{
  "best_metric": 0.9180880188941956,
  "best_model_checkpoint": "./fine_tuned_models/kvkk/Llama-3.1-8B-Instruct/checkpoint-2600",
  "epoch": 2.5669135802469136,
  "eval_steps": 100,
  "global_step": 2600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.019753086419753086,
      "grad_norm": 1.1142268180847168,
      "learning_rate": 1.9736842105263157e-06,
      "loss": 2.0286,
      "step": 20
    },
    {
      "epoch": 0.03950617283950617,
      "grad_norm": 1.1114156246185303,
      "learning_rate": 3.9473684210526315e-06,
      "loss": 2.012,
      "step": 40
    },
    {
      "epoch": 0.05925925925925926,
      "grad_norm": 1.0126515626907349,
      "learning_rate": 5.921052631578948e-06,
      "loss": 1.8797,
      "step": 60
    },
    {
      "epoch": 0.07901234567901234,
      "grad_norm": 1.0541017055511475,
      "learning_rate": 7.894736842105263e-06,
      "loss": 1.5698,
      "step": 80
    },
    {
      "epoch": 0.09876543209876543,
      "grad_norm": 0.8761700391769409,
      "learning_rate": 9.868421052631579e-06,
      "loss": 1.2906,
      "step": 100
    },
    {
      "epoch": 0.09876543209876543,
      "eval_loss": 1.1851595640182495,
      "eval_runtime": 402.5327,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 100
    },
    {
      "epoch": 0.11851851851851852,
      "grad_norm": 0.7023137211799622,
      "learning_rate": 1.1842105263157895e-05,
      "loss": 1.1435,
      "step": 120
    },
    {
      "epoch": 0.1382716049382716,
      "grad_norm": 0.7056591510772705,
      "learning_rate": 1.3815789473684211e-05,
      "loss": 1.1222,
      "step": 140
    },
    {
      "epoch": 0.1580246913580247,
      "grad_norm": 0.6742960810661316,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 1.1008,
      "step": 160
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.6822540760040283,
      "learning_rate": 1.7763157894736842e-05,
      "loss": 1.0732,
      "step": 180
    },
    {
      "epoch": 0.19753086419753085,
      "grad_norm": 0.7084145545959473,
      "learning_rate": 1.9736842105263158e-05,
      "loss": 1.0525,
      "step": 200
    },
    {
      "epoch": 0.19753086419753085,
      "eval_loss": 1.0499310493469238,
      "eval_runtime": 402.5074,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 200
    },
    {
      "epoch": 0.21728395061728395,
      "grad_norm": 0.8174996972084045,
      "learning_rate": 2.1710526315789474e-05,
      "loss": 1.0711,
      "step": 220
    },
    {
      "epoch": 0.23703703703703705,
      "grad_norm": 0.8514217138290405,
      "learning_rate": 2.368421052631579e-05,
      "loss": 1.0147,
      "step": 240
    },
    {
      "epoch": 0.25679012345679014,
      "grad_norm": 0.8636080622673035,
      "learning_rate": 2.5657894736842107e-05,
      "loss": 1.032,
      "step": 260
    },
    {
      "epoch": 0.2765432098765432,
      "grad_norm": 0.8430706262588501,
      "learning_rate": 2.7631578947368423e-05,
      "loss": 1.0326,
      "step": 280
    },
    {
      "epoch": 0.2962962962962963,
      "grad_norm": 0.9532914757728577,
      "learning_rate": 2.9605263157894735e-05,
      "loss": 1.0488,
      "step": 300
    },
    {
      "epoch": 0.2962962962962963,
      "eval_loss": 1.0270123481750488,
      "eval_runtime": 402.3395,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 300
    },
    {
      "epoch": 0.3160493827160494,
      "grad_norm": 0.9832524061203003,
      "learning_rate": 2.9997461206762896e-05,
