{
  "best_metric": 1.0,
  "best_model_checkpoint": "/content/drive/MyDrive/Colab Notebooks/16_label_check_point/checkpoint-563",
  "epoch": 4.997333333333334,
  "eval_steps": 500,
  "global_step": 2810,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.017777777777777778,
      "grad_norm": 10.457254409790039,
      "learning_rate": 1.7793594306049826e-06,
      "loss": 11.2334,
      "step": 10
    },
    {
      "epoch": 0.035555555555555556,
      "grad_norm": 6.316349506378174,
      "learning_rate": 3.558718861209965e-06,
      "loss": 11.1807,
      "step": 20
    },
    {
      "epoch": 0.05333333333333334,
      "grad_norm": 5.264202117919922,
      "learning_rate": 5.338078291814947e-06,
      "loss": 11.1463,
      "step": 30
    },
    {
      "epoch": 0.07111111111111111,
      "grad_norm": 5.205317497253418,
      "learning_rate": 7.11743772241993e-06,
      "loss": 11.0929,
      "step": 40
    },
    {
      "epoch": 0.08888888888888889,
      "grad_norm": 4.696351528167725,
      "learning_rate": 8.896797153024912e-06,
      "loss": 11.1015,
      "step": 50
    },
    {
      "epoch": 0.10666666666666667,
      "grad_norm": 5.9699320793151855,
      "learning_rate": 1.0676156583629894e-05,
      "loss": 11.0795,
      "step": 60
    },
    {
      "epoch": 0.12444444444444444,
      "grad_norm": 7.235191822052002,
      "learning_rate": 1.2455516014234877e-05,
      "loss": 11.046,
      "step": 70
    },
    {
      "epoch": 0.14222222222222222,
      "grad_norm": 14.865583419799805,
      "learning_rate": 1.423487544483986e-05,
      "loss": 11.0113,
      "step": 80
    },
    {
      "epoch": 0.16,
      "grad_norm": 17.412696838378906,
      "learning_rate": 1.601423487544484e-05,
      "loss": 10.8502,
      "step": 90
    },
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 20.41786003112793,
      "learning_rate": 1.7793594306049825e-05,
      "loss": 10.5185,
      "step": 100
    },
    {
      "epoch": 0.19555555555555557,
      "grad_norm": 24.58249282836914,
      "learning_rate": 1.9572953736654805e-05,
      "loss": 9.8301,
      "step": 110
    },
    {
      "epoch": 0.21333333333333335,
      "grad_norm": 43.27064895629883,
      "learning_rate": 2.135231316725979e-05,
      "loss": 8.8636,
      "step": 120
    },
    {
      "epoch": 0.2311111111111111,
      "grad_norm": 46.2359733581543,
      "learning_rate": 2.313167259786477e-05,
      "loss": 7.4813,
      "step": 130
    },
    {
      "epoch": 0.24888888888888888,
      "grad_norm": 71.36530303955078,
      "learning_rate": 2.4911032028469753e-05,
      "loss": 6.0609,
      "step": 140
    },
    {
      "epoch": 0.26666666666666666,
      "grad_norm": 61.219024658203125,
      "learning_rate": 2.669039145907473e-05,
      "loss": 4.5934,
      "step": 150
    },
    {
      "epoch": 0.28444444444444444,
      "grad_norm": 52.64271545410156,
      "learning_rate": 2.846975088967972e-05,
      "loss": 3.3049,
      "step": 160
    },
    {
      "epoch": 0.3022222222222222,
      "grad_norm": 40.045623779296875,
      "learning_rate": 3.02491103202847e-05,
      "loss": 2.0759,
      "step": 170
    },
    {
      "epoch": 0.32,
      "grad_norm": 32.63826370239258,
      "learning_rate": 3.202846975088968e-05,
      "loss": 1.2791,
      "step": 180
    },
    {
      "epoch": 0.3377777777777778,
      "grad_norm": 39.16119384765625,
      "learning_rate": 3.380782918149467e-05,
      "loss": 0.9052,
      "step": 190
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 36.06990051269531,
      "learning_rate": 3.558718861209965e-05,
      "loss": 0.6508,
      "step": 200
    },
    {
      "epoch": 0.37333333333333335,
      "grad_norm": 22.00220489501953,
      "learning_rate": 3.736654804270463e-05,
      "loss": 0.6293,
      "step": 210
    },
    {
      "epoch": 0.39111111111111113,
      "grad_norm": 27.334341049194336,
      "learning_rate": 3.914590747330961e-05,
      "loss": 0.5774,
      "step": 220
    },
    {
      "epoch": 0.4088888888888889,
      "grad_norm": 21.130746841430664,
      "learning_rate": 4.09252669039146e-05,
      "loss": 0.4318,
      "step": 230
    },
    {
      "epoch": 0.4266666666666667,
      "grad_norm": 21.37102508544922,
      "learning_rate": 4.270462633451958e-05,
      "loss": 0.414,
      "step": 240
    },
    {
      "epoch": 0.4444444444444444,
      "grad_norm": 86.8355712890625,
      "learning_rate": 4.448398576512456e-05,
      "loss": 0.4338,
      "step": 250
    },
    {
      "epoch": 0.4622222222222222,
      "grad_norm": 9.731348037719727,
      "learning_rate": 4.626334519572954e-05,
      "loss": 0.3978,
      "step": 260
    },
    {
      "epoch": 0.48,
      "grad_norm": 23.59229278564453,
      "learning_rate": 4.8042704626334526e-05,
      "loss": 0.2971,
      "step": 270
    },
    {
      "epoch": 0.49777777777777776,
      "grad_norm": 10.458292961120605,
      "learning_rate": 4.9822064056939506e-05,
      "loss": 0.3015,
      "step": 280
    },
    {
      "epoch": 0.5155555555555555,
      "grad_norm": 13.044571876525879,
      "learning_rate": 4.9822064056939506e-05,
      "loss": 0.2013,
      "step": 290
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 30.592700958251953,
      "learning_rate": 4.962435745353895e-05,
      "loss": 0.182,
      "step": 300
    },
    {
      "epoch": 0.5511111111111111,
      "grad_norm": 26.582555770874023,
      "learning_rate": 4.9426650850138396e-05,
      "loss": 0.2276,
      "step": 310
    },
    {
      "epoch": 0.5688888888888889,
      "grad_norm": 20.704526901245117,
      "learning_rate": 4.9228944246737844e-05,
      "loss": 0.1836,
      "step": 320
    },
    {
      "epoch": 0.5866666666666667,
      "grad_norm": 29.400476455688477,
      "learning_rate": 4.903123764333729e-05,
      "loss": 0.2921,
      "step": 330
    },
    {
      "epoch": 0.6044444444444445,
      "grad_norm": 23.031789779663086,
      "learning_rate": 4.8833531039936733e-05,
      "loss": 0.0848,
      "step": 340
    },
    {
      "epoch": 0.6222222222222222,
      "grad_norm": 0.9449447393417358,
      "learning_rate": 4.863582443653618e-05,
      "loss": 0.0741,
      "step": 350
    },
    {
      "epoch": 0.64,
      "grad_norm": 20.334348678588867,
