{
  "parameter_count": 1760624640,
  "parameter_summary": {
    "total": 1760624640,
    "parts": {
      "blocks": 1528823808,
      "lm_head": 115900416,
      "other": 115900416
    }
  },
  "n_layer": 24,
  "n_head": 18,
  "n_embd": 2304,
  "head_dim": 128,
  "depth_scale_mode": "none",
  "core_repetitions": 1,
  "core_layer_count": null,
  "resolved_core_layer_count": null,
  "prelude_layer_count": null,
  "resolved_prelude_layer_count": null,
  "coda_layer_count": null,
  "resolved_coda_layer_count": null,
  "effective_depth": 24,
  "expected_effective_depth": 24.0,
  "recurrence_rmsnorm": "none",
  "recurrence_emb_alpha": 0.0,
  "decoder_inject": false,
  "core_residual_branch_scale": 1.0,
  "resolved_core_residual_branch_scale": null,
  "core_repetition_schedule": "none",
  "core_repetition_crossover_fraction": null,
  "core_repetition_low": 1,
  "core_repetition_breakpoints": null,
  "lsched_lr_warmup_steps": 0,
  "lsched_lr_change_steps": [],
  "dep_grow_events": [],
  "target_core_repetitions": 1,
  "final_eval_active_core_repetitions": null,
  "observed_train_core_repetitions_mean": null,
  "observed_train_core_repetitions_min": null,
  "observed_train_core_repetitions_max": null,
  "observed_train_core_repetitions_histogram": null,
  "recurrence_truncation_schedule": "none",
  "recurrence_bp_window": 0,
  "matrix_lr": 0.010038349968232205,
  "embedding_lr": 0.0016061359949171529,
  "unembedding_lr": 0.0005678548267566215,
  "lr_multiplier": 1.0,
  "effective_matrix_lr": 0.010038349968232205,
  "effective_embedding_lr": 0.0016061359949171529,
  "effective_unembedding_lr": 0.0005678548267566215,
  "weight_decay": 0.1,
  "matrix_weight_decay": 0.1,
  "embedding_weight_decay": 0.1,
  "unembedding_weight_decay": 0.1,
  "embedding_wd_skip_norm": 0.0,
  "adam_betas": [
    0.8,
    0.99
  ],
  "adam_eps": 1e-08,
  "lr_schedule": "wsd",
  "warmup_steps": 5,
  "warmdown_ratio": 0.8,
  "output_multiplier": 1.0,
  "residual_branch_multiplier": 0.5,
  "wte_init_std": 0.005,
  "uniform_init_scale": 0.5,
  "total_batch_size": 524288,
  "device_batch_size": 8,
  "num_epochs": 1,
  "estimated_num_iterations": 20340,
  "lr_schedule_steps": 20149,
  "resolved_lr_schedule_steps": 20149,
  "max_train_steps": 20149,
  "resolved_max_train_steps": 20149,
  "completed_epochs": 0,
  "seed": 42,
  "train_token_limit": 10669517701,
  "val_token_limit": 10000000,
  "token_limit_cutoff_devices": 32,
  "active_train_tokens": 10669517701,
  "packed_train_sequences_per_epoch": 5207040,
  "packed_train_tokens_per_epoch": 10664017920,
  "active_val_packed_sequences": 4864,
  "active_val_packed_tokens": 9961472,
  "val_loss": 2.662481609143709,
  "final_val_loss": 2.662481609143709,
  "final_train_loss": 2.6769662301782877,
  "best_val_loss": 2.662481609143709,
  "best_val_bpb": 0.8653310518968823,
  "best_val_epoch": 1.0,
  "train_val_per_epoch": [
    {
      "epoch": 1,
      "step": 20149,
      "train_loss": 2.6769662301782877,
      "val_loss": 2.662481609143709,
      "val_bpb": 0.8653310518968823
    }
  ],
  "core_block_residual_rms": 1.9576795101165771,
  "last_train_mfu": 49.3306345373268,
  "last_train_tok_per_sec": 370021,
  "last_train_step_time_seconds": 1.4169137477874756,
  "total_training_time_seconds": 28616.788650512695,
  "total_wall_time_seconds": 28928.015945911407,
  "sequence_len": 2048,
  "window_pattern": "L"
}