{
  "best_metric": 0.41632866859436035,
  "best_model_checkpoint": "./output_v2/7b_cluster014_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_014/checkpoint-1000",
  "epoch": 2.8306059265811587,
  "global_step": 1600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.6451,
      "step": 10
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.5699,
      "step": 20
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.5073,
      "step": 30
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.4662,
      "step": 40
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4545,
      "step": 50
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4675,
      "step": 60
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4524,
      "step": 70
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4799,
      "step": 80
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.5122,
      "step": 90
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.461,
      "step": 100
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.4393,
      "step": 110
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.4981,
      "step": 120
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4686,
      "step": 130
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.469,
      "step": 140
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4926,
      "step": 150
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4213,
      "step": 160
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.4412,
      "step": 170
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.4607,
      "step": 180
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.4537,
      "step": 190
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.4358,
      "step": 200
    },
    {
      "epoch": 0.35,
      "eval_loss": 0.45071524381637573,
      "eval_runtime": 191.6209,
      "eval_samples_per_second": 5.219,
      "eval_steps_per_second": 2.609,
      "step": 200
    },
    {
      "epoch": 0.35,
      "mmlu_eval_accuracy": 0.4662069900433653,
      "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091,
      "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": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.6,
      "mmlu_eval_accuracy_high_school_biology": 0.375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.2727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3103448275862069,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "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": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "mmlu_eval_accuracy_prehistory": 0.5428571428571428,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.3333333333333333,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.2129778240663887,
      "step": 200
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.4516,
      "step": 210
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.428,
      "step": 220
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.4268,
      "step": 230
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.4796,
      "step": 240
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.4645,
      "step": 250
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.437,
      "step": 260
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.4531,
      "step": 270
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.4158,
      "step": 280
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.4918,
      "step": 290
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.4283,
      "step": 300
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.4114,
      "step": 310
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.4429,
      "step": 320
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.4476,
      "step": 330
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.4156,
      "step": 340
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.4497,
      "step": 350
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.4355,
      "step": 360
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.4493,
      "step": 370
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.4425,
      "step": 380
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.4114,
      "step": 390
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.4567,
      "step": 400
    },
    {
      "epoch": 0.71,
      "eval_loss": 0.4355984628200531,
      "eval_runtime": 174.9677,
      "eval_samples_per_second": 5.715,
      "eval_steps_per_second": 2.858,
      "step": 400
    },
    {
      "epoch": 0.71,
      "mmlu_eval_accuracy": 0.458230355534593,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.3793103448275862,
      "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": 0.3181818181818182,
      "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": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.34375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "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": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.6923076923076923,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
      "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": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.34782608695652173,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "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": 1.1960482836625594,
      "step": 400
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.4463,
      "step": 410
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.4467,
      "step": 420
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.422,
      "step": 430
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.4487,
      "step": 440
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4473,
      "step": 450
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.453,
      "step": 460
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.4064,
      "step": 470
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.4487,
      "step": 480
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.4366,
      "step": 490
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.4244,
      "step": 500
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.4112,
      "step": 510
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.4751,
      "step": 520
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.4209,
      "step": 530
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.3893,
      "step": 540
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.4137,
      "step": 550
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.4384,
      "step": 560
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.4461,
      "step": 570
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.423,
      "step": 580
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.3713,
      "step": 590
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.3571,
      "step": 600
    },
    {
      "epoch": 1.06,
      "eval_loss": 0.4240914583206177,
      "eval_runtime": 144.1354,
      "eval_samples_per_second": 6.938,
      "eval_steps_per_second": 3.469,
      "step": 600
    },
    {
      "epoch": 1.06,
