{ "architectures": [ "NanoChatForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": 0, "dtype": "bfloat16", "eos_token_id": 1, "final_logit_softcapping": 15.0, "hidden_act": "relu2", "hidden_size": 2304, "initializer_range": 0.02, "intermediate_size": 9216, "max_position_embeddings": 8192, "model_type": "nanochat", "num_attention_heads": 18, "num_hidden_layers": 36, "num_key_value_heads": 18, "pad_token_id": 1, "rms_norm_eps": 1e-06, "rope_parameters": { "rope_theta": 10000.0, "rope_type": "default" }, "tie_word_embeddings": false, "transformers_version": "5.0.0.dev0", "use_cache": true, "vocab_size": 65536 }