      "learning_rate": 4.843811783313563e-05,
      "loss": 0.1798,
      "step": 360
    },
    {
      "epoch": 0.6577777777777778,
      "grad_norm": 13.533489227294922,
      "learning_rate": 4.824041122973508e-05,
      "loss": 0.1007,
      "step": 370
    },
    {
      "epoch": 0.6755555555555556,
      "grad_norm": 1.5982263088226318,
      "learning_rate": 4.8042704626334526e-05,
      "loss": 0.1016,
      "step": 380
    },
    {
      "epoch": 0.6933333333333334,
      "grad_norm": 2.323336362838745,
      "learning_rate": 4.784499802293397e-05,
      "loss": 0.1131,
      "step": 390
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 24.458276748657227,
      "learning_rate": 4.7647291419533415e-05,
      "loss": 0.069,
      "step": 400
    },
    {
      "epoch": 0.7288888888888889,
      "grad_norm": 29.530794143676758,
      "learning_rate": 4.7449584816132864e-05,
      "loss": 0.2415,
      "step": 410
    },
    {
      "epoch": 0.7466666666666667,
      "grad_norm": 2.0636749267578125,
      "learning_rate": 4.725187821273231e-05,
      "loss": 0.1173,
      "step": 420
    },
    {
      "epoch": 0.7644444444444445,
      "grad_norm": 2.225900888442993,
      "learning_rate": 4.705417160933175e-05,
      "loss": 0.0918,
      "step": 430
    },
    {
      "epoch": 0.7822222222222223,
      "grad_norm": 7.136375904083252,
      "learning_rate": 4.68564650059312e-05,
      "loss": 0.1589,
      "step": 440
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.6402971744537354,
      "learning_rate": 4.665875840253064e-05,
      "loss": 0.2547,
      "step": 450
    },
    {
      "epoch": 0.8177777777777778,
      "grad_norm": 12.275321960449219,
      "learning_rate": 4.64610517991301e-05,
      "loss": 0.1082,
      "step": 460
    },
    {
      "epoch": 0.8355555555555556,
      "grad_norm": 0.6636475920677185,
      "learning_rate": 4.626334519572954e-05,
      "loss": 0.0986,
      "step": 470
    },
    {
      "epoch": 0.8533333333333334,
      "grad_norm": 27.784332275390625,
      "learning_rate": 4.606563859232899e-05,
      "loss": 0.1056,
      "step": 480
    },
    {
      "epoch": 0.8711111111111111,
      "grad_norm": 22.047527313232422,
      "learning_rate": 4.586793198892843e-05,
      "loss": 0.1654,
      "step": 490
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 15.613493919372559,
      "learning_rate": 4.5670225385527876e-05,
      "loss": 0.1517,
      "step": 500
    },
    {
      "epoch": 0.9066666666666666,
      "grad_norm": 20.039813995361328,
      "learning_rate": 4.5472518782127324e-05,
      "loss": 0.1283,
      "step": 510
    },
    {
      "epoch": 0.9244444444444444,
      "grad_norm": 21.469423294067383,
      "learning_rate": 4.527481217872677e-05,
      "loss": 0.0707,
      "step": 520
    },
    {
      "epoch": 0.9422222222222222,
      "grad_norm": 38.251953125,
      "learning_rate": 4.5077105575326214e-05,
      "loss": 0.1116,
      "step": 530
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.317658424377441,
      "learning_rate": 4.487939897192566e-05,
      "loss": 0.138,
      "step": 540
    },
    {
      "epoch": 0.9777777777777777,
      "grad_norm": 5.918329238891602,
      "learning_rate": 4.468169236852511e-05,
      "loss": 0.1053,
      "step": 550
    },
    {
      "epoch": 0.9955555555555555,
      "grad_norm": 12.31584358215332,
      "learning_rate": 4.448398576512456e-05,
      "loss": 0.1354,
      "step": 560
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 1.0,
      "eval_loss": 8.097552927210927e-05,
      "eval_runtime": 74.3955,
      "eval_samples_per_second": 107.533,
      "eval_steps_per_second": 3.36,
      "step": 563
    },
    {
      "epoch": 1.0124444444444445,
      "grad_norm": 10.894996643066406,
      "learning_rate": 4.4286279161724006e-05,
      "loss": 0.1134,
      "step": 570
    },
    {
      "epoch": 1.0302222222222222,
      "grad_norm": 16.037940979003906,
      "learning_rate": 4.408857255832345e-05,
      "loss": 0.0917,
      "step": 580
    },
    {
      "epoch": 1.048,
      "grad_norm": 0.164012148976326,
      "learning_rate": 4.3890865954922896e-05,
      "loss": 0.0816,
      "step": 590
    },
    {
      "epoch": 1.0657777777777777,
      "grad_norm": 0.05601898953318596,
      "learning_rate": 4.3693159351522344e-05,
      "loss": 0.0778,
      "step": 600
    },
    {
      "epoch": 1.0835555555555556,
      "grad_norm": 9.240290641784668,
      "learning_rate": 4.349545274812179e-05,
      "loss": 0.0893,
      "step": 610
    },
    {
      "epoch": 1.1013333333333333,
      "grad_norm": 8.448566436767578,
      "learning_rate": 4.3297746144721233e-05,
      "loss": 0.0905,
      "step": 620
    },
    {
      "epoch": 1.1191111111111112,
      "grad_norm": 1.9537012577056885,
      "learning_rate": 4.310003954132068e-05,
      "loss": 0.1561,
      "step": 630
    },
    {
      "epoch": 1.1368888888888888,
      "grad_norm": 25.13317108154297,
      "learning_rate": 4.290233293792013e-05,
      "loss": 0.0856,
      "step": 640
    },
    {
      "epoch": 1.1546666666666667,
      "grad_norm": 1.3837047815322876,
      "learning_rate": 4.270462633451958e-05,
      "loss": 0.0962,
      "step": 650
    },
    {
      "epoch": 1.1724444444444444,
      "grad_norm": 46.560367584228516,
      "learning_rate": 4.250691973111902e-05,
      "loss": 0.1009,
      "step": 660
    },
    {
      "epoch": 1.1902222222222223,
      "grad_norm": 42.40678787231445,
      "learning_rate": 4.230921312771847e-05,
      "loss": 0.1027,
      "step": 670
    },
    {
      "epoch": 1.208,
      "grad_norm": 1.2823681831359863,
      "learning_rate": 4.211150652431791e-05,
      "loss": 0.0609,
      "step": 680
    },
    {
      "epoch": 1.2257777777777779,
      "grad_norm": 3.9979896545410156,
      "learning_rate": 4.1913799920917364e-05,
      "loss": 0.0884,
      "step": 690
    },
    {
      "epoch": 1.2435555555555555,
      "grad_norm": 23.4843692779541,
      "learning_rate": 4.1716093317516805e-05,
      "loss": 0.049,
      "step": 700
    },
    {
      "epoch": 1.2613333333333334,
      "grad_norm": 0.9539620876312256,
      "learning_rate": 4.151838671411625e-05,
      "loss": 0.0481,
      "step": 710
    },
    {
      "epoch": 1.279111111111111,
