{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.09171156712140319,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.7825,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7565,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7408,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7496,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7424,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7112,
      "step": 60
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7409,
      "step": 70
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7646,
      "step": 80
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7129,
      "step": 90
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7671,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7166,
      "step": 110
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7113,
      "step": 120
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.6682,
      "step": 130
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7644,
      "step": 140
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6813,
      "step": 150
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.6447,
      "step": 160
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6587,
      "step": 170
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.6657,
      "step": 180
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.681,
      "step": 190
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7142,
      "step": 200
    },
    {
      "epoch": 0.09,
      "eval_loss": 0.6789492964744568,
      "eval_runtime": 280.115,
      "eval_samples_per_second": 3.57,
      "eval_steps_per_second": 0.892,
      "step": 200
    },
    {
      "epoch": 0.09,
      "mmlu_eval_accuracy": 0.4601645000494307,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.2727272727272727,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.6,
      "mmlu_eval_accuracy_high_school_biology": 0.34375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3488372093023256,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.75,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.5909090909090909,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.5454545454545454,
      "mmlu_eval_accuracy_marketing": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.42028985507246375,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.5909090909090909,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 0.9642877595465115,
      "step": 200
    }
  ],
  "max_steps": 5000,
  "num_train_epochs": 3,
  "total_flos": 5.724675843814195e+16,
  "trial_name": null,
  "trial_params": null
}