{
  "task": "reward_model",
  "strategy": "lora",
  "base_model": "bert-base-uncased",
  "learning_rate": 0.0001,
  "epochs": 3,
  "n_train_pairs": 4000,
  "param_stats": {
    "total_params": 112162562,
    "trainable_params": 2679553,
    "trainable_pct": 2.389,
    "total_M": 112.16,
    "trainable_M": 2.68
  },
  "build_notes": [
    "patched PeftModelForSequenceClassification.add_adapter() to tolerate low_cpu_mem_usage (PEFT prompt-learning signature mismatch)",
    "patched PeftModelForTokenClassification.add_adapter() to tolerate low_cpu_mem_usage (PEFT prompt-learning signature mismatch)",
    "patched PeftModelForQuestionAnswering.add_adapter() to tolerate low_cpu_mem_usage (PEFT prompt-learning signature mismatch)"
  ],
  "quantized": false,
  "val_history": [
    {
      "eval_loss": 0.009566658656694926,
      "eval_accuracy": 0.998,
      "eval_margin": 9.352208137512207,
      "eval_n": 500,
      "phase": "epoch_end",
      "strategy": "lora",
      "epoch": 0
    },
    {
      "eval_loss": 0.00644640389145934,
      "eval_accuracy": 0.998,
      "eval_margin": 10.501071691513062,
      "eval_n": 500,
      "phase": "epoch_end",
      "strategy": "lora",
      "epoch": 1
    },
    {
      "eval_loss": 0.00556696730927797,
      "eval_accuracy": 0.998,
      "eval_margin": 10.708506226539612,
      "eval_n": 500,
      "phase": "epoch_end",
      "strategy": "lora",
      "epoch": 2
    }
  ],
  "test_metrics": {
    "eval_loss": 0.011904432533119107,
    "eval_accuracy": 0.9973324441480493,
    "eval_margin": 10.969182755084748,
    "eval_n": 5998
  },
  "train_seconds": 387.97,
  "peak_gpu_mb": 4830.4,
  "data_source": "hf:nrizwan/safe_ai_assignment_1"
}