      "grad_norm": 0.23285511136054993,
      "learning_rate": 4.13206801107157e-05,
      "loss": 0.0388,
      "step": 720
    },
    {
      "epoch": 1.2968888888888888,
      "grad_norm": 24.404285430908203,
      "learning_rate": 4.112297350731515e-05,
      "loss": 0.0425,
      "step": 730
    },
    {
      "epoch": 1.3146666666666667,
      "grad_norm": 3.143155813217163,
      "learning_rate": 4.09252669039146e-05,
      "loss": 0.0313,
      "step": 740
    },
    {
      "epoch": 1.3324444444444445,
      "grad_norm": 0.3580031991004944,
      "learning_rate": 4.072756030051404e-05,
      "loss": 0.0466,
      "step": 750
    },
    {
      "epoch": 1.3502222222222222,
      "grad_norm": 39.96261978149414,
      "learning_rate": 4.052985369711349e-05,
      "loss": 0.0706,
      "step": 760
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 22.012971878051758,
      "learning_rate": 4.033214709371293e-05,
      "loss": 0.1249,
      "step": 770
    },
    {
      "epoch": 1.3857777777777778,
      "grad_norm": 0.20229700207710266,
      "learning_rate": 4.013444049031238e-05,
      "loss": 0.1089,
      "step": 780
    },
    {
      "epoch": 1.4035555555555557,
      "grad_norm": 36.518348693847656,
      "learning_rate": 3.9936733886911825e-05,
      "loss": 0.0514,
      "step": 790
    },
    {
      "epoch": 1.4213333333333333,
      "grad_norm": 0.011868222616612911,
      "learning_rate": 3.973902728351127e-05,
      "loss": 0.0298,
      "step": 800
    },
    {
      "epoch": 1.439111111111111,
      "grad_norm": 23.455787658691406,
      "learning_rate": 3.9541320680110714e-05,
      "loss": 0.0365,
      "step": 810
    },
    {
      "epoch": 1.456888888888889,
      "grad_norm": 0.5454981923103333,
      "learning_rate": 3.934361407671016e-05,
      "loss": 0.0603,
      "step": 820
    },
    {
      "epoch": 1.4746666666666668,
      "grad_norm": 0.2658223509788513,
      "learning_rate": 3.914590747330961e-05,
      "loss": 0.0364,
      "step": 830
    },
    {
      "epoch": 1.4924444444444445,
      "grad_norm": 32.6451301574707,
      "learning_rate": 3.894820086990906e-05,
      "loss": 0.1105,
      "step": 840
    },
    {
      "epoch": 1.5102222222222221,
      "grad_norm": 0.37322714924812317,
      "learning_rate": 3.87504942665085e-05,
      "loss": 0.0102,
      "step": 850
    },
    {
      "epoch": 1.528,
      "grad_norm": 1.2302775382995605,
      "learning_rate": 3.855278766310795e-05,
      "loss": 0.093,
      "step": 860
    },
    {
      "epoch": 1.545777777777778,
      "grad_norm": 0.11981203407049179,
      "learning_rate": 3.8355081059707396e-05,
      "loss": 0.0721,
      "step": 870
    },
    {
      "epoch": 1.5635555555555556,
      "grad_norm": 0.09180541336536407,
      "learning_rate": 3.8157374456306844e-05,
      "loss": 0.0408,
      "step": 880
    },
    {
      "epoch": 1.5813333333333333,
      "grad_norm": 29.872051239013672,
      "learning_rate": 3.7959667852906285e-05,
      "loss": 0.0393,
      "step": 890
    },
    {
      "epoch": 1.5991111111111111,
      "grad_norm": 1.0710923671722412,
      "learning_rate": 3.7761961249505734e-05,
      "loss": 0.0023,
      "step": 900
    },
    {
      "epoch": 1.616888888888889,
      "grad_norm": 20.386173248291016,
      "learning_rate": 3.756425464610518e-05,
      "loss": 0.0605,
      "step": 910
    },
    {
      "epoch": 1.6346666666666667,
      "grad_norm": 0.03605956584215164,
      "learning_rate": 3.736654804270463e-05,
      "loss": 0.0059,
      "step": 920
    },
    {
      "epoch": 1.6524444444444444,
      "grad_norm": 0.38812369108200073,
      "learning_rate": 3.716884143930408e-05,
      "loss": 0.0819,
      "step": 930
    },
    {
      "epoch": 1.6702222222222223,
      "grad_norm": 27.319766998291016,
      "learning_rate": 3.697113483590352e-05,
      "loss": 0.0201,
      "step": 940
    },
    {
      "epoch": 1.688,
      "grad_norm": 20.58792495727539,
      "learning_rate": 3.677342823250297e-05,
      "loss": 0.0678,
      "step": 950
    },
    {
      "epoch": 1.7057777777777776,
      "grad_norm": 0.33605310320854187,
      "learning_rate": 3.6575721629102416e-05,
      "loss": 0.0027,
      "step": 960
    },
    {
      "epoch": 1.7235555555555555,
      "grad_norm": 0.13025720417499542,
      "learning_rate": 3.6378015025701864e-05,
      "loss": 0.0029,
      "step": 970
    },
    {
      "epoch": 1.7413333333333334,
      "grad_norm": 31.07040023803711,
      "learning_rate": 3.6180308422301305e-05,
      "loss": 0.0626,
      "step": 980
    },
    {
      "epoch": 1.759111111111111,
      "grad_norm": 0.055677346885204315,
      "learning_rate": 3.598260181890075e-05,
      "loss": 0.0079,
      "step": 990
    },
    {
      "epoch": 1.7768888888888887,
      "grad_norm": 0.012918527238070965,
      "learning_rate": 3.5784895215500194e-05,
      "loss": 0.0308,
      "step": 1000
    },
    {
      "epoch": 1.7946666666666666,
      "grad_norm": 12.896405220031738,
      "learning_rate": 3.558718861209965e-05,
      "loss": 0.0551,
      "step": 1010
    },
    {
      "epoch": 1.8124444444444445,
      "grad_norm": 0.0037423851899802685,
      "learning_rate": 3.538948200869909e-05,
      "loss": 0.0437,
      "step": 1020
    },
    {
      "epoch": 1.8302222222222222,
      "grad_norm": 0.2329370528459549,
      "learning_rate": 3.519177540529854e-05,
      "loss": 0.0281,
      "step": 1030
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 0.027233602479100227,
      "learning_rate": 3.499406880189798e-05,
      "loss": 0.0136,
      "step": 1040
    },
    {
      "epoch": 1.8657777777777778,
      "grad_norm": 0.1924924999475479,
      "learning_rate": 3.4796362198497435e-05,
      "loss": 0.056,
      "step": 1050
    },
    {
      "epoch": 1.8835555555555556,
      "grad_norm": 0.02651727944612503,
      "learning_rate": 3.4598655595096876e-05,
      "loss": 0.0647,
      "step": 1060
    },
    {
      "epoch": 1.9013333333333333,
      "grad_norm": 38.13447570800781,
      "learning_rate": 3.4400948991696325e-05,
      "loss": 0.0216,
      "step": 1070
    },
    {
      "epoch": 1.919111111111111,
      "grad_norm": 11.278813362121582,
      "learning_rate": 3.420324238829577e-05,