      "loss": 1.0209,
      "step": 320
    },
    {
      "epoch": 0.3358024691358025,
      "grad_norm": 0.9382938146591187,
      "learning_rate": 2.998714883211034e-05,
      "loss": 1.0287,
      "step": 340
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 1.03255295753479,
      "learning_rate": 2.9968909651554313e-05,
      "loss": 1.0195,
      "step": 360
    },
    {
      "epoch": 0.37530864197530867,
      "grad_norm": 1.0193625688552856,
      "learning_rate": 2.994275331192262e-05,
      "loss": 1.0104,
      "step": 380
    },
    {
      "epoch": 0.3950617283950617,
      "grad_norm": 0.9543349146842957,
      "learning_rate": 2.9908693647482852e-05,
      "loss": 1.0032,
      "step": 400
    },
    {
      "epoch": 0.3950617283950617,
      "eval_loss": 1.0094538927078247,
      "eval_runtime": 402.3092,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.119,
      "step": 400
    },
    {
      "epoch": 0.4148148148148148,
      "grad_norm": 1.0028969049453735,
      "learning_rate": 2.9866748672625354e-05,
      "loss": 1.0,
      "step": 420
    },
    {
      "epoch": 0.4345679012345679,
      "grad_norm": 1.1169507503509521,
      "learning_rate": 2.9816940572335275e-05,
      "loss": 1.0028,
      "step": 440
    },
    {
      "epoch": 0.454320987654321,
      "grad_norm": 1.0066871643066406,
      "learning_rate": 2.9759295690458793e-05,
      "loss": 0.9991,
      "step": 460
    },
    {
      "epoch": 0.4740740740740741,
      "grad_norm": 1.1250879764556885,
      "learning_rate": 2.9693844515769658e-05,
      "loss": 1.0275,
      "step": 480
    },
    {
      "epoch": 0.49382716049382713,
      "grad_norm": 0.9638034701347351,
      "learning_rate": 2.9620621665843482e-05,
      "loss": 1.0097,
      "step": 500
    },
    {
      "epoch": 0.49382716049382713,
      "eval_loss": 0.9942829012870789,
      "eval_runtime": 402.3967,
      "eval_samples_per_second": 4.473,
      "eval_steps_per_second": 1.118,
      "step": 500
    },
    {
      "epoch": 0.5135802469135803,
      "grad_norm": 1.106367826461792,
      "learning_rate": 2.953966586874827e-05,
      "loss": 1.0219,
      "step": 520
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 1.0724130868911743,
      "learning_rate": 2.9451019942560854e-05,
      "loss": 0.9827,
      "step": 540
    },
    {
      "epoch": 0.5530864197530864,
      "grad_norm": 1.0933202505111694,
      "learning_rate": 2.9354730772720152e-05,
      "loss": 1.0033,
      "step": 560
    },
    {
      "epoch": 0.5728395061728395,
      "grad_norm": 1.089522123336792,
      "learning_rate": 2.9250849287229115e-05,
      "loss": 0.9558,
      "step": 580
    },
    {
      "epoch": 0.5925925925925926,
      "grad_norm": 1.0908564329147339,
      "learning_rate": 2.9139430429718557e-05,
      "loss": 0.9702,
      "step": 600
    },
    {
      "epoch": 0.5925925925925926,
      "eval_loss": 0.9849551916122437,
      "eval_runtime": 402.4211,
      "eval_samples_per_second": 4.473,
      "eval_steps_per_second": 1.118,
      "step": 600
    },
    {
      "epoch": 0.6123456790123457,
      "grad_norm": 1.0643543004989624,
      "learning_rate": 2.90205331303871e-05,
      "loss": 0.9678,
      "step": 620
    },
    {
      "epoch": 0.6320987654320988,
      "grad_norm": 1.1446090936660767,
      "learning_rate": 2.889422027483259e-05,
      "loss": 0.9617,
      "step": 640
    },
    {
      "epoch": 0.6518518518518519,
      "grad_norm": 1.2303221225738525,
