{
  "best_metric": 0.9985528219971056,
  "best_model_checkpoint": "videomae-base-finetuned-isl-numbers_aug/checkpoint-4180",
  "epoch": 4.2,
  "eval_steps": 500,
  "global_step": 4180,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0023923444976076554,
      "grad_norm": 14.015558242797852,
      "learning_rate": 1.1961722488038277e-06,
      "loss": 1.36,
      "step": 10
    },
    {
      "epoch": 0.004784688995215311,
      "grad_norm": 13.480476379394531,
      "learning_rate": 2.3923444976076554e-06,
      "loss": 1.4805,
      "step": 20
    },
    {
      "epoch": 0.007177033492822967,
      "grad_norm": 24.253854751586914,
      "learning_rate": 3.5885167464114835e-06,
      "loss": 1.1355,
      "step": 30
    },
    {
      "epoch": 0.009569377990430622,
      "grad_norm": 24.031822204589844,
      "learning_rate": 4.784688995215311e-06,
      "loss": 1.2175,
      "step": 40
    },
    {
      "epoch": 0.011961722488038277,
      "grad_norm": 33.942161560058594,
      "learning_rate": 5.980861244019139e-06,
      "loss": 1.5014,
      "step": 50
    },
    {
      "epoch": 0.014354066985645933,
      "grad_norm": 25.417530059814453,
      "learning_rate": 7.177033492822967e-06,
      "loss": 1.0382,
      "step": 60
    },
    {
      "epoch": 0.01674641148325359,
      "grad_norm": 33.37525939941406,
      "learning_rate": 8.373205741626795e-06,
      "loss": 0.818,
      "step": 70
    },
    {
      "epoch": 0.019138755980861243,
      "grad_norm": 11.674866676330566,
      "learning_rate": 9.569377990430622e-06,
      "loss": 1.0039,
      "step": 80
    },
    {
      "epoch": 0.0215311004784689,
      "grad_norm": 29.53816795349121,
      "learning_rate": 1.0765550239234451e-05,
      "loss": 1.0634,
      "step": 90
    },
    {
      "epoch": 0.023923444976076555,
      "grad_norm": 32.427513122558594,
      "learning_rate": 1.1961722488038278e-05,
      "loss": 1.1812,
      "step": 100
    },
    {
      "epoch": 0.02631578947368421,
      "grad_norm": 47.19343566894531,
      "learning_rate": 1.3157894736842106e-05,
      "loss": 0.9474,
      "step": 110
    },
    {
      "epoch": 0.028708133971291867,
      "grad_norm": 21.350784301757812,
      "learning_rate": 1.4354066985645934e-05,
      "loss": 0.7303,
      "step": 120
    },
    {
      "epoch": 0.03110047846889952,
      "grad_norm": 27.72856903076172,
      "learning_rate": 1.555023923444976e-05,
      "loss": 0.7567,
      "step": 130
    },
    {
      "epoch": 0.03349282296650718,
      "grad_norm": 24.274192810058594,
      "learning_rate": 1.674641148325359e-05,
      "loss": 0.5606,
      "step": 140
    },
    {
      "epoch": 0.03588516746411483,
      "grad_norm": 52.520484924316406,
      "learning_rate": 1.7942583732057417e-05,
      "loss": 0.718,
      "step": 150
    },
    {
      "epoch": 0.03827751196172249,
      "grad_norm": 37.529579162597656,
      "learning_rate": 1.9138755980861243e-05,
      "loss": 0.8521,
      "step": 160
    },
    {
      "epoch": 0.04066985645933014,
      "grad_norm": 4.263017177581787,
      "learning_rate": 2.0334928229665073e-05,
      "loss": 0.6431,
      "step": 170
    },
    {
      "epoch": 0.0430622009569378,
      "grad_norm": 15.213534355163574,
      "learning_rate": 2.1531100478468903e-05,
      "loss": 0.7901,
      "step": 180
    },
    {
      "epoch": 0.045454545454545456,
      "grad_norm": 31.36748504638672,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.6664,
      "step": 190
    },
    {
      "epoch": 0.04784688995215311,
      "grad_norm": 16.556751251220703,
      "learning_rate": 2.3923444976076556e-05,
      "loss": 0.5235,
      "step": 200
    },
    {
      "epoch": 0.050239234449760764,
