{ "architectures": [ "LlamaForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 131073, "dtype": "bfloat16", "eos_token_id": 131074, "head_dim": 128, "hidden_act": "silu", "hidden_size": 6144, "initializer_range": 0.005, "intermediate_size": 21504, "max_position_embeddings": 65536, "mlp_bias": false, "model_type": "llama", "num_attention_heads": 48, "num_hidden_layers": 60, "num_key_value_heads": 8, "pad_token_id": 131072, "pretraining_tp": 1, "rms_norm_eps": 3e-06, "rope_scaling": null, "rope_theta": 200000, "tie_word_embeddings": false, "transformers_version": "4.57.1", "use_cache": true, "vocab_size": 131102 }