{ "architectures": [ "Llama" ], "bos_token_id": 1, "context_len": 1024, "eos_token_id": 2, "eps": 1e-06, "hidden_size": 100, "model_type": "llama", "num_attention_heads": 10, "num_hidden_layers": 10, "pad_token_id": 2, "rope_theta": 10000.0, "torch_dtype": "float32", "transformers_version": "4.44.2", "vocab_size": 32000 }