{
  "best_metric": 0.8138102889060974,
  "best_model_checkpoint": "./output_v2/7b_cluster015_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_015/checkpoint-1400",
  "epoch": 1.340931947703654,
  "global_step": 2000,
  "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": 0.969,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.9032,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.81,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.876,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8858,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.8608,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.846,
      "step": 70
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.8466,
      "step": 80
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.8456,
      "step": 90
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.8895,
      "step": 100
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.862,
      "step": 110
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.8193,
      "step": 120
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8588,
      "step": 130
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.8516,
      "step": 140
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.8428,
      "step": 150
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.8829,
      "step": 160
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.882,
      "step": 170
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.8054,
      "step": 180
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8673,
      "step": 190
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.8389,
      "step": 200
    },
    {
      "epoch": 0.13,
      "eval_loss": 0.8394724130630493,
      "eval_runtime": 191.2112,
      "eval_samples_per_second": 5.23,
      "eval_steps_per_second": 2.615,
      "step": 200
    },
    {
      "epoch": 0.13,
      "mmlu_eval_accuracy": 0.4626311671628311,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "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": 0.5454545454545454,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_conceptual_physics": 0.46153846153846156,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.5,
      "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": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3023255813953488,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6976744186046512,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.5294117647058824,
      "mmlu_eval_accuracy_prehistory": 0.42857142857142855,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.3235294117647059,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0846105751924353,
      "step": 200
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.9027,
      "step": 210
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8621,
      "step": 220
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8405,
      "step": 230
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.8553,
      "step": 240
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8334,
      "step": 250
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.8791,
      "step": 260
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.8607,
      "step": 270
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8403,
      "step": 280
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.8471,
      "step": 290
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.8945,
      "step": 300
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8094,
      "step": 310
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.8571,
      "step": 320
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8469,
      "step": 330
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8609,
      "step": 340
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.8242,
      "step": 350
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.8679,
      "step": 360
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8583,
      "step": 370
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.8815,
      "step": 380
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.819,
      "step": 390
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8946,
      "step": 400
    },
    {
      "epoch": 0.27,
      "eval_loss": 0.8304864764213562,
      "eval_runtime": 191.0197,
      "eval_samples_per_second": 5.235,
      "eval_steps_per_second": 2.618,
      "step": 400
    },
    {
      "epoch": 0.27,
      "mmlu_eval_accuracy": 0.45184631712481954,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.5,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "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": 0.45454545454545453,
      "mmlu_eval_accuracy_college_mathematics": 0.18181818181818182,
      "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": 0.46153846153846156,
      "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": 0.4,
      "mmlu_eval_accuracy_high_school_biology": 0.34375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3023255813953488,
      "mmlu_eval_accuracy_high_school_mathematics": 0.27586206896551724,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
      "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": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "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": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.5588235294117647,
      "mmlu_eval_accuracy_prehistory": 0.4,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.3176470588235294,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.1180029823286726,
      "step": 400
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.8577,
      "step": 410
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.8594,
      "step": 420
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002,
      "loss": 0.8559,
      "step": 430
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8602,
      "step": 440
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0002,
      "loss": 0.8196,
      "step": 450
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0002,
      "loss": 0.8601,
      "step": 460
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8412,
      "step": 470
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.0002,
      "loss": 0.8543,
      "step": 480
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.0002,
      "loss": 0.8705,
      "step": 490
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.7979,
      "step": 500
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.0002,
      "loss": 0.8179,
      "step": 510
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0002,
      "loss": 0.8842,
      "step": 520
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.7691,
      "step": 530
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002,