      "loss": 0.0395,
      "step": 1080
    },
    {
      "epoch": 1.9368888888888889,
      "grad_norm": 0.5866456031799316,
      "learning_rate": 3.4005535784895214e-05,
      "loss": 0.0433,
      "step": 1090
    },
    {
      "epoch": 1.9546666666666668,
      "grad_norm": 2.6308796405792236,
      "learning_rate": 3.380782918149467e-05,
      "loss": 0.0118,
      "step": 1100
    },
    {
      "epoch": 1.9724444444444444,
      "grad_norm": 0.1729055494070053,
      "learning_rate": 3.361012257809411e-05,
      "loss": 0.0251,
      "step": 1110
    },
    {
      "epoch": 1.9902222222222221,
      "grad_norm": 10.363531112670898,
      "learning_rate": 3.341241597469356e-05,
      "loss": 0.0656,
      "step": 1120
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 1.0,
      "eval_loss": 2.5387274945387617e-05,
      "eval_runtime": 75.7395,
      "eval_samples_per_second": 105.625,
      "eval_steps_per_second": 3.301,
      "step": 1126
    },
    {
      "epoch": 2.010666666666667,
      "grad_norm": 29.318574905395508,
      "learning_rate": 3.3214709371293e-05,
      "loss": 0.2308,
      "step": 1130
    },
    {
      "epoch": 2.0284444444444443,
      "grad_norm": 67.69547271728516,
      "learning_rate": 3.301700276789245e-05,
      "loss": 0.0604,
      "step": 1140
    },
    {
      "epoch": 2.046222222222222,
      "grad_norm": 0.18736213445663452,
      "learning_rate": 3.2819296164491896e-05,
      "loss": 0.0546,
      "step": 1150
    },
    {
      "epoch": 2.064,
      "grad_norm": 15.664798736572266,
      "learning_rate": 3.2621589561091344e-05,
      "loss": 0.0304,
      "step": 1160
    },
    {
      "epoch": 2.081777777777778,
      "grad_norm": 0.13788799941539764,
      "learning_rate": 3.2423882957690785e-05,
      "loss": 0.0029,
      "step": 1170
    },
    {
      "epoch": 2.0995555555555554,
      "grad_norm": 4.969343185424805,
      "learning_rate": 3.2226176354290234e-05,
      "loss": 0.0099,
      "step": 1180
    },
    {
      "epoch": 2.1173333333333333,
      "grad_norm": 1.4300692081451416,
      "learning_rate": 3.202846975088968e-05,
      "loss": 0.0181,
      "step": 1190
    },
    {
      "epoch": 2.135111111111111,
      "grad_norm": 0.019833851605653763,
      "learning_rate": 3.183076314748913e-05,
      "loss": 0.0107,
      "step": 1200
    },
    {
      "epoch": 2.152888888888889,
      "grad_norm": 0.01474601961672306,
      "learning_rate": 3.163305654408857e-05,
      "loss": 0.0169,
      "step": 1210
    },
    {
      "epoch": 2.1706666666666665,
      "grad_norm": 0.004971742630004883,
      "learning_rate": 3.143534994068802e-05,
      "loss": 0.0254,
      "step": 1220
    },
    {
      "epoch": 2.1884444444444444,
      "grad_norm": 0.42502257227897644,
      "learning_rate": 3.123764333728747e-05,
      "loss": 0.0623,
      "step": 1230
    },
    {
      "epoch": 2.2062222222222223,
      "grad_norm": 30.118955612182617,
      "learning_rate": 3.1039936733886916e-05,
      "loss": 0.0456,
      "step": 1240
    },
    {
      "epoch": 2.224,
      "grad_norm": 7.1990132331848145,
      "learning_rate": 3.0842230130486364e-05,
      "loss": 0.0468,
      "step": 1250
    },
    {
      "epoch": 2.2417777777777776,
      "grad_norm": 0.021625256165862083,
      "learning_rate": 3.0644523527085805e-05,
      "loss": 0.0375,
      "step": 1260
    },
    {
      "epoch": 2.2595555555555555,
      "grad_norm": 0.08051316440105438,
      "learning_rate": 3.044681692368525e-05,
      "loss": 0.0097,
      "step": 1270
    },
    {
      "epoch": 2.2773333333333334,
      "grad_norm": 38.26923751831055,
      "learning_rate": 3.02491103202847e-05,
      "loss": 0.0247,
      "step": 1280
    },
    {
      "epoch": 2.295111111111111,
      "grad_norm": 1.5090163946151733,
      "learning_rate": 3.0051403716884146e-05,
      "loss": 0.0207,
      "step": 1290
    },
    {
      "epoch": 2.3128888888888888,
      "grad_norm": 14.290279388427734,
      "learning_rate": 2.985369711348359e-05,
      "loss": 0.0041,
      "step": 1300
    },
    {
      "epoch": 2.3306666666666667,
      "grad_norm": 0.025663571432232857,
      "learning_rate": 2.9655990510083035e-05,
      "loss": 0.0013,
      "step": 1310
    },
    {
      "epoch": 2.3484444444444446,
      "grad_norm": 13.073821067810059,
      "learning_rate": 2.9458283906682484e-05,
      "loss": 0.006,
      "step": 1320
    },
    {
      "epoch": 2.3662222222222224,
      "grad_norm": 0.03257250785827637,
      "learning_rate": 2.9260577303281932e-05,
      "loss": 0.0383,
      "step": 1330
    },
    {
      "epoch": 2.384,
      "grad_norm": 2.0085232257843018,
      "learning_rate": 2.906287069988138e-05,
      "loss": 0.0552,
      "step": 1340
    },
    {
      "epoch": 2.401777777777778,
      "grad_norm": 19.939329147338867,
      "learning_rate": 2.8865164096480825e-05,
      "loss": 0.0714,
      "step": 1350
    },
    {
      "epoch": 2.4195555555555557,
      "grad_norm": 0.028250480070710182,
      "learning_rate": 2.866745749308027e-05,
      "loss": 0.0227,
      "step": 1360
    },
    {
      "epoch": 2.437333333333333,
      "grad_norm": 0.6903110146522522,
      "learning_rate": 2.846975088967972e-05,
      "loss": 0.0085,
      "step": 1370
    },
    {
      "epoch": 2.455111111111111,
      "grad_norm": 0.22341494262218475,
      "learning_rate": 2.8272044286279166e-05,
      "loss": 0.0093,
      "step": 1380
    },
    {
      "epoch": 2.472888888888889,
      "grad_norm": 0.015194721519947052,
      "learning_rate": 2.807433768287861e-05,
      "loss": 0.0413,
      "step": 1390
    },
    {
      "epoch": 2.490666666666667,
      "grad_norm": 0.035975806415081024,
      "learning_rate": 2.7876631079478055e-05,
      "loss": 0.0013,
      "step": 1400
    },
    {
      "epoch": 2.5084444444444447,
      "grad_norm": 0.32504796981811523,
      "learning_rate": 2.76789244760775e-05,
      "loss": 0.0547,
      "step": 1410
    },
    {
      "epoch": 2.526222222222222,
      "grad_norm": 0.34928593039512634,
      "learning_rate": 2.748121787267695e-05,
      "loss": 0.0556,
      "step": 1420
    },
    {
      "epoch": 2.544,
      "grad_norm": 0.045603252947330475,
      "learning_rate": 2.7283511269276396e-05,
      "loss": 0.0156,
