{
  "best_metric": 0.4203888475894928,
  "best_model_checkpoint": "./output_v2/7b_cluster08_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_08/checkpoint-1000",
  "epoch": 1.7306652244456462,
  "global_step": 1600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.5043,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.4612,
      "step": 20
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.4969,
      "step": 30
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.4995,
      "step": 40
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.4247,
      "step": 50
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.4557,
      "step": 60
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.5582,
      "step": 70
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.4484,
      "step": 80
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.438,
      "step": 90
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.4528,
      "step": 100
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.4848,
      "step": 110
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.4192,
      "step": 120
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.4481,
      "step": 130
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.438,
      "step": 140
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.4484,
      "step": 150
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.4156,
      "step": 160
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.3978,
      "step": 170
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.5086,
      "step": 180
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.5037,
      "step": 190
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.4201,
      "step": 200
    },
    {
      "epoch": 0.22,
      "eval_loss": 0.44470372796058655,
      "eval_runtime": 145.3872,
      "eval_samples_per_second": 6.878,
      "eval_steps_per_second": 3.439,
      "step": 200
    },
    {
      "epoch": 0.22,
      "mmlu_eval_accuracy": 0.4603515989210557,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.3181818181818182,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.2727272727272727,
      "mmlu_eval_accuracy_conceptual_physics": 0.5,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "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": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.5454545454545454,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.5151515151515151,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.37681159420289856,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.2632288357763315,
      "step": 200
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.4912,
      "step": 210
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.4233,
      "step": 220
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.4275,
      "step": 230
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.4223,
      "step": 240
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.4336,
      "step": 250
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.4466,
      "step": 260
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.474,
      "step": 270
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.4583,
      "step": 280
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.4914,
      "step": 290
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.438,
      "step": 300
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.4737,
      "step": 310
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.4109,
      "step": 320
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.4382,
      "step": 330
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.4327,
      "step": 340
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.426,
      "step": 350
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.438,
      "step": 360
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.3864,
      "step": 370
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.4809,
      "step": 380
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.4108,
      "step": 390
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.4188,
      "step": 400
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.4346640706062317,
      "eval_runtime": 148.072,
      "eval_samples_per_second": 6.753,
      "eval_steps_per_second": 3.377,
      "step": 400
    },
    {
      "epoch": 0.43,
      "mmlu_eval_accuracy": 0.4649911940484824,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.2727272727272727,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.18181818181818182,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "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": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3488372093023256,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7,
      "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": 0.3333333333333333,
      "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": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.36231884057971014,
      "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": 0.6842105263157895,
      "mmlu_loss": 1.1255856356911809,
      "step": 400
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.4742,
      "step": 410
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.3976,
      "step": 420
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.4379,
      "step": 430
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.4952,
      "step": 440
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.3877,
      "step": 450
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.4486,
      "step": 460
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.4336,
      "step": 470
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.4962,
      "step": 480
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.4339,
      "step": 490
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.4264,
      "step": 500
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.4082,
      "step": 510
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.5009,
      "step": 520
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.425,
      "step": 530
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.4571,
      "step": 540
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.4694,
      "step": 550
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.4323,
      "step": 560
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.3936,
      "step": 570
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.457,
      "step": 580
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.4735,
      "step": 590
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.4292,
      "step": 600
    },
    {
      "epoch": 0.65,
      "eval_loss": 0.4281369745731354,
      "eval_runtime": 145.9924,
      "eval_samples_per_second": 6.85,
      "eval_steps_per_second": 3.425,
      "step": 600
    },
    {
      "epoch": 0.65,
      "mmlu_eval_accuracy": 0.4533110503935826,
      "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.3181818181818182,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.18181818181818182,
      "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": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.5714285714285714,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.3076923076923077,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "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": 0.5,
      "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": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "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": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.5909090909090909,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1304532200214756,
      "step": 600
