{
    "architectures": [
        "NanoChatForCausalLM"
    ],
    "hidden_size": 1280,
    "intermediate_size": 5120,
    "max_position_embeddings": 2048,
    "model_type": "nanochat",
    "num_attention_heads": 10,
    "num_hidden_layers": 20,
    "num_key_value_heads": 10,
    "rope_theta": 10000.0,
    "torch_dtype": "bfloat16",
    "transformers_version": "4.0.0",
    "vocab_size": 65536
}