{
  "cfg": {
    "name": "meter_mac_v2",
    "seed": 0,
    "ckpt": "best.pt",
    "out_dir": "meter_mac_v2",
    "encoded": "encoded",
    "attn": "sdpa",
    "T": 2048,
    "micro_batch": 16,
    "eval_micro": 16,
    "lr_enc": 5e-05,
    "lr_head": 0.0008,
    "wd": 0.01,
    "clip": 1.0,
    "warmup_frac": 0.1,
    "epochs": 12,
    "patience": 12,
    "head_dropout": 0.33,
    "use_cap": true,
    "scalar_mix": true,
    "w_mac": 1.0,
    "w_scan": 0.0,
    "select": "mac",
    "prose_per_epoch": 150000,
    "scan_passes": 2,
    "scan_max_verses": 8,
    "mac_class_w": "auto",
    "scan_class_w": "auto",
    "log_every": 50
  },
  "mcfg": {
    "head_dropout": 0.33,
    "w_mac": 1.0,
    "w_scan": 0.0,
    "use_cap": true,
    "scalar_mix": true,
    "mac_class_w": [
      3.935332098208277,
      0.5727731680236648
    ],
    "scan_class_w": [
      1.0,
      1.0,
      1.0,
      1.0
    ]
  },
  "pretrain_cfg": {
    "name": "greekcharbert_anneal_staged",
    "seed": 0,
    "d_model": 1024,
    "depth": 32,
    "char_window": 256,
    "attn": "flex",
    "compile": true,
    "qk_norm": true,
    "seq_len": 8192,
    "micro_batch": 4,
    "grad_accum": 2,
    "lr": 0.0003,
    "wd": 0.1,
    "clip": 1.0,
    "warmup_frac": 0.04,
    "decay_frac": 0.1,
    "tier_weights": [
      1.0,
      1.0,
      0.3
    ],
    "anneal_phases": [
      [
        0.5,
        [
          1.0,
          1.0,
          0.3
        ]
      ],
      [
        0.9,
        [
          1.0,
          1.0,
          0.0
        ]
      ],
      [
        1.0,
        [
          1.0,
          0.0,
          0.0
        ]
      ]
    ],
    "anneal_patience": 3,
    "lam": 0.1,
    "total_steps": 120000,
    "num_workers": 12,
    "log_every": 20,
    "ckpt_every": 2000,
    "eval_every_ckpt": true,
    "eval_n": 256,
    "out_dir": "greekcharbert_anneal_staged"
  },
  "epoch": 4,
  "dev": {
    "mac_bal": 0.8886,
    "mac_acc": 0.9303,
    "mac_n": 2498
  },
  "T": 2048,
  "_source_checkpoint": "best.pt"
}