      "step": 1430
    },
    {
      "epoch": 2.561777777777778,
      "grad_norm": 11.350424766540527,
      "learning_rate": 2.708580466587584e-05,
      "loss": 0.011,
      "step": 1440
    },
    {
      "epoch": 2.5795555555555554,
      "grad_norm": 0.07788264751434326,
      "learning_rate": 2.6888098062475286e-05,
      "loss": 0.0607,
      "step": 1450
    },
    {
      "epoch": 2.5973333333333333,
      "grad_norm": 0.06617221236228943,
      "learning_rate": 2.669039145907473e-05,
      "loss": 0.0102,
      "step": 1460
    },
    {
      "epoch": 2.615111111111111,
      "grad_norm": 34.64754867553711,
      "learning_rate": 2.6492684855674182e-05,
      "loss": 0.0268,
      "step": 1470
    },
    {
      "epoch": 2.632888888888889,
      "grad_norm": 9.72877311706543,
      "learning_rate": 2.6294978252273626e-05,
      "loss": 0.0051,
      "step": 1480
    },
    {
      "epoch": 2.6506666666666665,
      "grad_norm": 21.619274139404297,
      "learning_rate": 2.609727164887307e-05,
      "loss": 0.0092,
      "step": 1490
    },
    {
      "epoch": 2.6684444444444444,
      "grad_norm": 4.081634521484375,
      "learning_rate": 2.589956504547252e-05,
      "loss": 0.0369,
      "step": 1500
    },
    {
      "epoch": 2.6862222222222223,
      "grad_norm": 0.009345272555947304,
      "learning_rate": 2.5701858442071967e-05,
      "loss": 0.018,
      "step": 1510
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 73.03565979003906,
      "learning_rate": 2.5504151838671416e-05,
      "loss": 0.0945,
      "step": 1520
    },
    {
      "epoch": 2.7217777777777776,
      "grad_norm": 1.0712828636169434,
      "learning_rate": 2.530644523527086e-05,
      "loss": 0.0156,
      "step": 1530
    },
    {
      "epoch": 2.7395555555555555,
      "grad_norm": 0.023015221580863,
      "learning_rate": 2.5108738631870305e-05,
      "loss": 0.0408,
      "step": 1540
    },
    {
      "epoch": 2.7573333333333334,
      "grad_norm": 3.0739543437957764,
      "learning_rate": 2.4911032028469753e-05,
      "loss": 0.0384,
      "step": 1550
    },
    {
      "epoch": 2.7751111111111113,
      "grad_norm": 0.017695285379886627,
      "learning_rate": 2.4713325425069198e-05,
      "loss": 0.034,
      "step": 1560
    },
    {
      "epoch": 2.7928888888888888,
      "grad_norm": 0.013055549003183842,
      "learning_rate": 2.4515618821668646e-05,
      "loss": 0.074,
      "step": 1570
    },
    {
      "epoch": 2.8106666666666666,
      "grad_norm": 6.4839582443237305,
      "learning_rate": 2.431791221826809e-05,
      "loss": 0.0016,
      "step": 1580
    },
    {
      "epoch": 2.8284444444444445,
      "grad_norm": 0.17747992277145386,
      "learning_rate": 2.412020561486754e-05,
      "loss": 0.0579,
      "step": 1590
    },
    {
      "epoch": 2.846222222222222,
      "grad_norm": 0.07140109688043594,
      "learning_rate": 2.3922499011466984e-05,
      "loss": 0.0303,
      "step": 1600
    },
    {
      "epoch": 2.864,
      "grad_norm": 0.0027039784472435713,
      "learning_rate": 2.3724792408066432e-05,
      "loss": 0.0337,
      "step": 1610
    },
    {
      "epoch": 2.8817777777777778,
      "grad_norm": 0.015552740544080734,
      "learning_rate": 2.3527085804665877e-05,
      "loss": 0.0131,
      "step": 1620
    },
    {
      "epoch": 2.8995555555555557,
      "grad_norm": 0.014052975922822952,
      "learning_rate": 2.332937920126532e-05,
      "loss": 0.0434,
      "step": 1630
    },
    {
      "epoch": 2.9173333333333336,
      "grad_norm": 0.7165421843528748,
      "learning_rate": 2.313167259786477e-05,
      "loss": 0.0069,
      "step": 1640
    },
    {
      "epoch": 2.935111111111111,
      "grad_norm": 0.020382430404424667,
      "learning_rate": 2.2933965994464214e-05,
      "loss": 0.0419,
      "step": 1650
    },
    {
      "epoch": 2.952888888888889,
      "grad_norm": 0.07127852737903595,
      "learning_rate": 2.2736259391063662e-05,
      "loss": 0.0303,
      "step": 1660
    },
    {
      "epoch": 2.970666666666667,
      "grad_norm": 2.5529978275299072,
      "learning_rate": 2.2538552787663107e-05,
      "loss": 0.0063,
      "step": 1670
    },
    {
      "epoch": 2.9884444444444442,
      "grad_norm": 0.0018354392377659678,
      "learning_rate": 2.2340846184262555e-05,
      "loss": 0.0323,
      "step": 1680
    },
    {
      "epoch": 2.999111111111111,
      "eval_accuracy": 1.0,
      "eval_loss": 6.3951174524845555e-06,
      "eval_runtime": 39.8564,
      "eval_samples_per_second": 200.721,
      "eval_steps_per_second": 6.273,
      "step": 1686
    },
    {
      "epoch": 3.007111111111111,
      "grad_norm": 0.0571288987994194,
      "learning_rate": 2.2143139580862003e-05,
      "loss": 0.0435,
      "step": 1690
    },
    {
      "epoch": 3.024888888888889,
      "grad_norm": 0.015790535137057304,
      "learning_rate": 2.1945432977461448e-05,
      "loss": 0.0058,
      "step": 1700
    },
    {
      "epoch": 3.042666666666667,
      "grad_norm": 0.9516779184341431,
      "learning_rate": 2.1747726374060896e-05,
      "loss": 0.0109,
      "step": 1710
    },
    {
      "epoch": 3.0604444444444443,
      "grad_norm": 0.05595998093485832,
      "learning_rate": 2.155001977066034e-05,
      "loss": 0.01,
      "step": 1720
    },
    {
      "epoch": 3.078222222222222,
      "grad_norm": 0.07043913751840591,
      "learning_rate": 2.135231316725979e-05,
      "loss": 0.0017,
      "step": 1730
    },
    {
      "epoch": 3.096,
      "grad_norm": 0.00207032123580575,
      "learning_rate": 2.1154606563859234e-05,
      "loss": 0.0026,
      "step": 1740
    },
    {
      "epoch": 3.113777777777778,
      "grad_norm": 0.0049515170976519585,
      "learning_rate": 2.0956899960458682e-05,
      "loss": 0.0049,
      "step": 1750
    },
    {
      "epoch": 3.1315555555555554,
      "grad_norm": 0.08053428679704666,
      "learning_rate": 2.0759193357058127e-05,
      "loss": 0.0169,
      "step": 1760
    },
    {
      "epoch": 3.1493333333333333,
      "grad_norm": 0.0009357993258163333,
      "learning_rate": 2.0561486753657575e-05,
      "loss": 0.0124,
      "step": 1770
    },
    {
      "epoch": 3.167111111111111,
      "grad_norm": 0.014298039488494396,
