dappnode_package.json

{
  "name": "ollama.dnp.dappnode.eth",
  "version": "0.1.0",
  "upstream": [
    {
      "repo": "ollama/ollama",
      "version": "v0.17.7",
      "arg": "OLLAMA_VERSION"
    }
  ],
  "mainService": "ollama",
  "shortDescription": "Local LLM inference engine with GPU acceleration",
  "description": "Run large language models locally on your DAppNode with GPU acceleration. Ollama provides a fast and efficient LLM inference engine with AMD ROCm support.\n\n**Features:**\n- AMD GPU acceleration via ROCm\n- Complete privacy - all processing stays local\n- Support for multiple LLM models (Llama, Mistral, CodeLlama, etc.)\n\n**Requirements:**\n- AMD GPU with ROCm support\n- At least 8GB RAM (16GB+ recommended)\n- Sufficient storage for models (10GB+ recommended)",
  "type": "service",
  "author": "DAppNode Association <admin@dappnode.io> (https://github.com/dappnode)",
  "license": "GPL-3.0",
  "categories": [
    "AI"
  ],
  "links": {
    "Models library": "https://ollama.com/library"
  },
  "architectures": [
    "linux/amd64"
  ]
}