      "learning_rate": 2.876055867079146e-05,
      "loss": 0.9939,
      "step": 660
    },
    {
      "epoch": 0.671604938271605,
      "grad_norm": 1.1191219091415405,
      "learning_rate": 2.8619619012803674e-05,
      "loss": 0.986,
      "step": 680
    },
    {
      "epoch": 0.691358024691358,
      "grad_norm": 1.2467299699783325,
      "learning_rate": 2.847147584482188e-05,
      "loss": 0.9787,
      "step": 700
    },
    {
      "epoch": 0.691358024691358,
      "eval_loss": 0.9761277437210083,
      "eval_runtime": 402.3652,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 700
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 1.1553202867507935,
      "learning_rate": 2.8316207520784622e-05,
      "loss": 0.9545,
      "step": 720
    },
    {
      "epoch": 0.7308641975308642,
      "grad_norm": 1.162398338317871,
      "learning_rate": 2.81538961631744e-05,
      "loss": 0.9876,
      "step": 740
    },
    {
      "epoch": 0.7506172839506173,
      "grad_norm": 1.1037890911102295,
      "learning_rate": 2.7984627619582507e-05,
      "loss": 0.9526,
      "step": 760
    },
    {
      "epoch": 0.7703703703703704,
      "grad_norm": 1.1114506721496582,
      "learning_rate": 2.7808491417303672e-05,
      "loss": 1.0013,
      "step": 780
    },
    {
      "epoch": 0.7901234567901234,
      "grad_norm": 1.140010118484497,
      "learning_rate": 2.7625580715984452e-05,
      "loss": 0.9754,
      "step": 800
    },
    {
      "epoch": 0.7901234567901234,
      "eval_loss": 0.9668111205101013,
      "eval_runtime": 402.4946,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 800
    },
    {
      "epoch": 0.8098765432098766,
      "grad_norm": 1.1689811944961548,
      "learning_rate": 2.7435992258350444e-05,
      "loss": 0.979,
      "step": 820
    },
    {
      "epoch": 0.8296296296296296,
      "grad_norm": 1.045784592628479,
      "learning_rate": 2.7239826319038413e-05,
      "loss": 0.9723,
      "step": 840
    },
    {
      "epoch": 0.8493827160493828,
      "grad_norm": 1.1769437789916992,
      "learning_rate": 2.703718665156033e-05,
      "loss": 0.9709,
      "step": 860
    },
    {
      "epoch": 0.8691358024691358,
      "grad_norm": 1.0712182521820068,
      "learning_rate": 2.6828180433427437e-05,
      "loss": 0.964,
      "step": 880
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 1.4096373319625854,
      "learning_rate": 2.6612918209463323e-05,
      "loss": 0.9775,
      "step": 900
    },
    {
      "epoch": 0.8888888888888888,
      "eval_loss": 0.963696300983429,
      "eval_runtime": 402.3913,
      "eval_samples_per_second": 4.473,
      "eval_steps_per_second": 1.118,
      "step": 900
    },
    {
      "epoch": 0.908641975308642,
      "grad_norm": 1.0928242206573486,
      "learning_rate": 2.6391513833336013e-05,
      "loss": 0.9557,
      "step": 920
    },
    {
      "epoch": 0.928395061728395,
      "grad_norm": 1.1433868408203125,
      "learning_rate": 2.616408440733997e-05,
      "loss": 0.9321,
      "step": 940
    },
    {
      "epoch": 0.9481481481481482,
      "grad_norm": 1.1303808689117432,
      "learning_rate": 2.5930750220459893e-05,
      "loss": 0.9873,
      "step": 960
    },
    {
      "epoch": 0.9679012345679012,
      "grad_norm": 1.2201975584030151,
      "learning_rate": 2.569163468474906e-05,
      "loss": 0.9635,
      "step": 980
    },
    {
      "epoch": 0.9876543209876543,
      "grad_norm": 1.1534947156906128,