      "learning_rate": 2.036378015025702e-05,
      "loss": 0.0007,
      "step": 1780
    },
    {
      "epoch": 3.1848888888888887,
      "grad_norm": 1.3868434429168701,
      "learning_rate": 2.0166073546856464e-05,
      "loss": 0.0122,
      "step": 1790
    },
    {
      "epoch": 3.2026666666666666,
      "grad_norm": 0.0033609354868531227,
      "learning_rate": 1.9968366943455912e-05,
      "loss": 0.0097,
      "step": 1800
    },
    {
      "epoch": 3.2204444444444444,
      "grad_norm": 10.693577766418457,
      "learning_rate": 1.9770660340055357e-05,
      "loss": 0.0637,
      "step": 1810
    },
    {
      "epoch": 3.2382222222222223,
      "grad_norm": 0.006601002067327499,
      "learning_rate": 1.9572953736654805e-05,
      "loss": 0.0718,
      "step": 1820
    },
    {
      "epoch": 3.2560000000000002,
      "grad_norm": 0.03210904076695442,
      "learning_rate": 1.937524713325425e-05,
      "loss": 0.0014,
      "step": 1830
    },
    {
      "epoch": 3.2737777777777777,
      "grad_norm": 0.007109949365258217,
      "learning_rate": 1.9177540529853698e-05,
      "loss": 0.0065,
      "step": 1840
    },
    {
      "epoch": 3.2915555555555556,
      "grad_norm": 0.0019361014710739255,
      "learning_rate": 1.8979833926453143e-05,
      "loss": 0.0008,
      "step": 1850
    },
    {
      "epoch": 3.3093333333333335,
      "grad_norm": 0.08277800679206848,
      "learning_rate": 1.878212732305259e-05,
      "loss": 0.0004,
      "step": 1860
    },
    {
      "epoch": 3.327111111111111,
      "grad_norm": 0.003757915459573269,
      "learning_rate": 1.858442071965204e-05,
      "loss": 0.0074,
      "step": 1870
    },
    {
      "epoch": 3.344888888888889,
      "grad_norm": 13.0957670211792,
      "learning_rate": 1.8386714116251484e-05,
      "loss": 0.0177,
      "step": 1880
    },
    {
      "epoch": 3.3626666666666667,
      "grad_norm": 0.07693179696798325,
      "learning_rate": 1.8189007512850932e-05,
      "loss": 0.0093,
      "step": 1890
    },
    {
      "epoch": 3.3804444444444446,
      "grad_norm": 0.007269900757819414,
      "learning_rate": 1.7991300909450377e-05,
      "loss": 0.0004,
      "step": 1900
    },
    {
      "epoch": 3.398222222222222,
      "grad_norm": 0.149241641163826,
      "learning_rate": 1.7793594306049825e-05,
      "loss": 0.0008,
      "step": 1910
    },
    {
      "epoch": 3.416,
      "grad_norm": 0.13814906775951385,
      "learning_rate": 1.759588770264927e-05,
      "loss": 0.0228,
      "step": 1920
    },
    {
      "epoch": 3.433777777777778,
      "grad_norm": 0.012578528374433517,
      "learning_rate": 1.7398181099248718e-05,
      "loss": 0.0031,
      "step": 1930
    },
    {
      "epoch": 3.4515555555555557,
      "grad_norm": 0.002241666428744793,
      "learning_rate": 1.7200474495848162e-05,
      "loss": 0.0018,
      "step": 1940
    },
    {
      "epoch": 3.469333333333333,
      "grad_norm": 0.005206266883760691,
      "learning_rate": 1.7002767892447607e-05,
      "loss": 0.0469,
      "step": 1950
    },
    {
      "epoch": 3.487111111111111,
      "grad_norm": 11.583967208862305,
      "learning_rate": 1.6805061289047055e-05,
      "loss": 0.0691,
      "step": 1960
    },
    {
      "epoch": 3.504888888888889,
      "grad_norm": 0.006195690017193556,
      "learning_rate": 1.66073546856465e-05,
      "loss": 0.0199,
      "step": 1970
    },
    {
      "epoch": 3.522666666666667,
      "grad_norm": 8.7911958694458,
      "learning_rate": 1.6409648082245948e-05,
      "loss": 0.0023,
      "step": 1980
    },
    {
      "epoch": 3.5404444444444443,
      "grad_norm": 0.002020699204877019,
      "learning_rate": 1.6211941478845393e-05,
      "loss": 0.0022,
      "step": 1990
    },
    {
      "epoch": 3.558222222222222,
      "grad_norm": 0.006164974998682737,
      "learning_rate": 1.601423487544484e-05,
      "loss": 0.0004,
      "step": 2000
    },
    {
      "epoch": 3.576,
      "grad_norm": 0.004664299543946981,
      "learning_rate": 1.5816528272044286e-05,
      "loss": 0.0013,
      "step": 2010
    },
    {
      "epoch": 3.5937777777777775,
      "grad_norm": 0.47931966185569763,
      "learning_rate": 1.5618821668643734e-05,
      "loss": 0.0003,
      "step": 2020
    },
    {
      "epoch": 3.6115555555555554,
      "grad_norm": 43.603816986083984,
      "learning_rate": 1.5421115065243182e-05,
      "loss": 0.0085,
      "step": 2030
    },
    {
      "epoch": 3.6293333333333333,
      "grad_norm": 0.0044282907620072365,
      "learning_rate": 1.5223408461842625e-05,
      "loss": 0.0589,
      "step": 2040
    },
    {
      "epoch": 3.647111111111111,
      "grad_norm": 0.001094396342523396,
      "learning_rate": 1.5025701858442073e-05,
      "loss": 0.0006,
      "step": 2050
    },
    {
      "epoch": 3.664888888888889,
      "grad_norm": 0.008125217631459236,
      "learning_rate": 1.4827995255041518e-05,
      "loss": 0.0002,
      "step": 2060
    },
    {
      "epoch": 3.6826666666666665,
      "grad_norm": 19.899005889892578,
      "learning_rate": 1.4630288651640966e-05,
      "loss": 0.0241,
      "step": 2070
    },
    {
      "epoch": 3.7004444444444444,
      "grad_norm": 0.19454504549503326,
      "learning_rate": 1.4432582048240412e-05,
      "loss": 0.0005,
      "step": 2080
    },
    {
      "epoch": 3.7182222222222223,
      "grad_norm": 0.003989567514508963,
      "learning_rate": 1.423487544483986e-05,
      "loss": 0.0045,
      "step": 2090
    },
    {
      "epoch": 3.7359999999999998,
      "grad_norm": 0.0012507745996117592,
      "learning_rate": 1.4037168841439305e-05,
      "loss": 0.0011,
      "step": 2100
    },
    {
      "epoch": 3.7537777777777777,
      "grad_norm": 0.033174123615026474,
      "learning_rate": 1.383946223803875e-05,
      "loss": 0.0005,
      "step": 2110
    },
    {
      "epoch": 3.7715555555555556,
      "grad_norm": 0.04268620163202286,
      "learning_rate": 1.3641755634638198e-05,
      "loss": 0.0111,
      "step": 2120
    },
    {
      "epoch": 3.7893333333333334,
      "grad_norm": 0.041019320487976074,
      "learning_rate": 1.3444049031237643e-05,
      "loss": 0.0018,
      "step": 2130
    },
    {
      "epoch": 3.8071111111111113,
      "grad_norm": 0.1453588902950287,
      "learning_rate": 1.3246342427837091e-05,
