{
  "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": "logit",
  "attention_backend": "torch",
  "expert_backend": "torch",
  "norm_lm": 6.077180549043502,
  "norm_max": 104.82153460451599,
  "lr_lm_head": 0.003553694160237239,
  "lr_matrix": 0.028766129661875316,
  "lr_embed": 0.28812423010202187,
  "lr_norm": 0.7000202775898238,
  "adam_beta1": 0.9427557834609671,
  "adam_beta2": 0.9649890434637186,
  "muon_beta": 0.9262914664906484,
  "matrix_rms": 1.015611833228007,
  "embed_init_std": 0.10171903038785925,
  "embed_norm": 3.797179034987117,
  "residual_scale": 1.911265252824758,
  "norm_init": 0.7866334804197714,
  "router_temperature": 0.9711917630526921,
  "warmup_steps": 118,
  "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": 0.5,
  "qb_inverse_rank_offset": null
}