{ "architectures": [ "NanochronoForCausalLM" ], "attention_multiplier": 1.2, "aux_gate_dim": 12, "aux_layers": [ 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27 ], "bos_token_id": 32759, "dtype": "bfloat16", "eos_token_id": 32763, "final_logit_softcapping": 15.0, "hidden_size": 1792, "intermediate_size": 7168, "layer_types": [ "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention" ], "max_position_embeddings": 2048, "mix_gate_dim": 24, "model_type": "sn38-nanochrono", "num_attention_heads": 14, "num_hidden_layers": 28, "num_key_value_heads": 14, "pad_token_id": 32763, "rope_theta": 100000.0, "sliding_window": 512, "tap_layer": 14, "tie_word_embeddings": false, "transformers_version": "5.15.0", "use_cache": true, "vocab_size": 32768 }