{ "architectures": [ "TransformerForCausalLM" ], "attention_bias": false, "bos_token_id": 151643, "elementwise_affine": true, "eos_token_id": 151645, "fuse_cross_entropy": false, "fuse_norm": true, "fuse_swiglu": true, "hidden_act": "swish", "hidden_ratio": 4, "hidden_size": 3584, "initializer_range": 0.02, "intermediate_size": 18944, "max_position_embeddings": 32768, "model_type": "transformer", "norm_eps": 1e-06, "num_heads": 28, "num_hidden_layers": 28, "num_kv_heads": 4, "qk_norm": false, "qkv_bias": true, "rope_theta": 1000000.0, "tie_word_embeddings": false, "torch_dtype": "bfloat16", "transformers_version": "4.53.3", "use_cache": true, "use_l2warp": false, "vocab_size": 151936, "window_size": null }