{ "vocab_size": 50304, "d_model": 384, "n_layers": 8, "n_dense_layers": 8, "n_q_heads": 6, "n_kv_heads": 2, "head_dim": 64, "qk_norm": true, "rope_theta": 10000.0, "attn_softcap": 0.0, "dense_ffn": 1024, "expert_ffn": 64, "n_experts": 4, "top_k": 2, "n_shared": 0, "gating": "sigmoid", "norm_topk_prob": false, "balancing": "aux_free", "aux_loss_coef": 0.001, "z_loss_coef": 0.001, "bias_update_rate": 0.001, "router_init_std": 0.02, "tie_embeddings": true, "scale_embeddings": false, "final_z_loss_coef": 0.0001, "logit_softcap": 0.0, "diffusion": false, "mask_token_id": 50257, "mask_eps": 0.001, "n_mtp": 0, "mtp_weight": 0.1, "init_std": 0.02, "expert_backend": "grouped", "fused_ce": true, "ce_chunk": 4096, "fp8_head": false, "fp8_x_scale": 1.0, "fp8_w_scale": 1.0, "fp8_grad_scale": 1.0, "preset": "30M" }