      "mmlu_eval_accuracy": 0.4612172240268643,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.41379310344827586,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.0,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.2727272727272727,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "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": 0.3793103448275862,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.5454545454545454,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.4117647058823529,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.4057971014492754,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.2888787748296953,
      "step": 600
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.3828,
      "step": 610
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.3616,
      "step": 620
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.3479,
      "step": 630
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.3916,
      "step": 640
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.3856,
      "step": 650
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.3858,
      "step": 660
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.414,
      "step": 670
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.3672,
      "step": 680
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.4017,
      "step": 690
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.3742,
      "step": 700
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.3476,
      "step": 710
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.3673,
      "step": 720
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.3516,
      "step": 730
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.4049,
      "step": 740
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.3387,
      "step": 750
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.3401,
      "step": 760
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.3464,
      "step": 770
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.3745,
      "step": 780
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.3799,
      "step": 790
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.3845,
      "step": 800
    },
    {
      "epoch": 1.42,
      "eval_loss": 0.42181774973869324,
      "eval_runtime": 144.8192,
      "eval_samples_per_second": 6.905,
      "eval_steps_per_second": 3.453,
      "step": 800
    },
    {
      "epoch": 1.42,
      "mmlu_eval_accuracy": 0.46493289206525323,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.41379310344827586,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.0,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.3588235294117647,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.4057971014492754,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0728023134502667,
      "step": 800
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.3813,
      "step": 810
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.3932,
      "step": 820
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.3392,
      "step": 830
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.4011,
      "step": 840
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.4297,
      "step": 850
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.0002,
      "loss": 0.3765,
      "step": 860
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.4088,
      "step": 870
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.3801,
      "step": 880
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.3758,
      "step": 890
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.3849,
      "step": 900
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.3891,
      "step": 910
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.3816,
      "step": 920
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.3729,
      "step": 930
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.0002,
      "loss": 0.3546,
      "step": 940
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.3721,
      "step": 950
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.348,
      "step": 960
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.3976,
      "step": 970
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.3845,
      "step": 980
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.0002,
      "loss": 0.391,
      "step": 990
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002,
      "loss": 0.3798,
      "step": 1000
    },
    {
      "epoch": 1.77,
      "eval_loss": 0.41632866859436035,
      "eval_runtime": 146.4332,
      "eval_samples_per_second": 6.829,
      "eval_steps_per_second": 3.415,
      "step": 1000
    },
    {
      "epoch": 1.77,
      "mmlu_eval_accuracy": 0.46639585350385643,
      "mmlu_eval_accuracy_abstract_algebra": 0.5454545454545454,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.375,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.2727272727272727,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.46153846153846156,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.75,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.42028985507246375,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.3059391756359653,
      "step": 1000
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.0002,
      "loss": 0.3785,
      "step": 1010
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.0002,
      "loss": 0.3682,
      "step": 1020
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.0002,
      "loss": 0.3561,
      "step": 1030
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0002,
      "loss": 0.3796,
      "step": 1040
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.0002,
      "loss": 0.3552,
      "step": 1050
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.0002,
      "loss": 0.3337,
      "step": 1060
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0002,
      "loss": 0.3629,
      "step": 1070
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002,
      "loss": 0.4007,
      "step": 1080
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.0002,
      "loss": 0.3408,
      "step": 1090
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.0002,
      "loss": 0.3738,
      "step": 1100
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.0002,
      "loss": 0.3835,
      "step": 1110
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0002,
      "loss": 0.394,
      "step": 1120
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 0.3833,
      "step": 1130
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.0002,
      "loss": 0.3088,
      "step": 1140
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.0002,
      "loss": 0.3275,
      "step": 1150
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002,
      "loss": 0.2665,
      "step": 1160
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.0002,
      "loss": 0.3261,
      "step": 1170
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.0002,
      "loss": 0.2848,
      "step": 1180
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.0002,
      "loss": 0.3344,
      "step": 1190
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.0002,
      "loss": 0.3102,
      "step": 1200
    },
    {
      "epoch": 2.12,
      "eval_loss": 0.4272211790084839,
      "eval_runtime": 149.2515,
      "eval_samples_per_second": 6.7,
      "eval_steps_per_second": 3.35,
      "step": 1200
    },
    {
      "epoch": 2.12,
      "mmlu_eval_accuracy": 0.4741182631719215,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4827586206896552,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.5454545454545454,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4186046511627907,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.43478260869565216,
