{
    "framework": "pytorch",
    "task": "chat",
    "model": {
        "type": "llama2"
    },
    "pipeline": {
        "type": "llama2-text-generation-chat-pipeline"
    }
}