{ "architectures": [ "NanoChatForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 32759, "dtype": "bfloat16", "eos_token_id": 32763, "final_logit_softcapping": 15.0, "head_dim": 128, "hidden_act": "silu", "hidden_size": 1152, "initializer_range": 0.02, "intermediate_size": 3072, "max_position_embeddings": 32768, "model_type": "nanochat", "num_attention_heads": 9, "num_hidden_layers": 18, "num_key_value_heads": 9, "pad_token_id": 32763, "rms_norm_eps": 1e-06, "rope_parameters": { "rope_theta": 100000.0, "rope_type": "default" }, "tie_word_embeddings": false, "transformers_version": "4.57.3", "use_cache": true, "use_rope": false, "vocab_size": 32768 }