      "learning_rate": 2.5446864270055825e-05,
      "loss": 0.9749,
      "step": 1000
    },
    {
      "epoch": 0.9876543209876543,
      "eval_loss": 0.9554852843284607,
      "eval_runtime": 402.4957,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 1000
    },
    {
      "epoch": 1.0069135802469136,
      "grad_norm": 1.2161285877227783,
      "learning_rate": 2.5196568437132855e-05,
      "loss": 0.946,
      "step": 1020
    },
    {
      "epoch": 1.0266666666666666,
      "grad_norm": 1.110687255859375,
      "learning_rate": 2.4940879569164453e-05,
      "loss": 0.9553,
      "step": 1040
    },
    {
      "epoch": 1.0464197530864197,
      "grad_norm": 1.1861077547073364,
      "learning_rate": 2.467993290174819e-05,
      "loss": 0.9416,
      "step": 1060
    },
    {
      "epoch": 1.066172839506173,
      "grad_norm": 1.2155508995056152,
      "learning_rate": 2.4413866451367836e-05,
      "loss": 0.9075,
      "step": 1080
    },
    {
      "epoch": 1.085925925925926,
      "grad_norm": 1.3916659355163574,
      "learning_rate": 2.414282094239555e-05,
      "loss": 0.9344,
      "step": 1100
    },
    {
      "epoch": 1.085925925925926,
      "eval_loss": 0.9527194499969482,
      "eval_runtime": 402.3177,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.119,
      "step": 1100
    },
    {
      "epoch": 1.105679012345679,
      "grad_norm": 1.2178341150283813,
      "learning_rate": 2.386693973266173e-05,
      "loss": 0.9271,
      "step": 1120
    },
    {
      "epoch": 1.125432098765432,
      "grad_norm": 1.295937418937683,
      "learning_rate": 2.3586368737632124e-05,
      "loss": 0.9078,
      "step": 1140
    },
    {
      "epoch": 1.145185185185185,
      "grad_norm": 1.3823224306106567,
      "learning_rate": 2.3301256353232112e-05,
      "loss": 0.9113,
      "step": 1160
    },
    {
      "epoch": 1.1649382716049383,
      "grad_norm": 1.3728779554367065,
      "learning_rate": 2.3011753377359084e-05,
      "loss": 0.9135,
      "step": 1180
    },
    {
      "epoch": 1.1846913580246914,
      "grad_norm": 1.2611992359161377,
      "learning_rate": 2.2718012930124407e-05,
      "loss": 0.9178,
      "step": 1200
    },
    {
      "epoch": 1.1846913580246914,
      "eval_loss": 0.9479934573173523,
      "eval_runtime": 402.3084,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.119,
      "step": 1200
    },
    {
      "epoch": 1.2044444444444444,
      "grad_norm": 1.2411032915115356,
      "learning_rate": 2.2420190372867144e-05,
      "loss": 0.9159,
      "step": 1220
    },
    {
      "epoch": 1.2241975308641975,
      "grad_norm": 1.3021560907363892,
      "learning_rate": 2.2118443225982365e-05,
      "loss": 0.8981,
      "step": 1240
    },
    {
      "epoch": 1.2439506172839505,
      "grad_norm": 1.3610398769378662,
      "learning_rate": 2.1812931085607537e-05,
      "loss": 0.9121,
      "step": 1260
    },
    {
      "epoch": 1.2637037037037038,
      "grad_norm": 1.310373067855835,
      "learning_rate": 2.1503815539211034e-05,
      "loss": 0.8987,
      "step": 1280
    },
    {
      "epoch": 1.2834567901234568,
      "grad_norm": 1.2247895002365112,
      "learning_rate": 2.119126008012741e-05,
      "loss": 0.9115,
      "step": 1300
    },
    {
      "epoch": 1.2834567901234568,
      "eval_loss": 0.9459941983222961,
      "eval_runtime": 402.3726,
      "eval_samples_per_second": 4.473,
      "eval_steps_per_second": 1.118,
      "step": 1300
    },
    {
