Object

{
  "defaultBranch": "main",
  "mode": "interactive",
  "service": {
    "provider": "ollama",
    "model": "qwen2.5-coder:latest",
    "endpoint": "http://localhost:11434",
    "maxConcurrent": 1,
    "tokenLimit": 2024,
    "temperature": 0.4,
    "maxParsingAttempts": 3,
    "authentication": {
      "type": "None"
    }
  }
}