      "loss": 0.0276,
      "step": 2140
    },
    {
      "epoch": 3.824888888888889,
      "grad_norm": 0.0027844184078276157,
      "learning_rate": 1.3048635824436536e-05,
      "loss": 0.0018,
      "step": 2150
    },
    {
      "epoch": 3.8426666666666667,
      "grad_norm": 0.0006637629121541977,
      "learning_rate": 1.2850929221035984e-05,
      "loss": 0.015,
      "step": 2160
    },
    {
      "epoch": 3.8604444444444446,
      "grad_norm": 0.00038110537570901215,
      "learning_rate": 1.265322261763543e-05,
      "loss": 0.0017,
      "step": 2170
    },
    {
      "epoch": 3.878222222222222,
      "grad_norm": 0.0005764598026871681,
      "learning_rate": 1.2455516014234877e-05,
      "loss": 0.0443,
      "step": 2180
    },
    {
      "epoch": 3.896,
      "grad_norm": 6.875983238220215,
      "learning_rate": 1.2257809410834323e-05,
      "loss": 0.0357,
      "step": 2190
    },
    {
      "epoch": 3.913777777777778,
      "grad_norm": 51.81392288208008,
      "learning_rate": 1.206010280743377e-05,
      "loss": 0.074,
      "step": 2200
    },
    {
      "epoch": 3.9315555555555557,
      "grad_norm": 0.0016795358387753367,
      "learning_rate": 1.1862396204033216e-05,
      "loss": 0.0046,
      "step": 2210
    },
    {
      "epoch": 3.9493333333333336,
      "grad_norm": 1.101694107055664,
      "learning_rate": 1.166468960063266e-05,
      "loss": 0.0004,
      "step": 2220
    },
    {
      "epoch": 3.967111111111111,
      "grad_norm": 0.03407168760895729,
      "learning_rate": 1.1466982997232107e-05,
      "loss": 0.0004,
      "step": 2230
    },
    {
      "epoch": 3.984888888888889,
      "grad_norm": 0.012767530046403408,
      "learning_rate": 1.1269276393831553e-05,
      "loss": 0.0426,
      "step": 2240
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 1.0,
      "eval_loss": 7.458427717210725e-06,
      "eval_runtime": 41.6071,
      "eval_samples_per_second": 192.275,
      "eval_steps_per_second": 6.009,
      "step": 2249
    },
    {
      "epoch": 4.001777777777778,
      "grad_norm": 0.0036902178544551134,
      "learning_rate": 1.1071569790431002e-05,
      "loss": 0.0003,
      "step": 2250
    },
    {
      "epoch": 4.019555555555556,
      "grad_norm": 0.00998405460268259,
      "learning_rate": 1.0873863187030448e-05,
      "loss": 0.002,
      "step": 2260
    },
    {
      "epoch": 4.037333333333334,
      "grad_norm": 23.953229904174805,
      "learning_rate": 1.0676156583629894e-05,
      "loss": 0.0187,
      "step": 2270
    },
    {
      "epoch": 4.0551111111111116,
      "grad_norm": 0.008150537498295307,
      "learning_rate": 1.0478449980229341e-05,
      "loss": 0.0153,
      "step": 2280
    },
    {
      "epoch": 4.072888888888889,
      "grad_norm": 0.5894471406936646,
      "learning_rate": 1.0280743376828787e-05,
      "loss": 0.0002,
      "step": 2290
    },
    {
      "epoch": 4.0906666666666665,
      "grad_norm": 0.07007890194654465,
      "learning_rate": 1.0083036773428232e-05,
      "loss": 0.0367,
      "step": 2300
    },
    {
      "epoch": 4.108444444444444,
      "grad_norm": 0.07020383328199387,
      "learning_rate": 9.885330170027678e-06,
      "loss": 0.0017,
      "step": 2310
    },
    {
      "epoch": 4.126222222222222,
      "grad_norm": 0.0013433824060484767,
      "learning_rate": 9.687623566627125e-06,
      "loss": 0.0176,
      "step": 2320
    },
    {
      "epoch": 4.144,
      "grad_norm": 0.0036678831093013287,
      "learning_rate": 9.489916963226571e-06,
      "loss": 0.001,
      "step": 2330
    },
    {
      "epoch": 4.161777777777778,
      "grad_norm": 20.646207809448242,
      "learning_rate": 9.29221035982602e-06,
      "loss": 0.0071,
      "step": 2340
    },
    {
      "epoch": 4.179555555555556,
      "grad_norm": 0.004499041475355625,
      "learning_rate": 9.094503756425466e-06,
      "loss": 0.0004,
      "step": 2350
    },
    {
      "epoch": 4.197333333333333,
      "grad_norm": 0.0007168107549659908,
      "learning_rate": 8.896797153024912e-06,
      "loss": 0.0025,
      "step": 2360
    },
    {
      "epoch": 4.215111111111111,
      "grad_norm": 0.015021364204585552,
      "learning_rate": 8.699090549624359e-06,
      "loss": 0.0277,
      "step": 2370
    },
    {
      "epoch": 4.232888888888889,
      "grad_norm": 0.006119410507380962,
      "learning_rate": 8.501383946223804e-06,
      "loss": 0.0009,
      "step": 2380
    },
    {
      "epoch": 4.250666666666667,
      "grad_norm": 0.0018322835676372051,
      "learning_rate": 8.30367734282325e-06,
      "loss": 0.0009,
      "step": 2390
    },
    {
      "epoch": 4.2684444444444445,
      "grad_norm": 0.0025883447378873825,
      "learning_rate": 8.105970739422696e-06,
      "loss": 0.0179,
      "step": 2400
    },
    {
      "epoch": 4.286222222222222,
      "grad_norm": 0.010295086540281773,
      "learning_rate": 7.908264136022143e-06,
      "loss": 0.0002,
      "step": 2410
    },
    {
      "epoch": 4.304,
      "grad_norm": 0.27159199118614197,
      "learning_rate": 7.710557532621591e-06,
      "loss": 0.0003,
      "step": 2420
    },
    {
      "epoch": 4.321777777777778,
      "grad_norm": 0.014537914656102657,
      "learning_rate": 7.5128509292210365e-06,
      "loss": 0.0108,
      "step": 2430
    },
    {
      "epoch": 4.339555555555555,
      "grad_norm": 0.001482433988712728,
      "learning_rate": 7.315144325820483e-06,
      "loss": 0.0003,
      "step": 2440
    },
    {
      "epoch": 4.357333333333333,
      "grad_norm": 0.0015277402708306909,
      "learning_rate": 7.11743772241993e-06,
      "loss": 0.0496,
      "step": 2450
    },
    {
      "epoch": 4.375111111111111,
      "grad_norm": 0.005141290370374918,
      "learning_rate": 6.919731119019375e-06,
      "loss": 0.0001,
      "step": 2460
    },
    {
      "epoch": 4.392888888888889,
      "grad_norm": 27.9423770904541,
      "learning_rate": 6.722024515618821e-06,
      "loss": 0.0393,
      "step": 2470
    },
    {
      "epoch": 4.410666666666667,
      "grad_norm": 0.010187560692429543,
      "learning_rate": 6.524317912218268e-06,
      "loss": 0.0001,
      "step": 2480
    },
    {
      "epoch": 4.428444444444445,
      "grad_norm": 0.002554529346525669,