      "epoch": 1.3032098765432099,
      "grad_norm": 1.2773354053497314,
      "learning_rate": 2.0875430021084665e-05,
      "loss": 0.924,
      "step": 1320
    },
    {
      "epoch": 1.322962962962963,
      "grad_norm": 1.2930251359939575,
      "learning_rate": 2.0556492406769176e-05,
      "loss": 0.9201,
      "step": 1340
    },
    {
      "epoch": 1.342716049382716,
      "grad_norm": 1.255893349647522,
      "learning_rate": 2.0234615925474648e-05,
      "loss": 0.9378,
      "step": 1360
    },
    {
      "epoch": 1.3624691358024692,
      "grad_norm": 1.2988559007644653,
      "learning_rate": 1.9909970819881708e-05,
      "loss": 0.9004,
      "step": 1380
    },
    {
      "epoch": 1.3822222222222222,
      "grad_norm": 1.328128457069397,
      "learning_rate": 1.9582728797015385e-05,
      "loss": 0.921,
      "step": 1400
    },
    {
      "epoch": 1.3822222222222222,
      "eval_loss": 0.9410489797592163,
      "eval_runtime": 402.2547,
      "eval_samples_per_second": 4.475,
      "eval_steps_per_second": 1.119,
      "step": 1400
    },
    {
      "epoch": 1.4019753086419753,
      "grad_norm": 1.3988349437713623,
      "learning_rate": 1.9253062937428073e-05,
      "loss": 0.9056,
      "step": 1420
    },
    {
      "epoch": 1.4217283950617283,
      "grad_norm": 1.256790041923523,
      "learning_rate": 1.892114760365605e-05,
      "loss": 0.9073,
      "step": 1440
    },
    {
      "epoch": 1.4414814814814814,
      "grad_norm": 1.3246599435806274,
      "learning_rate": 1.8587158347997932e-05,
      "loss": 0.9097,
      "step": 1460
    },
    {
      "epoch": 1.4612345679012346,
      "grad_norm": 1.2596899271011353,
      "learning_rate": 1.8251271819663863e-05,
      "loss": 0.9104,
      "step": 1480
    },
    {
      "epoch": 1.4809876543209877,
      "grad_norm": 1.2959362268447876,
      "learning_rate": 1.7913665671344526e-05,
      "loss": 0.8994,
      "step": 1500
    },
    {
      "epoch": 1.4809876543209877,
      "eval_loss": 0.9367458820343018,
      "eval_runtime": 402.3021,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.119,
      "step": 1500
    },
    {
      "epoch": 1.5007407407407407,
      "grad_norm": 1.2440412044525146,
      "learning_rate": 1.7574518465249413e-05,
      "loss": 0.9056,
      "step": 1520
    },
    {
      "epoch": 1.520493827160494,
      "grad_norm": 1.441502571105957,
      "learning_rate": 1.7234009578664066e-05,
      "loss": 0.9264,
      "step": 1540
    },
    {
      "epoch": 1.5402469135802468,
      "grad_norm": 1.3158551454544067,
      "learning_rate": 1.6892319109076183e-05,
      "loss": 0.8845,
      "step": 1560
    },
    {
      "epoch": 1.56,
      "grad_norm": 1.5368109941482544,
      "learning_rate": 1.654962777892081e-05,
      "loss": 0.9141,
      "step": 1580
    },
    {
      "epoch": 1.579753086419753,
      "grad_norm": 1.3186088800430298,
      "learning_rate": 1.6206116839995033e-05,
      "loss": 0.8878,
      "step": 1600
    },
    {
      "epoch": 1.579753086419753,
      "eval_loss": 0.9328420758247375,
      "eval_runtime": 402.4724,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 1600
    },
    {
      "epoch": 1.5995061728395061,
      "grad_norm": 1.448764443397522,
      "learning_rate": 1.5861967977592642e-05,
      "loss": 0.8936,
      "step": 1620
    },
    {
      "epoch": 1.6192592592592594,
      "grad_norm": 1.3021979331970215,
      "learning_rate": 1.5517363214409528e-05,