      "loss": 0.8867,
      "step": 540
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0002,
      "loss": 0.8812,
      "step": 550
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8507,
      "step": 560
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.0002,
      "loss": 0.8627,
      "step": 570
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0002,
      "loss": 0.8451,
      "step": 580
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8396,
      "step": 590
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.0002,
      "loss": 0.8756,
      "step": 600
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.8252214193344116,
      "eval_runtime": 190.9921,
      "eval_samples_per_second": 5.236,
      "eval_steps_per_second": 2.618,
      "step": 600
    },
    {
      "epoch": 0.4,
      "mmlu_eval_accuracy": 0.46243828535426434,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.4375,
      "mmlu_eval_accuracy_business_ethics": 0.45454545454545453,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.36363636363636365,
      "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": 0.4375,
      "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": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "mmlu_eval_accuracy_high_school_psychology": 0.7333333333333333,
      "mmlu_eval_accuracy_high_school_statistics": 0.21739130434782608,
      "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": 0.6923076923076923,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.72,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5757575757575758,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.4857142857142857,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.3058823529411765,
      "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": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.7272727272727273,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.6842105263157895,
      "mmlu_loss": 1.1610933113113706,
      "step": 600
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0002,
      "loss": 0.8505,
      "step": 610
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.8277,
      "step": 620
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0002,
      "loss": 0.798,
      "step": 630
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0002,
      "loss": 0.8582,
      "step": 640
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8542,
      "step": 650
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.0002,
      "loss": 0.8167,
      "step": 660
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0002,
      "loss": 0.8035,
      "step": 670
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8348,
      "step": 680
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0002,
      "loss": 0.8423,
      "step": 690
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0002,
      "loss": 0.8668,
      "step": 700
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8037,
      "step": 710
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0002,
      "loss": 0.8491,
      "step": 720
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0002,
      "loss": 0.8498,
      "step": 730
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8048,
      "step": 740
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002,
      "loss": 0.8662,
      "step": 750
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.0002,
      "loss": 0.8479,
      "step": 760
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8457,
      "step": 770
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0002,
      "loss": 0.8865,
      "step": 780
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.0002,
      "loss": 0.8411,
      "step": 790
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.8207,
      "step": 800
    },
    {
      "epoch": 0.54,
      "eval_loss": 0.8217844367027283,
      "eval_runtime": 191.1923,
      "eval_samples_per_second": 5.23,
      "eval_steps_per_second": 2.615,
      "step": 800
    },
    {
      "epoch": 0.54,
      "mmlu_eval_accuracy": 0.4627342112287351,
      "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": 0.4482758620689655,
      "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": 0.36363636363636365,
      "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": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "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": 0.375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "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": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "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": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.5454545454545454,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.6060606060606061,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "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": 0.4057971014492754,
      "mmlu_eval_accuracy_public_relations": 0.5,
      "mmlu_eval_accuracy_security_studies": 0.5185185185185185,
      "mmlu_eval_accuracy_sociology": 0.6363636363636364,
      "mmlu_eval_accuracy_us_foreign_policy": 0.45454545454545453,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1044702904187045,
      "step": 800
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0002,
      "loss": 0.903,
      "step": 810
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.0002,
      "loss": 0.852,
      "step": 820
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8646,
      "step": 830
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.0002,
      "loss": 0.8278,
      "step": 840
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0002,
      "loss": 0.8349,
      "step": 850
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8422,
      "step": 860
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.0002,
      "loss": 0.8313,
      "step": 870
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0002,
      "loss": 0.8228,
      "step": 880
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8204,
      "step": 890
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.0002,
      "loss": 0.8327,
      "step": 900
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.0002,
      "loss": 0.8523,
      "step": 910
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.8661,
      "step": 920
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.0002,
      "loss": 0.8103,
      "step": 930
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0002,
      "loss": 0.8723,
      "step": 940
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8503,
      "step": 950
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0002,
      "loss": 0.8252,
      "step": 960
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.0002,
      "loss": 0.8143,
      "step": 970
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.824,