      "learning_rate": 6.326611308817715e-06,
      "loss": 0.0031,
      "step": 2490
    },
    {
      "epoch": 4.4462222222222225,
      "grad_norm": 4.20240592956543,
      "learning_rate": 6.1289047054171615e-06,
      "loss": 0.0016,
      "step": 2500
    },
    {
      "epoch": 4.464,
      "grad_norm": 0.013741197995841503,
      "learning_rate": 5.931198102016608e-06,
      "loss": 0.0003,
      "step": 2510
    },
    {
      "epoch": 4.481777777777777,
      "grad_norm": 0.043951794505119324,
      "learning_rate": 5.7334914986160535e-06,
      "loss": 0.0125,
      "step": 2520
    },
    {
      "epoch": 4.499555555555555,
      "grad_norm": 0.11376281827688217,
      "learning_rate": 5.535784895215501e-06,
      "loss": 0.003,
      "step": 2530
    },
    {
      "epoch": 4.517333333333333,
      "grad_norm": 0.05734412372112274,
      "learning_rate": 5.338078291814947e-06,
      "loss": 0.0009,
      "step": 2540
    },
    {
      "epoch": 4.535111111111111,
      "grad_norm": 0.0010632964549586177,
      "learning_rate": 5.140371688414394e-06,
      "loss": 0.0196,
      "step": 2550
    },
    {
      "epoch": 4.552888888888889,
      "grad_norm": 0.0036729658022522926,
      "learning_rate": 4.942665085013839e-06,
      "loss": 0.0071,
      "step": 2560
    },
    {
      "epoch": 4.570666666666667,
      "grad_norm": 5.985267162322998,
      "learning_rate": 4.744958481613286e-06,
      "loss": 0.0311,
      "step": 2570
    },
    {
      "epoch": 4.588444444444445,
      "grad_norm": 0.996809720993042,
      "learning_rate": 4.547251878212733e-06,
      "loss": 0.0012,
      "step": 2580
    },
    {
      "epoch": 4.606222222222222,
      "grad_norm": 0.11869648844003677,
      "learning_rate": 4.349545274812179e-06,
      "loss": 0.0003,
      "step": 2590
    },
    {
      "epoch": 4.624,
      "grad_norm": 0.006143218372017145,
      "learning_rate": 4.151838671411625e-06,
      "loss": 0.0013,
      "step": 2600
    },
    {
      "epoch": 4.641777777777778,
      "grad_norm": 0.024631284177303314,
      "learning_rate": 3.954132068011071e-06,
      "loss": 0.0388,
      "step": 2610
    },
    {
      "epoch": 4.6595555555555555,
      "grad_norm": 0.0017836794722825289,
      "learning_rate": 3.7564254646105183e-06,
      "loss": 0.0015,
      "step": 2620
    },
    {
      "epoch": 4.677333333333333,
      "grad_norm": 0.003801500890403986,
      "learning_rate": 3.558718861209965e-06,
      "loss": 0.0003,
      "step": 2630
    },
    {
      "epoch": 4.695111111111111,
      "grad_norm": 0.004178278613835573,
      "learning_rate": 3.3610122578094107e-06,
      "loss": 0.001,
      "step": 2640
    },
    {
      "epoch": 4.712888888888889,
      "grad_norm": 0.0044832993298769,
      "learning_rate": 3.1633056544088575e-06,
      "loss": 0.0194,
      "step": 2650
    },
    {
      "epoch": 4.730666666666667,
      "grad_norm": 0.0029173328075557947,
      "learning_rate": 2.965599051008304e-06,
      "loss": 0.0302,
      "step": 2660
    },
    {
      "epoch": 4.748444444444445,
      "grad_norm": 0.0005038917297497392,
      "learning_rate": 2.7678924476077504e-06,
      "loss": 0.011,
      "step": 2670
    },
    {
      "epoch": 4.766222222222222,
      "grad_norm": 0.01968969963490963,
      "learning_rate": 2.570185844207197e-06,
      "loss": 0.0002,
      "step": 2680
    },
    {
      "epoch": 4.784,
      "grad_norm": 0.02507755346596241,
      "learning_rate": 2.372479240806643e-06,
      "loss": 0.0012,
      "step": 2690
    },
    {
      "epoch": 4.801777777777778,
      "grad_norm": 6.288967609405518,
      "learning_rate": 2.1747726374060897e-06,
      "loss": 0.0195,
      "step": 2700
    },
    {
      "epoch": 4.819555555555556,
      "grad_norm": 0.19547367095947266,
      "learning_rate": 1.9770660340055357e-06,
      "loss": 0.0034,
      "step": 2710
    },
    {
      "epoch": 4.8373333333333335,
      "grad_norm": 0.007160472217947245,
      "learning_rate": 1.7793594306049826e-06,
      "loss": 0.0008,
      "step": 2720
    },
    {
      "epoch": 4.855111111111111,
      "grad_norm": 0.1027381643652916,
      "learning_rate": 1.5816528272044288e-06,
      "loss": 0.0006,
      "step": 2730
    },
    {
      "epoch": 4.872888888888889,
      "grad_norm": 0.6163949966430664,
      "learning_rate": 1.3839462238038752e-06,
      "loss": 0.0001,
      "step": 2740
    },
    {
      "epoch": 4.890666666666666,
      "grad_norm": 0.0065285759046673775,
      "learning_rate": 1.1862396204033214e-06,
      "loss": 0.0,
      "step": 2750
    },
    {
      "epoch": 4.908444444444444,
      "grad_norm": 2.0664429664611816,
      "learning_rate": 9.885330170027678e-07,
      "loss": 0.0033,
      "step": 2760
    },
    {
      "epoch": 4.926222222222222,
      "grad_norm": 0.04097575694322586,
      "learning_rate": 7.908264136022144e-07,
      "loss": 0.001,
      "step": 2770
    },
    {
      "epoch": 4.944,
      "grad_norm": 0.0015862607397139072,
      "learning_rate": 5.931198102016607e-07,
      "loss": 0.0001,
      "step": 2780
    },
    {
      "epoch": 4.961777777777778,
      "grad_norm": 0.0021847274620085955,
      "learning_rate": 3.954132068011072e-07,
      "loss": 0.0006,
      "step": 2790
    },
    {
      "epoch": 4.979555555555556,
      "grad_norm": 0.00527564063668251,
      "learning_rate": 1.977066034005536e-07,
      "loss": 0.0243,
      "step": 2800
    },
    {
      "epoch": 4.997333333333334,
      "grad_norm": 0.0007434898870997131,
      "learning_rate": 0.0,
      "loss": 0.0525,
      "step": 2810
    },
    {
      "epoch": 4.997333333333334,
      "eval_accuracy": 1.0,
      "eval_loss": 2.338912281629746e-06,
      "eval_runtime": 40.6863,
      "eval_samples_per_second": 196.626,
      "eval_steps_per_second": 6.145,
      "step": 2810
    },
    {
      "epoch": 4.997333333333334,
      "step": 2810,
      "total_flos": 8.94051665811918e+18,
      "train_loss": 0.011076973860487225,
      "train_runtime": 5179.8163,
      "train_samples_per_second": 69.501,
      "train_steps_per_second": 0.542
    }
  ],
  "logging_steps": 10,
  "max_steps": 2810,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.94051665811918e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}