      "loss": 0.9185,
      "step": 1640
    },
    {
      "epoch": 1.6390123456790122,
      "grad_norm": 1.3101555109024048,
      "learning_rate": 1.5172484814270635e-05,
      "loss": 0.9109,
      "step": 1660
    },
    {
      "epoch": 1.6587654320987655,
      "grad_norm": 1.3461183309555054,
      "learning_rate": 1.4827515185729368e-05,
      "loss": 0.9045,
      "step": 1680
    },
    {
      "epoch": 1.6785185185185185,
      "grad_norm": 1.3536429405212402,
      "learning_rate": 1.4482636785590474e-05,
      "loss": 0.8981,
      "step": 1700
    },
    {
      "epoch": 1.6785185185185185,
      "eval_loss": 0.9300320148468018,
      "eval_runtime": 402.5102,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 1700
    },
    {
      "epoch": 1.6982716049382716,
      "grad_norm": 1.2944140434265137,
      "learning_rate": 1.413803202240736e-05,
      "loss": 0.9127,
      "step": 1720
    },
    {
      "epoch": 1.7180246913580248,
      "grad_norm": 1.3316771984100342,
      "learning_rate": 1.3793883160004971e-05,
      "loss": 0.9157,
      "step": 1740
    },
    {
      "epoch": 1.7377777777777776,
      "grad_norm": 1.403693675994873,
      "learning_rate": 1.3450372221079193e-05,
      "loss": 0.8998,
      "step": 1760
    },
    {
      "epoch": 1.757530864197531,
      "grad_norm": 1.3952233791351318,
      "learning_rate": 1.3107680890923821e-05,
      "loss": 0.9139,
      "step": 1780
    },
    {
      "epoch": 1.777283950617284,
      "grad_norm": 1.3515853881835938,
      "learning_rate": 1.2765990421335937e-05,
      "loss": 0.9178,
      "step": 1800
    },
    {
      "epoch": 1.777283950617284,
      "eval_loss": 0.9266696572303772,
      "eval_runtime": 402.3456,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 1800
    },
    {
      "epoch": 1.797037037037037,
      "grad_norm": 1.3852437734603882,
      "learning_rate": 1.2425481534750591e-05,
      "loss": 0.9189,
      "step": 1820
    },
    {
      "epoch": 1.8167901234567903,
      "grad_norm": 1.372295618057251,
      "learning_rate": 1.2086334328655478e-05,
      "loss": 0.8969,
      "step": 1840
    },
    {
      "epoch": 1.836543209876543,
      "grad_norm": 1.6304378509521484,
      "learning_rate": 1.1748728180336137e-05,
      "loss": 0.903,
      "step": 1860
    },
    {
      "epoch": 1.8562962962962963,
      "grad_norm": 1.3534801006317139,
      "learning_rate": 1.141284165200207e-05,
      "loss": 0.9219,
      "step": 1880
    },
    {
      "epoch": 1.8760493827160494,
      "grad_norm": 1.4054875373840332,
      "learning_rate": 1.1078852396343955e-05,
      "loss": 0.9009,
      "step": 1900
    },
    {
      "epoch": 1.8760493827160494,
      "eval_loss": 0.923714280128479,
      "eval_runtime": 402.3854,
      "eval_samples_per_second": 4.473,
      "eval_steps_per_second": 1.118,
      "step": 1900
    },
    {
      "epoch": 1.8958024691358024,
      "grad_norm": 1.385722279548645,
      "learning_rate": 1.0746937062571928e-05,
      "loss": 0.9103,
      "step": 1920
    },
    {
      "epoch": 1.9155555555555557,
      "grad_norm": 1.3441500663757324,
      "learning_rate": 1.0417271202984614e-05,
      "loss": 0.9465,
      "step": 1940
    },
    {
      "epoch": 1.9353086419753085,
      "grad_norm": 1.5040432214736938,
      "learning_rate": 1.0090029180118293e-05,
      "loss": 0.881,
      "step": 1960
    },
    {
      "epoch": 1.9550617283950618,
      "grad_norm": 1.401288390159607,