      "step": 980
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.0002,
      "loss": 0.8643,
      "step": 990
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.0002,
      "loss": 0.827,
      "step": 1000
    },
    {
      "epoch": 0.67,
      "eval_loss": 0.8191137313842773,
      "eval_runtime": 191.0648,
      "eval_samples_per_second": 5.234,
      "eval_steps_per_second": 2.617,
      "step": 1000
    },
    {
      "epoch": 0.67,
      "mmlu_eval_accuracy": 0.4562828094173592,
      "mmlu_eval_accuracy_abstract_algebra": 0.09090909090909091,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "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": 0.0,
      "mmlu_eval_accuracy_college_computer_science": 0.36363636363636365,
      "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": 0.36363636363636365,
      "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": 0.4146341463414634,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "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": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_statistics": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.5769230769230769,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "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": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6744186046511628,
      "mmlu_eval_accuracy_moral_disputes": 0.42105263157894735,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5151515151515151,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.5161290322580645,
      "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": 0.5909090909090909,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.3333333333333333,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.060513006994992,
      "step": 1000
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8497,
      "step": 1010
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0002,
      "loss": 0.8419,
      "step": 1020
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0002,
      "loss": 0.7963,
      "step": 1030
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8297,
      "step": 1040
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.0002,
      "loss": 0.8797,
      "step": 1050
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002,
      "loss": 0.8088,
      "step": 1060
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8709,
      "step": 1070
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.0002,
      "loss": 0.8643,
      "step": 1080
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.0002,
      "loss": 0.7761,
      "step": 1090
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8927,
      "step": 1100
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.0002,
      "loss": 0.8244,
      "step": 1110
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0002,
      "loss": 0.8095,
      "step": 1120
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.8269,
      "step": 1130
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0002,
      "loss": 0.79,
      "step": 1140
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0002,
      "loss": 0.8232,
      "step": 1150
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8538,
      "step": 1160
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0002,
      "loss": 0.8721,
      "step": 1170
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002,
      "loss": 0.8653,
      "step": 1180
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8279,
      "step": 1190
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0002,
      "loss": 0.8398,
      "step": 1200
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.8146417140960693,
      "eval_runtime": 191.0331,
      "eval_samples_per_second": 5.235,
      "eval_steps_per_second": 2.617,
      "step": 1200
    },
    {
      "epoch": 0.8,
      "mmlu_eval_accuracy": 0.45965424254543474,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.7142857142857143,
      "mmlu_eval_accuracy_astronomy": 0.375,
      "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": 0.25,
      "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": 0.45454545454545453,
      "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": 0.36585365853658536,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.5,
      "mmlu_eval_accuracy_high_school_biology": 0.34375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3488372093023256,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.34615384615384615,
      "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": 0.5769230769230769,
      "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": 0.6111111111111112,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.5454545454545454,
      "mmlu_eval_accuracy_marketing": 0.76,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6627906976744186,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5151515151515151,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.31176470588235294,
      "mmlu_eval_accuracy_professional_medicine": 0.4838709677419355,
      "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": 0.5909090909090909,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.3333333333333333,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.1305760689099216,
      "step": 1200
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.0002,
      "loss": 0.794,
      "step": 1210
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8787,
      "step": 1220
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0002,
      "loss": 0.8212,
      "step": 1230
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.0002,
      "loss": 0.8622,
      "step": 1240
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8568,
      "step": 1250
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.0002,
      "loss": 0.8485,
      "step": 1260
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0002,
      "loss": 0.8174,
      "step": 1270
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8327,
      "step": 1280
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002,
      "loss": 0.8177,
      "step": 1290
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.0002,
      "loss": 0.8023,
      "step": 1300
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.0002,
      "loss": 0.8108,
      "step": 1310
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8222,
      "step": 1320
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0002,
      "loss": 0.8784,
      "step": 1330
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.0002,
      "loss": 0.8098,
      "step": 1340
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8441,
      "step": 1350
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0002,
      "loss": 0.8273,
      "step": 1360