      "grad_norm": 46.87775421142578,
      "learning_rate": 2.5119617224880382e-05,
      "loss": 0.8653,
      "step": 210
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 26.508085250854492,
      "learning_rate": 2.6315789473684212e-05,
      "loss": 0.8489,
      "step": 220
    },
    {
      "epoch": 0.05502392344497608,
      "grad_norm": 25.13356590270996,
      "learning_rate": 2.751196172248804e-05,
      "loss": 0.6788,
      "step": 230
    },
    {
      "epoch": 0.05741626794258373,
      "grad_norm": 31.49532699584961,
      "learning_rate": 2.8708133971291868e-05,
      "loss": 0.9036,
      "step": 240
    },
    {
      "epoch": 0.05980861244019139,
      "grad_norm": 42.16746520996094,
      "learning_rate": 2.9904306220095695e-05,
      "loss": 0.9568,
      "step": 250
    },
    {
      "epoch": 0.06220095693779904,
      "grad_norm": 22.292442321777344,
      "learning_rate": 3.110047846889952e-05,
      "loss": 0.3915,
      "step": 260
    },
    {
      "epoch": 0.0645933014354067,
      "grad_norm": 31.647417068481445,
      "learning_rate": 3.229665071770335e-05,
      "loss": 0.8292,
      "step": 270
    },
    {
      "epoch": 0.06698564593301436,
      "grad_norm": 31.727983474731445,
      "learning_rate": 3.349282296650718e-05,
      "loss": 0.8209,
      "step": 280
    },
    {
      "epoch": 0.06937799043062201,
      "grad_norm": 24.073183059692383,
      "learning_rate": 3.4688995215311004e-05,
      "loss": 0.4971,
      "step": 290
    },
    {
      "epoch": 0.07177033492822966,
      "grad_norm": 19.445539474487305,
      "learning_rate": 3.5885167464114834e-05,
      "loss": 0.7874,
      "step": 300
    },
    {
      "epoch": 0.07416267942583732,
      "grad_norm": 19.681711196899414,
      "learning_rate": 3.7081339712918663e-05,
      "loss": 0.6186,
      "step": 310
    },
    {
      "epoch": 0.07655502392344497,
      "grad_norm": 19.926891326904297,
      "learning_rate": 3.8277511961722486e-05,
      "loss": 0.5922,
      "step": 320
    },
    {
      "epoch": 0.07894736842105263,
      "grad_norm": 26.6019229888916,
      "learning_rate": 3.9473684210526316e-05,
      "loss": 0.4758,
      "step": 330
    },
    {
      "epoch": 0.08133971291866028,
      "grad_norm": 29.542890548706055,
      "learning_rate": 4.0669856459330146e-05,
      "loss": 0.6328,
      "step": 340
    },
    {
      "epoch": 0.08373205741626795,
      "grad_norm": 27.611955642700195,
      "learning_rate": 4.1866028708133976e-05,
      "loss": 0.8429,
      "step": 350
    },
    {
      "epoch": 0.0861244019138756,
      "grad_norm": 26.79853057861328,
      "learning_rate": 4.3062200956937806e-05,
      "loss": 0.7488,
      "step": 360
    },
    {
      "epoch": 0.08851674641148326,
      "grad_norm": 20.399744033813477,
      "learning_rate": 4.425837320574163e-05,
      "loss": 0.9143,
      "step": 370
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 31.50543785095215,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.9285,
      "step": 380
    },
    {
      "epoch": 0.09330143540669857,
      "grad_norm": 12.433365821838379,
      "learning_rate": 4.665071770334928e-05,
      "loss": 0.729,
      "step": 390
    },
    {
      "epoch": 0.09569377990430622,
      "grad_norm": 5.882745742797852,
      "learning_rate": 4.784688995215311e-05,
      "loss": 0.6221,
      "step": 400
    },
    {
      "epoch": 0.09808612440191387,
      "grad_norm": 47.791988372802734,
      "learning_rate": 4.904306220095694e-05,
      "loss": 0.697,
      "step": 410
    },
    {
      "epoch": 0.10047846889952153,
      "grad_norm": 32.664573669433594,
      "learning_rate": 4.997341839447103e-05,