      "learning_rate": 9.765384074525358e-06,
      "loss": 0.901,
      "step": 1980
    },
    {
      "epoch": 1.9748148148148148,
      "grad_norm": 1.439374327659607,
      "learning_rate": 9.443507593230829e-06,
      "loss": 0.904,
      "step": 2000
    },
    {
      "epoch": 1.9748148148148148,
      "eval_loss": 0.9219183921813965,
      "eval_runtime": 402.3558,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 2000
    },
    {
      "epoch": 1.9945679012345678,
      "grad_norm": 1.63518488407135,
      "learning_rate": 9.124569978915336e-06,
      "loss": 0.9222,
      "step": 2020
    },
    {
      "epoch": 2.013827160493827,
      "grad_norm": 1.3853759765625,
      "learning_rate": 8.808739919872588e-06,
      "loss": 0.8659,
      "step": 2040
    },
    {
      "epoch": 2.0335802469135804,
      "grad_norm": 1.3967058658599854,
      "learning_rate": 8.496184460788969e-06,
      "loss": 0.8865,
      "step": 2060
    },
    {
      "epoch": 2.0533333333333332,
      "grad_norm": 1.3898597955703735,
      "learning_rate": 8.187068914392464e-06,
      "loss": 0.8473,
      "step": 2080
    },
    {
      "epoch": 2.0730864197530865,
      "grad_norm": 1.4539365768432617,
      "learning_rate": 7.881556774017635e-06,
      "loss": 0.8638,
      "step": 2100
    },
    {
      "epoch": 2.0730864197530865,
      "eval_loss": 0.9233973622322083,
      "eval_runtime": 402.3351,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 2100
    },
    {
      "epoch": 2.0928395061728393,
      "grad_norm": 1.4145288467407227,
      "learning_rate": 7.579809627132857e-06,
      "loss": 0.8393,
      "step": 2120
    },
    {
      "epoch": 2.1125925925925926,
      "grad_norm": 1.436933994293213,
      "learning_rate": 7.281987069875591e-06,
      "loss": 0.8841,
      "step": 2140
    },
    {
      "epoch": 2.132345679012346,
      "grad_norm": 1.4067989587783813,
      "learning_rate": 6.988246622640921e-06,
      "loss": 0.8516,
      "step": 2160
    },
    {
      "epoch": 2.1520987654320987,
      "grad_norm": 1.5077627897262573,
      "learning_rate": 6.698743646767891e-06,
      "loss": 0.8749,
      "step": 2180
    },
    {
      "epoch": 2.171851851851852,
      "grad_norm": 1.599204182624817,
      "learning_rate": 6.413631262367882e-06,
      "loss": 0.8678,
      "step": 2200
    },
    {
      "epoch": 2.171851851851852,
      "eval_loss": 0.9224557876586914,
      "eval_runtime": 402.3427,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 2200
    },
    {
      "epoch": 2.1916049382716047,
      "grad_norm": 1.4863977432250977,
      "learning_rate": 6.1330602673382725e-06,
      "loss": 0.8517,
      "step": 2220
    },
    {
      "epoch": 2.211358024691358,
      "grad_norm": 1.3517553806304932,
      "learning_rate": 5.857179057604451e-06,
      "loss": 0.8445,
      "step": 2240
    },
    {
      "epoch": 2.2311111111111113,
      "grad_norm": 1.5938282012939453,
      "learning_rate": 5.586133548632161e-06,
      "loss": 0.8475,
      "step": 2260
    },
    {
      "epoch": 2.250864197530864,
      "grad_norm": 1.5121272802352905,
      "learning_rate": 5.320067098251815e-06,
      "loss": 0.852,
      "step": 2280
    },
    {
      "epoch": 2.2706172839506173,
      "grad_norm": 1.5417536497116089,
      "learning_rate": 5.0591204308355465e-06,
      "loss": 0.8463,
      "step": 2300
    },
    {
      "epoch": 2.2706172839506173,
      "eval_loss": 0.9209318161010742,