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002,
      "loss": 0.8209,
      "step": 1370
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8096,
      "step": 1380
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0002,
      "loss": 0.8163,
      "step": 1390
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.0002,
      "loss": 0.8194,
      "step": 1400
    },
    {
      "epoch": 0.94,
      "eval_loss": 0.8138102889060974,
      "eval_runtime": 191.1419,
      "eval_samples_per_second": 5.232,
      "eval_steps_per_second": 2.616,
      "step": 1400
    },
    {
      "epoch": 0.94,
      "mmlu_eval_accuracy": 0.47157305762995727,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.7142857142857143,
      "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": 0.25,
      "mmlu_eval_accuracy_college_computer_science": 0.45454545454545453,
      "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": 0.43902439024390244,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "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": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "mmlu_eval_accuracy_high_school_physics": 0.29411764705882354,
      "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.391304347826087,
      "mmlu_eval_accuracy_high_school_us_history": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_world_history": 0.6153846153846154,
      "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": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.686046511627907,
      "mmlu_eval_accuracy_moral_disputes": 0.4473684210526316,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5151515151515151,
      "mmlu_eval_accuracy_philosophy": 0.4411764705882353,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.2903225806451613,
      "mmlu_eval_accuracy_professional_law": 0.31176470588235294,
      "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": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 0.9451145088937202,
      "step": 1400
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8871,
      "step": 1410
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.0002,
      "loss": 0.8086,
      "step": 1420
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.0002,
      "loss": 0.8082,
      "step": 1430
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8396,
      "step": 1440
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.0002,
      "loss": 0.8239,
      "step": 1450
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.0002,
      "loss": 0.816,
      "step": 1460
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.817,
      "step": 1470
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002,
      "loss": 0.8399,
      "step": 1480
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002,
      "loss": 0.835,
      "step": 1490
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.7444,
      "step": 1500
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.0002,
      "loss": 0.7771,
      "step": 1510
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.0002,
      "loss": 0.7708,
      "step": 1520
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.7805,
      "step": 1530
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.0002,
      "loss": 0.766,
      "step": 1540
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.0002,
      "loss": 0.7637,
      "step": 1550
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7254,
      "step": 1560
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0002,
      "loss": 0.7519,
      "step": 1570
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.0002,
      "loss": 0.7441,
      "step": 1580
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7175,
      "step": 1590
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 1600
    },
    {
      "epoch": 1.07,
      "eval_loss": 0.8182914853096008,
      "eval_runtime": 191.1519,
      "eval_samples_per_second": 5.231,
      "eval_steps_per_second": 2.616,
      "step": 1600
    },
    {
      "epoch": 1.07,
      "mmlu_eval_accuracy": 0.46122104830313654,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.7142857142857143,
      "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": 0.4375,
      "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": 0.3181818181818182,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.2727272727272727,
      "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": 0.375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.36363636363636365,
      "mmlu_eval_accuracy_high_school_computer_science": 0.4444444444444444,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3488372093023256,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "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": 0.34782608695652173,
      "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": 0.3333333333333333,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.64,
      "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": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5151515151515151,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "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": 0.42028985507246375,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "mmlu_eval_accuracy_sociology": 0.5909090909090909,
      "mmlu_eval_accuracy_us_foreign_policy": 0.6363636363636364,
      "mmlu_eval_accuracy_virology": 0.5,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 0.9508307614113269,
      "step": 1600
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.0002,
      "loss": 0.7187,
      "step": 1610
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7525,
      "step": 1620
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.0002,
      "loss": 0.7606,
      "step": 1630
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.0002,
      "loss": 0.7593,
      "step": 1640
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7942,
      "step": 1650
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 1660
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.0002,
      "loss": 0.7519,
      "step": 1670
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.728,
      "step": 1680
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0002,
      "loss": 0.773,
      "step": 1690
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.0002,
      "loss": 0.7509,
      "step": 1700
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7435,
      "step": 1710
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0002,
      "loss": 0.7797,
      "step": 1720
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.0002,
      "loss": 0.7727,
      "step": 1730
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.7488,
      "step": 1740
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0002,
      "loss": 0.734,
      "step": 1750
