{
  "best_metric": 0.7810184359550476,
  "best_model_checkpoint": "./output_v2/7b_cluster022_Nous-Hermes-llama-2-7b_partitioned_v3_standardized_022/checkpoint-800",
  "epoch": 0.2822118352588412,
  "global_step": 800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 0.0002,
      "loss": 0.8325,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.8202,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7466,
      "step": 30
    },
    {
      "epoch": 0.01,
      "learning_rate": 0.0002,
      "loss": 0.7549,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7569,
      "step": 50
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.7691,
      "step": 60
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0002,
      "loss": 0.744,
      "step": 70
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.7708,
      "step": 80
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.0002,
      "loss": 0.8071,
      "step": 90
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7303,
      "step": 100
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.6861,
      "step": 110
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0002,
      "loss": 0.7592,
      "step": 120
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7361,
      "step": 130
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.76,
      "step": 140
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.0002,
      "loss": 0.7617,
      "step": 150
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7073,
      "step": 160
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7581,
      "step": 170
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0002,
      "loss": 0.7636,
      "step": 180
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7712,
      "step": 190
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7547,
      "step": 200
    },
    {
      "epoch": 0.07,
      "eval_loss": 0.8007758259773254,
      "eval_runtime": 185.4331,
      "eval_samples_per_second": 5.393,
      "eval_steps_per_second": 2.696,
      "step": 200
    },
    {
      "epoch": 0.07,
      "mmlu_eval_accuracy": 0.4659766842502547,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.6428571428571429,
      "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": 0.4375,
      "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": 0.36363636363636365,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "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": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.2682926829268293,
      "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": 0.3181818181818182,
      "mmlu_eval_accuracy_high_school_computer_science": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.5555555555555556,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.37209302325581395,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "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": 0.34782608695652173,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.36363636363636365,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.7272727272727273,
      "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": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.45161290322580644,
      "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": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.3888888888888889,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0009409290575484,
      "step": 200
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 0.7814,
      "step": 210
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7313,
      "step": 220
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7217,
      "step": 230
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.0002,
      "loss": 0.7299,
      "step": 240
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7229,
      "step": 250
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0002,
      "loss": 0.7271,
      "step": 260
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 270
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 280
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0002,
      "loss": 0.7434,
      "step": 290
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.6741,
      "step": 300
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7386,
      "step": 310
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0002,
      "loss": 0.7441,
      "step": 320
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7243,
      "step": 330
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7534,
      "step": 340
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0002,
      "loss": 0.7187,
      "step": 350
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 360
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7597,
      "step": 370
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0002,
      "loss": 0.7398,
      "step": 380
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.6924,
      "step": 390
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7035,
      "step": 400
    },
    {
      "epoch": 0.14,
      "eval_loss": 0.7900036573410034,
      "eval_runtime": 188.9686,
      "eval_samples_per_second": 5.292,
      "eval_steps_per_second": 2.646,
      "step": 400
    },
    {
      "epoch": 0.14,
      "mmlu_eval_accuracy": 0.45327850420813054,
      "mmlu_eval_accuracy_abstract_algebra": 0.2727272727272727,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "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": 0.18181818181818182,
      "mmlu_eval_accuracy_college_medicine": 0.2727272727272727,
      "mmlu_eval_accuracy_college_physics": 0.45454545454545453,
      "mmlu_eval_accuracy_computer_security": 0.2727272727272727,
      "mmlu_eval_accuracy_conceptual_physics": 0.34615384615384615,
      "mmlu_eval_accuracy_econometrics": 0.16666666666666666,
      "mmlu_eval_accuracy_electrical_engineering": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.36585365853658536,
      "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": 0.32558139534883723,
      "mmlu_eval_accuracy_high_school_mathematics": 0.20689655172413793,
      "mmlu_eval_accuracy_high_school_microeconomics": 0.38461538461538464,
      "mmlu_eval_accuracy_high_school_physics": 0.35294117647058826,
      "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": 0.5,
      "mmlu_eval_accuracy_human_aging": 0.6956521739130435,
      "mmlu_eval_accuracy_human_sexuality": 0.4166666666666667,
      "mmlu_eval_accuracy_international_law": 0.6923076923076923,
      "mmlu_eval_accuracy_jurisprudence": 0.2727272727272727,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.18181818181818182,
      "mmlu_eval_accuracy_management": 0.45454545454545453,
      "mmlu_eval_accuracy_marketing": 0.76,
      "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": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.5,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.35294117647058826,
      "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": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.0370562722026213,
      "step": 400
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002,
      "loss": 0.7245,
      "step": 410
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.8027,
      "step": 420
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0002,
      "loss": 0.7174,
      "step": 430
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7471,