      "loss": 0.6012,
      "step": 420
    },
    {
      "epoch": 0.10287081339712918,
      "grad_norm": 5.583775997161865,
      "learning_rate": 4.984051036682616e-05,
      "loss": 0.4423,
      "step": 430
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 19.180892944335938,
      "learning_rate": 4.970760233918128e-05,
      "loss": 0.7457,
      "step": 440
    },
    {
      "epoch": 0.1076555023923445,
      "grad_norm": 16.862192153930664,
      "learning_rate": 4.9574694311536416e-05,
      "loss": 0.8563,
      "step": 450
    },
    {
      "epoch": 0.11004784688995216,
      "grad_norm": 36.542415618896484,
      "learning_rate": 4.944178628389155e-05,
      "loss": 0.6835,
      "step": 460
    },
    {
      "epoch": 0.11244019138755981,
      "grad_norm": 26.89089584350586,
      "learning_rate": 4.930887825624668e-05,
      "loss": 0.5772,
      "step": 470
    },
    {
      "epoch": 0.11483253588516747,
      "grad_norm": 64.69689178466797,
      "learning_rate": 4.917597022860181e-05,
      "loss": 0.6574,
      "step": 480
    },
    {
      "epoch": 0.11722488038277512,
      "grad_norm": 32.71666717529297,
      "learning_rate": 4.904306220095694e-05,
      "loss": 0.7411,
      "step": 490
    },
    {
      "epoch": 0.11961722488038277,
      "grad_norm": 28.70256805419922,
      "learning_rate": 4.8910154173312074e-05,
      "loss": 0.7735,
      "step": 500
    },
    {
      "epoch": 0.12200956937799043,
      "grad_norm": 47.490455627441406,
      "learning_rate": 4.87772461456672e-05,
      "loss": 0.9173,
      "step": 510
    },
    {
      "epoch": 0.12440191387559808,
      "grad_norm": 0.40462446212768555,
      "learning_rate": 4.8644338118022334e-05,
      "loss": 0.3499,
      "step": 520
    },
    {
      "epoch": 0.12679425837320574,
      "grad_norm": 13.079553604125977,
      "learning_rate": 4.8511430090377467e-05,
      "loss": 0.6362,
      "step": 530
    },
    {
      "epoch": 0.1291866028708134,
      "grad_norm": 16.84446907043457,
      "learning_rate": 4.837852206273259e-05,
      "loss": 0.6439,
      "step": 540
    },
    {
      "epoch": 0.13157894736842105,
      "grad_norm": 1.5127249956130981,
      "learning_rate": 4.824561403508772e-05,
      "loss": 0.688,
      "step": 550
    },
    {
      "epoch": 0.1339712918660287,
      "grad_norm": 5.6877923011779785,
      "learning_rate": 4.811270600744285e-05,
      "loss": 0.4104,
      "step": 560
    },
    {
      "epoch": 0.13636363636363635,
      "grad_norm": 12.39535140991211,
      "learning_rate": 4.797979797979798e-05,
      "loss": 0.5339,
      "step": 570
    },
    {
      "epoch": 0.13875598086124402,
      "grad_norm": 55.42013168334961,
      "learning_rate": 4.784688995215311e-05,
      "loss": 0.4111,
      "step": 580
    },
    {
      "epoch": 0.14114832535885166,
      "grad_norm": 26.363779067993164,
      "learning_rate": 4.7713981924508244e-05,
      "loss": 0.8168,
      "step": 590
    },
    {
      "epoch": 0.14354066985645933,
      "grad_norm": 22.93970489501953,
      "learning_rate": 4.758107389686337e-05,
      "loss": 0.6611,
      "step": 600
    },
    {
      "epoch": 0.145933014354067,
      "grad_norm": 17.63263511657715,
      "learning_rate": 4.7448165869218504e-05,
      "loss": 0.6952,
      "step": 610
    },
    {
      "epoch": 0.14832535885167464,
      "grad_norm": 42.79911804199219,
      "learning_rate": 4.731525784157364e-05,
      "loss": 0.6323,
      "step": 620
    },
    {
      "epoch": 0.1507177033492823,
      "grad_norm": 52.291927337646484,
      "learning_rate": 4.718234981392876e-05,
      "loss": 0.5826,
      "step": 630
    },
    {
      "epoch": 0.15311004784688995,
