{ "_name_or_path": "", "architectures": [ "MouhForCausalLM" ], "attn_every": 6, "bos_token_id": 257, "capacity_factor": 1.25, "chunk_size_feed_forward": 0, "dtype": null, "enforce_capacity_at_inference": false, "eos_token_id": 258, "expert_intermediate_size": 2304, "gate_rank": 64, "hidden_size": 1024, "id2label": { "0": "LABEL_0", "1": "LABEL_1" }, "initializer_range": 0.02, "is_encoder_decoder": false, "label2id": { "LABEL_0": 0, "LABEL_1": 1 }, "layer_norm_eps": 1e-05, "max_position_embeddings": 65536, "model_type": "mouh", "mouh_parameter_report": { "active": 832265617, "active_ratio": 0.550533848228217, "embedding": 524288, "expert": 1358954496, "non_expert": 152788369, "total": 1511742865 }, "mouh_preset": "mouh-1.5b", "num_attention_heads": 16, "num_experts": 8, "num_experts_per_tok": 4, "num_hidden_layers": 24, "output_attentions": false, "output_hidden_states": false, "pad_token_id": 256, "problem_type": null, "return_dict": true, "rope_base": 10000.0, "rope_scaling_factor": 1.0, "router_aux_loss_coef": 0.01, "router_bias_update_rate": 0.001, "router_type_prior_strength": 1.0, "ssm_chunk_size": 64, "ssm_conv_kernel": 4, "ssm_expand": 2, "ssm_head_dim": 64, "ssm_state_size": 128, "tie_word_embeddings": true, "use_triton": true, "vocab_size": 512 }