{
    "architectures": [
        "LlamaForCausalLM"
    ],
    "model_type": "llama",
    "torch_dtype": "float16",
    "transformers_version": "4.38.0",
    "vocab_size": 32000,
    "bos_token_id": 1,
    "eos_token_id": 2,
    "hidden_size": 5120,
    "intermediate_size": 14080,
    "num_attention_heads": 40,
    "num_hidden_layers": 36,
    "num_key_value_heads": 40,
    "max_position_embeddings":8192
}