      "eval_runtime": 402.2865,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.119,
      "step": 2300
    },
    {
      "epoch": 2.29037037037037,
      "grad_norm": 1.4710192680358887,
      "learning_rate": 4.803431562867145e-06,
      "loss": 0.8782,
      "step": 2320
    },
    {
      "epoch": 2.3101234567901234,
      "grad_norm": 1.4649263620376587,
      "learning_rate": 4.5531357299441774e-06,
      "loss": 0.8881,
      "step": 2340
    },
    {
      "epoch": 2.3298765432098767,
      "grad_norm": 1.5059787034988403,
      "learning_rate": 4.3083653152509434e-06,
      "loss": 0.859,
      "step": 2360
    },
    {
      "epoch": 2.3496296296296295,
      "grad_norm": 1.481704831123352,
      "learning_rate": 4.069249779540108e-06,
      "loss": 0.8733,
      "step": 2380
    },
    {
      "epoch": 2.3693827160493828,
      "grad_norm": 1.5068329572677612,
      "learning_rate": 3.835915592660032e-06,
      "loss": 0.832,
      "step": 2400
    },
    {
      "epoch": 2.3693827160493828,
      "eval_loss": 0.9204065203666687,
      "eval_runtime": 402.3595,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.118,
      "step": 2400
    },
    {
      "epoch": 2.389135802469136,
      "grad_norm": 1.5572285652160645,
      "learning_rate": 3.6084861666639856e-06,
      "loss": 0.8649,
      "step": 2420
    },
    {
      "epoch": 2.408888888888889,
      "grad_norm": 1.519260287284851,
      "learning_rate": 3.387081790536677e-06,
      "loss": 0.8452,
      "step": 2440
    },
    {
      "epoch": 2.428641975308642,
      "grad_norm": 1.5407612323760986,
      "learning_rate": 3.171819566572563e-06,
      "loss": 0.8637,
      "step": 2460
    },
    {
      "epoch": 2.448395061728395,
      "grad_norm": 1.5315485000610352,
      "learning_rate": 2.962813348439669e-06,
      "loss": 0.8348,
      "step": 2480
    },
    {
      "epoch": 2.468148148148148,
      "grad_norm": 1.5446561574935913,
      "learning_rate": 2.7601736809615883e-06,
      "loss": 0.8572,
      "step": 2500
    },
    {
      "epoch": 2.468148148148148,
      "eval_loss": 0.9187188148498535,
      "eval_runtime": 402.5088,
      "eval_samples_per_second": 4.472,
      "eval_steps_per_second": 1.118,
      "step": 2500
    },
    {
      "epoch": 2.487901234567901,
      "grad_norm": 1.7411221265792847,
      "learning_rate": 2.564007741649559e-06,
      "loss": 0.8592,
      "step": 2520
    },
    {
      "epoch": 2.5076543209876543,
      "grad_norm": 1.5264136791229248,
      "learning_rate": 2.374419284015552e-06,
      "loss": 0.841,
      "step": 2540
    },
    {
      "epoch": 2.5274074074074075,
      "grad_norm": 1.6637004613876343,
      "learning_rate": 2.191508582696329e-06,
      "loss": 0.8427,
      "step": 2560
    },
    {
      "epoch": 2.5471604938271604,
      "grad_norm": 1.4840551614761353,
      "learning_rate": 2.0153723804174945e-06,
      "loss": 0.8604,
      "step": 2580
    },
    {
      "epoch": 2.5669135802469136,
      "grad_norm": 1.530154824256897,
      "learning_rate": 1.8461038368256027e-06,
      "loss": 0.8788,
      "step": 2600
    },
    {
      "epoch": 2.5669135802469136,
      "eval_loss": 0.9180880188941956,
      "eval_runtime": 402.2861,
      "eval_samples_per_second": 4.474,
      "eval_steps_per_second": 1.119,
      "step": 2600
    }
  ],
  "logging_steps": 20,
  "max_steps": 3036,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.8563300999472087e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}