{
  "vocab_size": 8192,
  "d_model": 512,
  "n_layers": 8,
  "head_dim": 64,
  "n_experts": 64,
  "routed_experts": 4,
  "context": 1024,
  "rope_base": 1024.0,
  "qb_bias_space": "sigmoid",
  "attention_backend": "torch",
  "expert_backend": "torch",
  "norm_lm": 16.656438811550174,
  "norm_max": 69.55572269195358,
  "lr_lm_head": 0.0070686706674333525,
  "lr_matrix": 0.032834104445745554,
  "lr_embed": 0.1973737779461892,
  "lr_norm": 0.8242994374559834,
  "adam_beta1": 0.961945807887999,
  "adam_beta2": 0.9743463428650329,
  "muon_beta": 0.8901712682741592,
  "matrix_rms": 0.8877552778767859,
  "embed_init_std": 0.18445015298144068,
  "embed_norm": 6.062301111439426,
  "residual_scale": 2.1538663014966506,
  "norm_init": 0.44338432684171436,
  "router_temperature": 0.6474627651589311,
  "warmup_steps": 129,
  "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": 0.1,
  "qb_upper_slack": 1.0,
  "qb_inverse_rank_offset": null
}