{ "architectures": [ "LLaMAForCausalLM" ], "bos_token_id": 50256, "eos_token_id": 50256, "hidden_act": "silu", "hidden_size": 1536, "intermediate_size": 5376, "initializer_range": 0.02, "max_sequence_length": 1024, "model_type": "llama", "num_attention_heads": 24, "num_hidden_layers": 14, "rms_norm_eps": 1e-05, "transformers_version": "4.28.1", "use_cache": true, "vocab_size": 50257 }