{ "activation": "silu", "architectures": [ "Param1MoEForCausalLM" ], "attention_dropout": 0.0, "auto_map": { "AutoConfig": "configuration_param1MoE.Param7BMoEConfig", "AutoModelForCausalLM": "modeling_param1MoE.Param1MoEForCausalLM" }, "bos_token_id": 1, "eos_token_id": 2, "head_dim": 96, "hidden_act": "silu", "hidden_size": 1536, "initializer_range": 0.02, "intermediate_size": 512, "max_position_embeddings": 4096, "model_type": "param1MoE", "num_attention_heads": 16, "num_experts_per_tok": 8, "num_hidden_layers": 40, "num_key_value_heads": 4, "num_local_experts": 64, "output_router_logits": false, "quantization_config": { "_load_in_4bit": true, "_load_in_8bit": false, "bnb_4bit_compute_dtype": "bfloat16", "bnb_4bit_quant_storage": "uint8", "bnb_4bit_quant_type": "nf4", "bnb_4bit_use_double_quant": false, "llm_int8_enable_fp32_cpu_offload": false, "llm_int8_has_fp16_weight": false, "llm_int8_skip_modules": [ "lm_head", "q_proj", "k_proj", "v_proj", "gate", "router", "o_proj" ], "llm_int8_threshold": 6.0, "load_in_4bit": true, "load_in_8bit": false, "quant_method": "bitsandbytes" }, "rms_norm_eps": 1e-05, "rope_theta": 1000.0, "router_aux_loss_coef": 0.02, "router_jitter_noise": 0.0, "sliding_window": null, "tie_word_embeddings": false, "torch_dtype": "float16", "transformers_version": "4.53.3", "use_cache": true, "vocab_size": 128000 }