{
  "model_id": "MatterWaveAI/grt-30m",
  "architecture": "GRTForCausalLM",
  "stored_parameters": 29305498,
  "non_embedding_parameters": 9740197,
  "dense_equiv_parameters": 55680446,
  "nn_linear_layers": 0,
  "pretraining": {
    "teacher_family": "Qwen 3 (151,669-token BPE)",
    "tokens_trained": 10789404672,
    "steps": 20580,
    "optimizer": "PairAdam",
    "val_loss": 4.4776,
    "val_ppl": 88.02,
    "val_top1_acc": 0.272,
    "wikitext2_test_ppl": 153.01,
    "wikitext2_top1_acc": 0.2571
  },
  "slm_bench": {
    "benchmark_suite": "SLM-Bench (10-Task Core Suite)",
    "evaluation_harness": "grt-slm-bench (training/slm_bench.py)",
    "hardware": "AMD Ryzen AI Max+ 395 (Radeon 8060S / ROCm gfx1151)",
    "macro_average_accuracy": {
      "random_baseline": 0.345,
      "zero_shot": 0.3723,
      "pair_adam": 0.3957,
      "symplectic_adam": 0.4768,
      "huygens_symplectic_adam": 0.4934
    },
    "tasks": {
      "boolq": {
        "split": "validation",
        "n_eval": 3270,
        "random_baseline": 0.5,
        "zero_shot_acc": 0.5141,
        "pair_adam_acc": 0.6055,
        "symplectic_adam_acc": 0.6125,
        "huygens_symplectic_adam_acc": 0.6223
      },
      "arc_easy": {
        "split": "test",
        "n_eval": 2376,
        "random_baseline": 0.25,
        "zero_shot_acc": 0.2959,
        "pair_adam_acc": 0.2896,
        "symplectic_adam_acc": 0.359,
        "huygens_symplectic_adam_acc": 0.3632
      },
      "arc_challenge": {
        "split": "test",
        "n_eval": 1172,
        "random_baseline": 0.25,
        "zero_shot_acc": 0.2082,
        "pair_adam_acc": 0.2355,
        "symplectic_adam_acc": 0.2568,
        "huygens_symplectic_adam_acc": 0.2568
      },
      "openbookqa": {
        "split": "test",
        "n_eval": 500,
        "random_baseline": 0.25,
        "zero_shot_acc": 0.282,
        "pair_adam_acc": 0.3,
        "symplectic_adam_acc": 0.304,
        "huygens_symplectic_adam_acc": 0.358
      },
      "piqa": {
        "split": "validation",
        "n_eval": 1838,
        "random_baseline": 0.5,
        "zero_shot_acc": 0.5528,
        "pair_adam_acc": 0.5408,
        "symplectic_adam_acc": 0.5528,
        "huygens_symplectic_adam_acc": 0.5528
      },
      "winogrande": {
        "split": "validation",
        "n_eval": 1267,
        "random_baseline": 0.5,
        "zero_shot_acc": 0.5028,
        "pair_adam_acc": 0.5099,
        "symplectic_adam_acc": 0.5122,
        "huygens_symplectic_adam_acc": 0.5122
      },
      "hellaswag": {
        "split": "validation",
        "n_eval": 1000,
        "random_baseline": 0.25,
        "zero_shot_acc": 0.292,
        "pair_adam_acc": 0.316,
        "symplectic_adam_acc": 0.316,
        "huygens_symplectic_adam_acc": 0.316
      },
      "qnli": {
        "split": "validation",
        "n_eval": 1000,
        "random_baseline": 0.5,
        "zero_shot_acc": 0.516,
        "pair_adam_acc": 0.54,
        "symplectic_adam_acc": 0.624,
        "huygens_symplectic_adam_acc": 0.72
      },
      "commonsenseqa": {
        "split": "validation",
        "n_eval": 1221,
        "random_baseline": 0.2,
        "zero_shot_acc": 0.217,
        "pair_adam_acc": 0.2277,
        "symplectic_adam_acc": 0.2465,
        "huygens_symplectic_adam_acc": 0.2465
      },
      "gsm8k": {
        "split": "test",
        "n_eval": 500,
        "random_baseline": 0.25,
        "zero_shot_acc": 0.342,
        "zero_shot_ppl": 27.99,
        "pair_adam_acc": 0.392,
        "pair_adam_ppl": 10.65,
        "symplectic_adam_acc": 0.984,
        "symplectic_adam_ppl": 10.98,
        "huygens_symplectic_adam_acc": 0.986,
        "huygens_symplectic_adam_ppl": 9.04
      }
    }
  }
}