{
  "depth": 20,
  "aspect_ratio": 64,
  "head_dim": 128,
  "max_seq_len": 2048,
  "vocab_size": 32768,
  "params_total": 896533746,
  "params_scaling": 435160240,
  "flops_per_token": 3240107184,
  "unique_train_tokens": 622758683,
  "target_tokens_ratio": 4569182520,
  "target_tokens_epochs": 2491034732,
  "horizon": "epochs",
  "num_iterations": 6312,
  "total_batch_size": 524288,
  "device_batch_size": 32,
  "world_size": 8,
  "device_batch_size_source": "ladder.json device_batch_size",
  "grad_accum_steps": 1,
  "train_tokens": 3309305856,
  "epochs": 5.31,
  "tokens_per_param": 7.6,
  "train_flops": 10722505678078869504,
  "window_pattern": "SSSL",
  "attention": "fa3",
  "fp8": false,
  "eval_tokens": 10485760,
  "eval_every": 315,
  "save_every": 594,
  "keep_checkpoints": true,
  "base_overrides": {},
  "warmdown_ratio": 0.199985,
  "stream": true,
  "decay_at": 5050,
  "plain_at": 5681,
  "loader_tails": true,
  "regularization": {
    "wd_schedule": "floor",
    "wd_floor": 0.5,
    "table_wd_mult": 1.0,
    "dropout": 0.0,
    "dropout_from": 0.5
  },
  "dropout_from_step": null,
  "weight_decay": {
    "horizon": "trained",
    "base": 0.28,
    "nanochat_scaled": 0.070843,
    "scaled": 0.070843,
    "decay_per_step": 0.001417,
    "ema_tokens": 370032765
  },
  "wd_horizon": null,
  "save_steps": [],
  "seed": null,
  "arm": null,
  "device_type": "cuda",
  "frozen_at": "2026-10-14T15:50:00"
}