{
  "vocab_size": 8192,
  "d_model": 512,
  "n_layers": 8,
  "head_dim": 64,
  "n_experts": 64,
  "routed_experts": 2,
  "context": 1024,
  "rope_base": 1024.0,
  "qb_bias_space": "sigmoid",
  "attention_backend": "torch",
  "expert_backend": "torch",
  "norm_lm": 6.249480829286974,
  "norm_max": 96.13180859137935,
  "lr_lm_head": 0.0085358500110653,
  "lr_matrix": 0.04075739496093826,
  "lr_embed": 0.1235913456606203,
  "lr_norm": 0.6397345453678573,
  "adam_beta1": 0.9477204225975786,
  "adam_beta2": 0.960551068103405,
  "muon_beta": 0.9363747440687935,
  "matrix_rms": 1.2741794066611543,
  "embed_init_std": 0.14465988865912652,
  "embed_norm": 3.181372653501972,
  "residual_scale": 2.021690545828811,
  "norm_init": 0.6650629367907821,
  "router_temperature": 1.1897449806635323,
  "warmup_steps": 166,
  "micro_batch_size": 128,
  "global_batch_tokens": 524288,
  "total_tokens": 1500000000,
  "compile": true,
  "data_dir": "data/fineweb10B_sp8192",
  "eval_every": 0,
  "save_every": 0,
  "qb_diagnostics": false,
  "qb_logit_histogram": "compact",
  "qb_load_profile": "uniform",
  "qb_lower_slack": 0.0,
  "qb_upper_slack": 2.0,
  "qb_inverse_rank_offset": null
}