      "mmlu_eval_accuracy_high_school_us_history": 0.6363636363636364,
      "mmlu_eval_accuracy_high_school_world_history": 0.5,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.64,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.22580645161290322,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.4057971014492754,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.3267292114212992,
      "step": 1200
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.0002,
      "loss": 0.3279,
      "step": 1210
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.0002,
      "loss": 0.3034,
      "step": 1220
    },
    {
      "epoch": 2.18,
      "learning_rate": 0.0002,
      "loss": 0.2832,
      "step": 1230
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.0002,
      "loss": 0.3297,
      "step": 1240
    },
    {
      "epoch": 2.21,
      "learning_rate": 0.0002,
      "loss": 0.2971,
      "step": 1250
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.0002,
      "loss": 0.3342,
      "step": 1260
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0002,
      "loss": 0.3061,
      "step": 1270
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.0002,
      "loss": 0.3044,
      "step": 1280
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0002,
      "loss": 0.2828,
      "step": 1290
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0002,
      "loss": 0.3011,
      "step": 1300
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.0002,
      "loss": 0.285,
      "step": 1310
    },
    {
      "epoch": 2.34,
      "learning_rate": 0.0002,
      "loss": 0.2897,
      "step": 1320
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.0002,
      "loss": 0.2908,
      "step": 1330
    },
    {
      "epoch": 2.37,
      "learning_rate": 0.0002,
      "loss": 0.3331,
      "step": 1340
    },
    {
      "epoch": 2.39,
      "learning_rate": 0.0002,
      "loss": 0.3105,
      "step": 1350
    },
    {
      "epoch": 2.41,
      "learning_rate": 0.0002,
      "loss": 0.321,
      "step": 1360
    },
    {
      "epoch": 2.42,
      "learning_rate": 0.0002,
      "loss": 0.3539,
      "step": 1370
    },
    {
      "epoch": 2.44,
      "learning_rate": 0.0002,
      "loss": 0.3114,
      "step": 1380
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.0002,
      "loss": 0.2943,
      "step": 1390
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.0002,
      "loss": 0.3232,
      "step": 1400
    },
    {
      "epoch": 2.48,
      "eval_loss": 0.4280230402946472,
      "eval_runtime": 147.2124,
      "eval_samples_per_second": 6.793,
      "eval_steps_per_second": 3.396,
      "step": 1400
    },
    {
      "epoch": 2.48,
      "mmlu_eval_accuracy": 0.46287474793754874,
      "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": 0.41379310344827586,
      "mmlu_eval_accuracy_college_biology": 0.375,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.35714285714285715,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.40625,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5238095238095238,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.4418604651162791,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "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": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.6923076923076923,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.391304347826087,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.2063124309593019,
      "step": 1400
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.0002,
      "loss": 0.3221,
      "step": 1410
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.0002,
      "loss": 0.303,
      "step": 1420
    },
    {
      "epoch": 2.53,
      "learning_rate": 0.0002,
      "loss": 0.2884,
      "step": 1430
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.0002,
      "loss": 0.291,
      "step": 1440
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.0002,
      "loss": 0.3169,
      "step": 1450
    },
    {
      "epoch": 2.58,
      "learning_rate": 0.0002,
      "loss": 0.301,
      "step": 1460
    },
    {
      "epoch": 2.6,
      "learning_rate": 0.0002,
      "loss": 0.3229,
      "step": 1470
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002,
      "loss": 0.312,
      "step": 1480
    },
    {
      "epoch": 2.64,
      "learning_rate": 0.0002,
      "loss": 0.3365,
      "step": 1490
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.0002,
      "loss": 0.2919,
      "step": 1500
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0002,
      "loss": 0.3054,
      "step": 1510
    },
    {
      "epoch": 2.69,
      "learning_rate": 0.0002,
      "loss": 0.3009,
      "step": 1520
    },
    {
      "epoch": 2.71,
      "learning_rate": 0.0002,
      "loss": 0.2621,
      "step": 1530
    },
    {
      "epoch": 2.72,
      "learning_rate": 0.0002,
      "loss": 0.2999,
      "step": 1540
    },
    {
      "epoch": 2.74,
      "learning_rate": 0.0002,
      "loss": 0.3183,
      "step": 1550
    },
    {
      "epoch": 2.76,
      "learning_rate": 0.0002,
      "loss": 0.2864,
      "step": 1560
    },
    {
      "epoch": 2.78,
      "learning_rate": 0.0002,
      "loss": 0.2863,
      "step": 1570
    },
    {
      "epoch": 2.8,
      "learning_rate": 0.0002,
      "loss": 0.3426,
      "step": 1580
    },
    {
      "epoch": 2.81,
      "learning_rate": 0.0002,
      "loss": 0.3245,
      "step": 1590
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.0002,
      "loss": 0.3233,
      "step": 1600
    },
    {
      "epoch": 2.83,
      "eval_loss": 0.4239733815193176,
      "eval_runtime": 145.7087,
      "eval_samples_per_second": 6.863,
      "eval_steps_per_second": 3.432,
      "step": 1600
    },
    {
      "epoch": 2.83,
      "mmlu_eval_accuracy": 0.4674582653302792,
      "mmlu_eval_accuracy_abstract_algebra": 0.45454545454545453,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "mmlu_eval_accuracy_college_mathematics": 0.45454545454545453,
      "mmlu_eval_accuracy_college_medicine": 0.4090909090909091,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.24390243902439024,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.3,
      "mmlu_eval_accuracy_high_school_biology": 0.53125,
      "mmlu_eval_accuracy_high_school_chemistry": 0.22727272727272727,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.47619047619047616,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3953488372093023,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.4117647058823529,
      "mmlu_eval_accuracy_high_school_psychology": 0.75,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.6086956521739131,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.6363636363636364,
      "mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
      "mmlu_eval_accuracy_moral_disputes": 0.5263157894736842,
      "mmlu_eval_accuracy_moral_scenarios": 0.25,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.37681159420289856,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.631578947368421,
      "mmlu_loss": 1.322269008302813,
      "step": 1600
    }
  ],
  "max_steps": 5000,
  "num_train_epochs": 9,
  "total_flos": 2.32154379234902e+17,
  "trial_name": null,
  "trial_params": null
}