{
  "id": "month-20261028",
  "title": "Learning from the full September corpus",
  "date": "2026-10-28",
  "kind": "training",
  "status": "completed",
  "hypothesis": "Scaling the same model to the retained monthly corpus will improve held-out next-token prediction within a fixed compute budget.",
  "baseline": "Random initialization validation NLL 5.029830; pilot is contextual, not a comparable validation set.",
  "data": "684,367 retained training games; 155,948,725 prediction targets per pass. Validation: 7,000 games / 1,595,704 targets. 305,193,023 repeated target exposures (1.957 target-pass equivalents), not unique data.",
  "config": {
    "accumulation": 2,
    "dataset_fingerprint": "e59e843f726f5ebcb54cd353f76a4e62c216829fa5085b830f9a534a9e9b441f",
    "epochs_schedule": 2,
    "lr": 0.0003,
    "max_updates": 41854,
    "microbatch": 8,
    "model": {
      "attn_implementation": "sdpa",
      "attn_pdrop": 0,
      "bos_token_id": 69,
      "embd_pdrop": 0,
      "eos_token_id": 70,
      "n_ctx": 1024,
      "n_embd": 768,
      "n_head": 12,
      "n_layer": 12,
      "n_positions": 1024,
      "pad_token_id": 71,
      "resid_pdrop": 0,
      "use_cache": false,
      "vocab_size": 128
    },
    "optimizer": "FP32 fused AdamW betas .9/.95 matrix wd .1 clip1",
    "parameters": 85940736,
    "precision": "FP32 weights/gradients/optimizer; BF16 autocast SDPA",
    "sampling": "seeded length buckets, disjoint full games; right pad; zero-target dummy rank on partial global batches",
    "schedule_steps": 41854,
    "seed": 2026,
    "tokenizer": {
      "sha256": "0db7085edec22885bd8c51aac4d6867afc04203ba913ef561be97213df019029",
      "version": "chess-square-triple-v1"
    },
    "torch": "2.9.0+cu128",
    "training_code": {
      "train_chess.py": "8147238e3684c68343c80a365d3c13f6d061e921b50fe9bcb28d75a18423b279",
      "training_core.py": "140022ab400e6ef5f56fe4c3a1df28834afdc7f90e39b011968418a93795b856",
      "training_deadline.py": "41aa6a01d17823f24d1990d181b5402c16aed96848569b30210d5bffc6d2e8dd",
      "verified_training_data.py": "ad3952fe45d8721d05dbe80d6f077e69a4851a3fae7ee5114def2e33ccc57037"
    },
    "warmup_steps": 2093,
    "world_size": 2
  },
  "compute": "Two RTX 4090 GPUs. 41,854 updates in 86.01 minutes total job wall time; 63,058 useful targets/s during optimizer work.",
  "results": [
    {
      "label": "Best validation NLL",
      "value": 0.6058686751172994
    },
    {
      "label": "Validation games",
      "value": 7000
    },
    {
      "label": "Teacher-forced triple legality (128-game subset)",
      "value": 0.6463697762769101
    },
    {
      "label": "Raw illegal-token probability",
      "value": 0.009694814930694023
    }
  ],
  "limitations": [
    "Test set and free-running match strength unmeasured in this run.",
    "Teacher-forced legality uses 128 fixed validation games, not free-running rollout.",
    "Full BF16 DDP exact resume equivalence remains untested; this run did not resume.",
    "Single seed/configuration; not a hyperparameter optimum or production-readiness claim.",
    "Same-month splits share players/opening positions; exact full-game ID/sequence duplicates are grouped.",
    "Promotion accuracy is dominated by no-promotion tokens; do not interpret it as promotion strength.",
    "The later Sunfish match uses a legal-move filter plus a mate helper and is a separate evaluation."
  ],
  "checkpoint": "6133648b2926b38542ee3ee1251a1ba111d61da9a8ea69d1c5f0559b09933643",
  "sourceSnapshot": "month-20261028",
  "history": [
    {
      "step": 5300,
      "nll": 0.8899340847576866
    },
    {
      "step": 10600,
      "nll": 0.7769294259080382
    },
    {
      "step": 15900,
      "nll": 0.7174767785439452
    },
    {
      "step": 21200,
      "nll": 0.6790902393498442
    },
    {
      "step": 21387,
      "nll": 0.6762016430910819
    },
    {
      "step": 26500,
      "nll": 0.6517212626194913
    },
    {
      "step": 31800,
      "nll": 0.6303233843590866
    },
    {
      "step": 37100,
      "nll": 0.6128119806352297
    },
    {
      "step": 41854,
      "nll": 0.6058686751172994
    }
  ],
  "evidence": [
    {
      "label": "completion-summary.json",
      "sha256": "2d1f69d863327fcfa150ec9e0e30409a8ffe277ce5b4c8dfd4bd85045ac94747",
      "selectedFields": {
        "best_validation_nll": 0.6058686751172994,
        "final_validation": {
          "games": 7000,
          "legality_games": 128,
          "legality_subset": "N evenly spaced game indices across immutable validation shard order; N=None means all",
          "legality_targets": 28556,
          "nll": 0.6058686751172994,
          "phase_accuracy": {
            "destination": 0.7693100791588615,
            "eos": 0.111,
            "promotion": 0.999884811771104,
            "source": 0.5690109674300562
          },
          "phase_counts": {
            "destination": 529568,
            "eos": 7000,
            "promotion": 529568,
            "source": 529568
          },
          "protocol": "raw128 teacher-forced next-token logits; triple destination/promotion conditioned on gold prefix; not free-running",
          "raw_illegal_token_probability": 0.009694814930694023,
          "raw_reserved_probability": 0.0000015190403278775776,
          "seconds": 19.11114326200004,
          "teacher_forced_triple_legality": 0.6463697762769101,
          "valid_targets": 1595704
        },
        "test_evaluated": false,
        "total_wall_seconds_from_guardian_receipt": 5160.336741685867,
        "training_prediction_target_exposures": 305193023,
        "updates": 41854
      }
    }
  ]
}