{
  "depth": 20,
  "aspect_ratio": 64,
  "head_dim": 128,
  "max_seq_len": 2048,
  "vocab_size": 16384,
  "params_total": 644875506,
  "params_scaling": 414188720,
  "flops_per_token": 3114278064,
  "unique_train_tokens": 421945781,
  "target_tokens_ratio": 4348981560,
  "target_tokens_epochs": 1687783124,
  "horizon": "epochs",
  "num_iterations": 4462,
  "total_batch_size": 524288,
  "device_batch_size": 32,
  "world_size": 4,
  "device_batch_size_source": "ladder.json device_batch_size",
  "grad_accum_steps": 2,
  "train_tokens": 2339373056,
  "epochs": 5.54,
  "tokens_per_param": 5.65,
  "train_flops": 7285458191813443584,
  "window_pattern": "SSSL",
  "attention": "fa3",
  "fp8": false,
  "eval_tokens": 6291456,
  "eval_every": 223,
  "save_every": 402,
  "keep_checkpoints": true,
  "base_overrides": {},
  "warmdown_ratio": 0.199708,
  "stream": true,
  "decay_at": 3571,
  "plain_at": 4016,
  "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": 0.065924,
    "scaled": 0.065924,
    "decay_per_step": 0.001318,
    "ema_tokens": 397644859
  },
  "wd_horizon": null,
  "save_steps": [],
  "seed": null,
  "arm": null,
  "device_type": "cuda",
  "frozen_at": "2026-10-11T01:36:55"
}