{
  "architectures": [
    "LlamaForCausalLM"
  ],
  "model_type": "llama",
  "torch_dtype": "float16",
  "prompt_template": {
    "type": "alpaca",
    "description": "Alpaca instruction-following format"
  }
}