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.0002,
      "loss": 0.7716,
      "step": 1760
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.7514,
      "step": 1770
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0002,
      "loss": 0.789,
      "step": 1780
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002,
      "loss": 0.7544,
      "step": 1790
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7649,
      "step": 1800
    },
    {
      "epoch": 1.21,
      "eval_loss": 0.8216288685798645,
      "eval_runtime": 191.3902,
      "eval_samples_per_second": 5.225,
      "eval_steps_per_second": 2.612,
      "step": 1800
    },
    {
      "epoch": 1.21,
      "mmlu_eval_accuracy": 0.45547954117728906,
      "mmlu_eval_accuracy_abstract_algebra": 0.18181818181818182,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "mmlu_eval_accuracy_astronomy": 0.3125,
      "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": 0.45454545454545453,
      "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": 0.2727272727272727,
      "mmlu_eval_accuracy_conceptual_physics": 0.38461538461538464,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.34146341463414637,
      "mmlu_eval_accuracy_formal_logic": 0.21428571428571427,
      "mmlu_eval_accuracy_global_facts": 0.2,
      "mmlu_eval_accuracy_high_school_biology": 0.375,
      "mmlu_eval_accuracy_high_school_chemistry": 0.4090909090909091,
      "mmlu_eval_accuracy_high_school_computer_science": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_european_history": 0.7777777777777778,
      "mmlu_eval_accuracy_high_school_geography": 0.6818181818181818,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.3448275862068966,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
      "mmlu_eval_accuracy_high_school_statistics": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5384615384615384,
      "mmlu_eval_accuracy_human_aging": 0.6521739130434783,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.64,
      "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": 0.25,
      "mmlu_eval_accuracy_nutrition": 0.48484848484848486,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.34705882352941175,
      "mmlu_eval_accuracy_professional_medicine": 0.41935483870967744,
      "mmlu_eval_accuracy_professional_psychology": 0.43478260869565216,
      "mmlu_eval_accuracy_public_relations": 0.6666666666666666,
      "mmlu_eval_accuracy_security_studies": 0.5555555555555556,
      "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": 1.018148140189075,
      "step": 1800
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002,
      "loss": 0.7312,
      "step": 1810
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0002,
      "loss": 0.7612,
      "step": 1820
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7495,
      "step": 1830
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0002,
      "loss": 0.7255,
      "step": 1840
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0002,
      "loss": 0.7511,
      "step": 1850
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7879,
      "step": 1860
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.0002,
      "loss": 0.7397,
      "step": 1870
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0002,
      "loss": 0.8136,
      "step": 1880
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7964,
      "step": 1890
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.0002,
      "loss": 0.7584,
      "step": 1900
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0002,
      "loss": 0.7701,
      "step": 1910
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7423,
      "step": 1920
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.0002,
      "loss": 0.7411,
      "step": 1930
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.0002,
      "loss": 0.7673,
      "step": 1940
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.717,
      "step": 1950
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.0002,
      "loss": 0.7825,
      "step": 1960
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.0002,
      "loss": 0.7682,
      "step": 1970
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7266,
      "step": 1980
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.0002,
      "loss": 0.7417,
      "step": 1990
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.0002,
      "loss": 0.81,
      "step": 2000
    },
    {
      "epoch": 1.34,
      "eval_loss": 0.8217906951904297,
      "eval_runtime": 191.5711,
      "eval_samples_per_second": 5.22,
      "eval_steps_per_second": 2.61,
      "step": 2000
    },
    {
      "epoch": 1.34,
      "mmlu_eval_accuracy": 0.4667080313375052,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "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": 0.375,
      "mmlu_eval_accuracy_college_chemistry": 0.25,
      "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": 0.6363636363636364,
      "mmlu_eval_accuracy_conceptual_physics": 0.4230769230769231,
      "mmlu_eval_accuracy_econometrics": 0.25,
      "mmlu_eval_accuracy_electrical_engineering": 0.4375,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2926829268292683,
      "mmlu_eval_accuracy_formal_logic": 0.14285714285714285,
      "mmlu_eval_accuracy_global_facts": 0.4,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_geography": 0.7727272727272727,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6190476190476191,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.1724137931034483,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.4230769230769231,
      "mmlu_eval_accuracy_high_school_physics": 0.23529411764705882,
      "mmlu_eval_accuracy_high_school_psychology": 0.7166666666666667,
      "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": 0.3333333333333333,
      "mmlu_eval_accuracy_international_law": 0.8461538461538461,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5,
      "mmlu_eval_accuracy_machine_learning": 0.36363636363636365,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.68,
      "mmlu_eval_accuracy_medical_genetics": 0.7272727272727273,
      "mmlu_eval_accuracy_miscellaneous": 0.6395348837209303,
      "mmlu_eval_accuracy_moral_disputes": 0.47368421052631576,
      "mmlu_eval_accuracy_moral_scenarios": 0.24,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.5714285714285714,
      "mmlu_eval_accuracy_professional_accounting": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.4057971014492754,
      "mmlu_eval_accuracy_public_relations": 0.6666666666666666,
      "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": 0.6842105263157895,
      "mmlu_loss": 1.0138715160499976,
      "step": 2000
    }
  ],
  "max_steps": 5000,
  "num_train_epochs": 4,
  "total_flos": 4.28111794915541e+17,
  "trial_name": null,
  "trial_params": null
}