      "step": 440
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7263,
      "step": 450
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0002,
      "loss": 0.7001,
      "step": 460
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7767,
      "step": 470
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7406,
      "step": 480
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0002,
      "loss": 0.7371,
      "step": 490
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7152,
      "step": 500
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.746,
      "step": 510
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0002,
      "loss": 0.7178,
      "step": 520
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.7056,
      "step": 530
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.6961,
      "step": 540
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.0002,
      "loss": 0.673,
      "step": 550
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 560
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7508,
      "step": 570
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0002,
      "loss": 0.7122,
      "step": 580
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7154,
      "step": 590
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002,
      "loss": 0.7587,
      "step": 600
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.785829484462738,
      "eval_runtime": 189.436,
      "eval_samples_per_second": 5.279,
      "eval_steps_per_second": 2.639,
      "step": 600
    },
    {
      "epoch": 0.21,
      "mmlu_eval_accuracy": 0.4748863496985387,
      "mmlu_eval_accuracy_abstract_algebra": 0.36363636363636365,
      "mmlu_eval_accuracy_anatomy": 0.5714285714285714,
      "mmlu_eval_accuracy_astronomy": 0.5625,
      "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": 0.45454545454545453,
      "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": 0.34146341463414637,
      "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": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_european_history": 0.6111111111111112,
      "mmlu_eval_accuracy_high_school_geography": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_government_and_politics": 0.6666666666666666,
      "mmlu_eval_accuracy_high_school_macroeconomics": 0.3488372093023256,
      "mmlu_eval_accuracy_high_school_mathematics": 0.2413793103448276,
      "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": 0.30434782608695654,
      "mmlu_eval_accuracy_high_school_us_history": 0.7272727272727273,
      "mmlu_eval_accuracy_high_school_world_history": 0.5,
      "mmlu_eval_accuracy_human_aging": 0.7391304347826086,
      "mmlu_eval_accuracy_human_sexuality": 0.5,
      "mmlu_eval_accuracy_international_law": 0.7692307692307693,
      "mmlu_eval_accuracy_jurisprudence": 0.45454545454545453,
      "mmlu_eval_accuracy_logical_fallacies": 0.5555555555555556,
      "mmlu_eval_accuracy_machine_learning": 0.2727272727272727,
      "mmlu_eval_accuracy_management": 0.6363636363636364,
      "mmlu_eval_accuracy_marketing": 0.76,
      "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": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.45714285714285713,
      "mmlu_eval_accuracy_professional_accounting": 0.25806451612903225,
      "mmlu_eval_accuracy_professional_law": 0.3411764705882353,
      "mmlu_eval_accuracy_professional_medicine": 0.3225806451612903,
      "mmlu_eval_accuracy_professional_psychology": 0.42028985507246375,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.5925925925925926,
      "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": 0.7894736842105263,
      "mmlu_loss": 0.9638749470559798,
      "step": 600
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.8107,
      "step": 610
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7193,
      "step": 620
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.0002,
      "loss": 0.7275,
      "step": 630
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7553,
      "step": 640
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7385,
      "step": 650
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.0002,
      "loss": 0.7071,
      "step": 660
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7395,
      "step": 670
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7512,
      "step": 680
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0002,
      "loss": 0.7063,
      "step": 690
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.723,
      "step": 700
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7068,
      "step": 710
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.0002,
      "loss": 0.7211,
      "step": 720
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.7123,
      "step": 730
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.6394,
      "step": 740
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0002,
      "loss": 0.679,
      "step": 750
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7402,
      "step": 760
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.0002,
      "loss": 0.7634,
      "step": 770
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7253,
      "step": 780
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7497,
      "step": 790
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.0002,
      "loss": 0.7008,
      "step": 800
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.7810184359550476,
      "eval_runtime": 187.2684,
      "eval_samples_per_second": 5.34,
      "eval_steps_per_second": 2.67,
      "step": 800
    },
    {
      "epoch": 0.28,
      "mmlu_eval_accuracy": 0.46145585660156935,
      "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": 0.18181818181818182,
      "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": 0.3125,
      "mmlu_eval_accuracy_elementary_mathematics": 0.3170731707317073,
      "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": 0.6666666666666666,
      "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": 0.3023255813953488,
      "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": 0.5769230769230769,
      "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": 0.5555555555555556,
      "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": 0.5,
      "mmlu_eval_accuracy_moral_scenarios": 0.23,
      "mmlu_eval_accuracy_nutrition": 0.5454545454545454,
      "mmlu_eval_accuracy_philosophy": 0.47058823529411764,
      "mmlu_eval_accuracy_prehistory": 0.5142857142857142,
      "mmlu_eval_accuracy_professional_accounting": 0.3548387096774194,
      "mmlu_eval_accuracy_professional_law": 0.3352941176470588,
      "mmlu_eval_accuracy_professional_medicine": 0.3870967741935484,
      "mmlu_eval_accuracy_professional_psychology": 0.42028985507246375,
      "mmlu_eval_accuracy_public_relations": 0.5833333333333334,
      "mmlu_eval_accuracy_security_studies": 0.48148148148148145,
      "mmlu_eval_accuracy_sociology": 0.6818181818181818,
      "mmlu_eval_accuracy_us_foreign_policy": 0.5454545454545454,
      "mmlu_eval_accuracy_virology": 0.4444444444444444,
      "mmlu_eval_accuracy_world_religions": 0.7368421052631579,
      "mmlu_loss": 1.074835775843177,
      "step": 800
    }
  ],
  "max_steps": 5000,
  "num_train_epochs": 2,
  "total_flos": 1.842037467536425e+17,
  "trial_name": null,
  "trial_params": null
}