{
  "vocab_size": 8192,
  "d_model": 384,
  "n_layers": 6,
  "head_dim": 64,
  "n_experts": 160,
  "routed_experts": 2,
  "context": 1024,
  "rope_base": 1024.0,
  "attention_backend": "torch",
  "expert_backend": "torch",
  "norm_lm": 14.85117682804544,
  "norm_max": 75.46884004363606,
  "lr_lm_head": 0.005469599322971592,
  "lr_matrix": 0.031504192412483305,
  "lr_embed": 0.318139201357217,
  "lr_norm": 0.5010942407685995,
  "adam_beta1": 0.9781894638349105,
  "adam_beta2": 0.9772230178286921,
  "muon_beta": 0.883478680846439,
  "matrix_rms": 0.915686752300296,
  "embed_init_std": 0.3873248839256641,
  "embed_norm": 5.521504655293985,
  "residual_scale": 0.7740658568703003,
  "norm_init": 0.8209734484201097,
  "router_temperature": 1.137749373697223,
  "warmup_steps": 145,
  "micro_batch_size": 128,
  "global_batch_tokens": 524288,
  "total_tokens": 1500000000,
  "compile": true,
  "data_dir": "data/fineweb10B_sp8192",
  "qb_bias_space": "logit",
  "eval_every": 0,
  "save_every": 0,
  "qb_diagnostics": false,
  "qb_logit_histogram": "compact",
  "qb_load_profile": "uniform",
  "qb_lower_slack": 0.1,
  "qb_upper_slack": 0.5,
  "qb_inverse_rank_offset": null
}