{ "architectures": [ "Doge2ForCausalLM" ], "attention_bias": false, "attention_dropout": 0.0, "bos_token_id": null, "dynamic_mask_ratio": 0.0, "eos_token_id": 2, "hidden_act": "silu", "hidden_dropout": 0.0, "hidden_size": 576, "initializer_range": 0.02, "intermediate_size": 1536, "is_moe": true, "keep_window_size": 2048, "max_position_embeddings": 2048, "model_type": "doge2", "norm_topk_prob": false, "num_attention_heads": 9, "num_experts": 900, "num_experts_per_tok": 30, "num_hidden_layers": 30, "num_key_value_heads": 3, "output_router_logits": false, "pad_token_id": 0, "rms_norm_eps": 1e-06, "rope_scaling": null, "rope_theta": 10000.0, "router_aux_loss_coef": 0.001, "tie_word_embeddings": false, "torch_dtype": "float32", "transformers_version": "4.51.3", "use_cache": true, "vocab_size": 32768 }