{
  "model_id": "MatterWaveAI/grt-10m-base",
  "architecture": "GRTForCausalLM",
  "stored_parameters": 8684515,
  "non_embedding_parameters": 2234515,
  "dense_equiv_parameters": 17518848,
  "nn_linear_layers": 0,
  "pretraining": {
    "corpus": "nvidia/Nemotron-Pretraining-Dataset-sample",
    "steps": 336000,
    "optimizer": "PairAdam",
    "train_loss": 2.144,
    "val_loss": 4.033,
    "val_ppl": 56.4,
    "val_top1_acc": 0.4
  },
  "slm_bench_zero_shot": {
    "benchmark_suite": "SLM-Bench (9 Multiple-Choice Tasks + GSM8K Perplexity)",
    "evaluation_harness": "grt-slm-bench (training/slm_bench.py)",
    "mc_macro_average_accuracy": 0.3808,
    "random_macro_average": 0.3556,
    "tasks": {
      "boolq": {
        "split": "validation",
        "n_eval": 3270,
        "random": 0.5,
        "zero_shot_acc": 0.6214,
        "pair_adam_acc": 0.6113
      },
      "qnli": {
        "split": "validation",
        "n_eval": 1000,
        "random": 0.5,
        "zero_shot_acc": 0.527,
        "pair_adam_acc": 0.497
      },
      "piqa": {
        "split": "validation",
        "n_eval": 1838,
        "random": 0.5,
        "zero_shot_acc": 0.5218,
        "pair_adam_acc": 0.5082
      },
      "winogrande": {
        "split": "validation",
        "n_eval": 1267,
        "random": 0.5,
        "zero_shot_acc": 0.5051,
        "pair_adam_acc": 0.5075
      },
      "openbookqa": {
        "split": "test",
        "n_eval": 500,
        "random": 0.25,
        "zero_shot_acc": 0.3,
        "pair_adam_acc": 0.226
      },
      "arc_easy": {
        "split": "test",
        "n_eval": 2376,
        "random": 0.25,
        "zero_shot_acc": 0.2694,
        "pair_adam_acc": 0.266
      },
      "hellaswag": {
        "split": "validation",
        "n_eval": 1000,
        "random": 0.25,
        "zero_shot_acc": 0.256,
        "pair_adam_acc": 0.291
      },
      "arc_challenge": {
        "split": "test",
        "n_eval": 1172,
        "random": 0.25,
        "zero_shot_acc": 0.2253,
        "pair_adam_acc": 0.2491
      },
      "commonsenseqa": {
        "split": "validation",
        "n_eval": 1221,
        "random": 0.2,
        "zero_shot_acc": 0.2015,
        "pair_adam_acc": 0.2285
      },
      "gsm8k": {
        "split": "test",
        "n_eval": 500,
        "zero_shot_ppl": 145.85,
        "pair_adam_ppl": 37.27
      }
    }
  }
}