{ "vocab_size": 8192, "max_seq_len": 1024, "dim": 128, "n_layers": 4, "n_dense_layers": 1, "dense_hidden_dim": 352, "norm_eps": 1e-06, "init_std": 0.02, "tie_embeddings": true, "n_heads": 4, "q_lora_rank": 0, "kv_lora_rank": 32, "qk_nope_head_dim": 16, "qk_rope_head_dim": 16, "v_head_dim": 16, "rope_theta": 10000.0, "n_routed_experts": 8, "n_shared_experts": 1, "n_activated_experts": 2, "moe_hidden_dim": 64, "n_expert_groups": 1, "n_limited_groups": 1, "route_scale": 1.0, "bias_update_speed": 0.001, "seq_aux_loss_alpha": 0.0001, "moe_backend": "capacity", "capacity_factor": 2.0 }