{ "_name_or_path": "/apps/sharedstorage/oceanstor-a800/sft/deepspeed_chat_moe/Pretrained_model/final_with_router/", "activation": "silu", "architectures": [ "JiutianForCausalLM" ], "attention_dropout": 0.0, "auto_map": { "AutoConfig": "configuration_jiutian.JiutianConfig", "AutoModelForCausalLM": "modeling_jiutian.JiutianForCausalLM" }, "bos_token_id": 1, "eos_token_id": 0, "hidden_act": "silu", "hidden_size": 5120, "initializer_range": 0.02, "intermediate_size": 13824, "max_position_embeddings": 4096, "model_type": "jiutian", "num_attention_heads": 40, "num_experts_per_tok": 2, "num_hidden_layers": 40, "num_key_value_heads": 40, "num_local_experts": 8, "output_router_logits": true, "rms_norm_eps": 1e-08, "rope_theta": 10000.0, "router_aux_loss_coef": 0.01, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "bfloat16", "transformers_version": "4.36.2", "use_cache": true, "use_cope": false, "vocab_size": 69120 }