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.4535,
      "step": 610
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.4298,
      "step": 620
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.4762,
      "step": 630
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.4094,
      "step": 640
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.4249,
      "step": 650
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.4532,
      "step": 660
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.3749,
      "step": 670
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.4204,
      "step": 680
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.3707,
      "step": 690
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.4761,
      "step": 700
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.3654,
      "step": 710
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.4196,
      "step": 720
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.4136,
      "step": 730
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.4185,
      "step": 740
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.3943,
      "step": 750
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.4549,
      "step": 760
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.4459,
      "step": 770
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.3884,
      "step": 780
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.4566,
      "step": 790
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.4407,
      "step": 800
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.4228321611881256,
      "eval_runtime": 145.8899,
      "eval_samples_per_second": 6.854,
      "eval_steps_per_second": 3.427,
      "step": 800
    },
    {
      "epoch": 0.87,
      "mmlu_eval_accuracy": 0.459709154819756,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "mmlu_eval_accuracy_college_biology": 0.375,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.2727272727272727,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.18181818181818182,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.34375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.6833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "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": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5,
      "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": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
      "mmlu_eval_accuracy_moral_scenarios": 0.25,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0534737041079034,
      "step": 800
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.3953,
      "step": 810
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.3551,
      "step": 820
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.3915,
      "step": 830
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.3444,
      "step": 840
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.4325,
      "step": 850
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.4298,
      "step": 860
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.4051,
      "step": 870
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.3934,
      "step": 880
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.4189,
      "step": 890
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.4441,
      "step": 900
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.4313,
      "step": 910
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.4491,
      "step": 920
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.3924,
      "step": 930
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.4027,
      "step": 940
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.3962,
      "step": 950
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.3919,
      "step": 960
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.2931,
      "step": 970
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.5071,
      "step": 980
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.3284,
      "step": 990
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.359,
      "step": 1000
    },
    {
      "epoch": 1.08,
      "eval_loss": 0.4203888475894928,
      "eval_runtime": 150.4231,
      "eval_samples_per_second": 6.648,
      "eval_steps_per_second": 3.324,
      "step": 1000
    },
    {
      "epoch": 1.08,
      "mmlu_eval_accuracy": 0.46832755229244377,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.41379310344827586,
      "mmlu_eval_accuracy_college_biology": 0.4375,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.3181818181818182,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.18181818181818182,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "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": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "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": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "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": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.37142857142857144,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "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": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0159204223973788,
      "step": 1000
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.4234,
      "step": 1010
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.4452,
      "step": 1020
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.3596,
      "step": 1030
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.3379,
      "step": 1040
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.4443,
      "step": 1050
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.3825,
      "step": 1060
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.3872,
      "step": 1070
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.4029,
      "step": 1080
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.3287,
      "step": 1090
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.3646,
      "step": 1100
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.3707,
      "step": 1110
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.3713,
      "step": 1120
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.3834,
      "step": 1130
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.4071,
      "step": 1140
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.3694,
      "step": 1150
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.4209,
      "step": 1160
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.3257,
      "step": 1170
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.3688,
      "step": 1180
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.396,
      "step": 1190
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.3519,
      "step": 1200
    },
    {
      "epoch": 1.3,
      "eval_loss": 0.4303589463233948,
      "eval_runtime": 147.9243,
      "eval_samples_per_second": 6.76,
      "eval_steps_per_second": 3.38,
      "step": 1200
    },
    {
      "epoch": 1.3,
      "mmlu_eval_accuracy": 0.4575244244905048,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "mmlu_eval_accuracy_business_ethics": 0.5454545454545454,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.2727272727272727,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "mmlu_eval_accuracy_computer_security": 0.36363636363636365,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "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": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.8181818181818182,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.2608695652173913,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.46153846153846156,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5,
      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.36363636363636365,
      "mmlu_eval_accuracy_marketing": 0.72,
      "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": 0.28,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "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": 0.37681159420289856,
      "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": 0.6842105263157895,
      "mmlu_loss": 1.007975729654103,
      "step": 1200
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.3172,
      "step": 1210
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.4015,
      "step": 1220
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.3605,
      "step": 1230
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.3806,
      "step": 1240
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.0002,
      "loss": 0.387,
      "step": 1250
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0002,
      "loss": 0.3597,
      "step": 1260
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0002,
      "loss": 0.3484,
      "step": 1270
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.0002,
      "loss": 0.3474,
      "step": 1280
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.0002,
      "loss": 0.3546,
      "step": 1290
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.0002,
      "loss": 0.3973,
      "step": 1300
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002,
      "loss": 0.3887,
      "step": 1310
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002,
      "loss": 0.3583,
      "step": 1320
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.0002,
      "loss": 0.4175,
      "step": 1330
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0002,
      "loss": 0.3729,
      "step": 1340
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.0002,
      "loss": 0.3922,
      "step": 1350
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.0002,
      "loss": 0.4228,
      "step": 1360
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0002,
      "loss": 0.4216,
      "step": 1370
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002,
      "loss": 0.3686,
      "step": 1380
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002,
      "loss": 0.2974,
      "step": 1390
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.0002,
      "loss": 0.367,
      "step": 1400
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.4240153729915619,
      "eval_runtime": 153.5179,
      "eval_samples_per_second": 6.514,
      "eval_steps_per_second": 3.257,
      "step": 1400
    },
    {
      "epoch": 1.51,
      "mmlu_eval_accuracy": 0.464896484980687,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "mmlu_eval_accuracy_business_ethics": 0.6363636363636364,
      "mmlu_eval_accuracy_clinical_knowledge": 0.4482758620689655,
      "mmlu_eval_accuracy_college_biology": 0.375,
      "mmlu_eval_accuracy_college_chemistry": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.3181818181818182,
      "mmlu_eval_accuracy_college_physics": 0.36363636363636365,
      "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": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.2857142857142857,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.34375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.45454545454545453,
      "mmlu_eval_accuracy_high_school_computer_science": 0.7777777777777778,
      "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": 0.3023255813953488,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.6833333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.46153846153846156,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.5454545454545454,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6511627906976745,
      "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
      "mmlu_eval_accuracy_moral_scenarios": 0.26,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.4,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "mmlu_eval_accuracy_professional_psychology": 0.36231884057971014,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 0.9802373749721922,
      "step": 1400
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0002,
      "loss": 0.3926,
      "step": 1410
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0002,
      "loss": 0.3492,
      "step": 1420
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0002,
      "loss": 0.4083,
      "step": 1430
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0002,
      "loss": 0.421,
      "step": 1440
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0002,
      "loss": 0.3333,
      "step": 1450
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.0002,
      "loss": 0.3925,
      "step": 1460
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0002,
      "loss": 0.3849,
      "step": 1470
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.0002,
      "loss": 0.3789,
      "step": 1480
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0002,
      "loss": 0.3504,
      "step": 1490
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.0002,
      "loss": 0.3615,
      "step": 1500
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.0002,
      "loss": 0.4198,
      "step": 1510
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002,
      "loss": 0.3257,
      "step": 1520
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.0002,
      "loss": 0.4162,
      "step": 1530
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0002,
      "loss": 0.3853,
      "step": 1540
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.0002,
      "loss": 0.3603,
      "step": 1550
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0002,
      "loss": 0.3868,
      "step": 1560
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002,
      "loss": 0.3895,
      "step": 1570
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.0002,
      "loss": 0.3476,
      "step": 1580
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.0002,
      "loss": 0.3791,
      "step": 1590
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.0002,
      "loss": 0.3352,
      "step": 1600
    },
    {
      "epoch": 1.73,
      "eval_loss": 0.42534786462783813,
      "eval_runtime": 148.5639,
      "eval_samples_per_second": 6.731,
      "eval_steps_per_second": 3.366,
      "step": 1600
    },
    {
      "epoch": 1.73,
      "mmlu_eval_accuracy": 0.46498642485781333,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5,
      "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": 0.125,
      "mmlu_eval_accuracy_college_computer_science": 0.2727272727272727,
      "mmlu_eval_accuracy_college_mathematics": 0.36363636363636365,
      "mmlu_eval_accuracy_college_medicine": 0.45454545454545453,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.45454545454545453,
      "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": 0.4146341463414634,
      "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": 0.4090909090909091,
      "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": 0.3488372093023256,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.5384615384615384,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7,
      "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": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.3333333333333333,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "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": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.26,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "mmlu_eval_accuracy_professional_psychology": 0.391304347826087,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0468192458658554,
      "step": 1600
    }
  ],
  "max_steps": 5000,
  "num_train_epochs": 6,
  "total_flos": 2.2913517749885338e+17,
  "trial_name": null,
  "trial_params": null
}