      "grad_norm": 5.49782133102417,
      "learning_rate": 4.7049441786283896e-05,
      "loss": 0.6368,
      "step": 640
    },
    {
      "epoch": 0.15550239234449761,
      "grad_norm": 24.484716415405273,
      "learning_rate": 4.691653375863902e-05,
      "loss": 0.767,
      "step": 650
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 43.46366882324219,
      "learning_rate": 4.678362573099415e-05,
      "loss": 0.6198,
      "step": 660
    },
    {
      "epoch": 0.16028708133971292,
      "grad_norm": 0.9531241059303284,
      "learning_rate": 4.665071770334928e-05,
      "loss": 0.3393,
      "step": 670
    },
    {
      "epoch": 0.16267942583732056,
      "grad_norm": 23.464101791381836,
      "learning_rate": 4.6517809675704415e-05,
      "loss": 0.3039,
      "step": 680
    },
    {
      "epoch": 0.16507177033492823,
      "grad_norm": 61.715240478515625,
      "learning_rate": 4.638490164805955e-05,
      "loss": 0.635,
      "step": 690
    },
    {
      "epoch": 0.1674641148325359,
      "grad_norm": 4.636231422424316,
      "learning_rate": 4.6251993620414674e-05,
      "loss": 0.4028,
      "step": 700
    },
    {
      "epoch": 0.16985645933014354,
      "grad_norm": 19.922252655029297,
      "learning_rate": 4.611908559276981e-05,
      "loss": 0.4871,
      "step": 710
    },
    {
      "epoch": 0.1722488038277512,
      "grad_norm": 25.78838348388672,
      "learning_rate": 4.598617756512494e-05,
      "loss": 0.7949,
      "step": 720
    },
    {
      "epoch": 0.17464114832535885,
      "grad_norm": 8.514424324035645,
      "learning_rate": 4.5853269537480066e-05,
      "loss": 0.6953,
      "step": 730
    },
    {
      "epoch": 0.17703349282296652,
      "grad_norm": 44.999107360839844,
      "learning_rate": 4.57203615098352e-05,
      "loss": 0.7568,
      "step": 740
    },
    {
      "epoch": 0.17942583732057416,
      "grad_norm": 33.99626541137695,
      "learning_rate": 4.5587453482190326e-05,
      "loss": 0.3274,
      "step": 750
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 1.6210479736328125,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.2306,
      "step": 760
    },
    {
      "epoch": 0.18421052631578946,
      "grad_norm": 6.679223537445068,
      "learning_rate": 4.5321637426900585e-05,
      "loss": 0.6284,
      "step": 770
    },
    {
      "epoch": 0.18660287081339713,
      "grad_norm": 43.11864471435547,
      "learning_rate": 4.518872939925572e-05,
      "loss": 0.9184,
      "step": 780
    },
    {
      "epoch": 0.18899521531100477,
      "grad_norm": 39.860008239746094,
      "learning_rate": 4.5055821371610844e-05,
      "loss": 0.3346,
      "step": 790
    },
    {
      "epoch": 0.19138755980861244,
      "grad_norm": 58.848724365234375,
      "learning_rate": 4.492291334396598e-05,
      "loss": 0.7516,
      "step": 800
    },
    {
      "epoch": 0.1937799043062201,
      "grad_norm": 1.0923501253128052,
      "learning_rate": 4.479000531632111e-05,
      "loss": 0.5148,
      "step": 810
    },
    {
      "epoch": 0.19617224880382775,
      "grad_norm": 24.75638198852539,
      "learning_rate": 4.4657097288676236e-05,
      "loss": 0.5507,
      "step": 820
    },
    {
      "epoch": 0.19856459330143542,
      "grad_norm": 0.345415860414505,
      "learning_rate": 4.452418926103137e-05,
      "loss": 0.1649,
      "step": 830
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.8552821997105644,
      "eval_loss": 0.4412595331668854,
      "eval_runtime": 761.629,
      "eval_samples_per_second": 0.907,
      "eval_steps_per_second": 0.227,
      "step": 836
    },
    {
      "epoch": 1.000956937799043,
      "grad_norm": 55.6067008972168,
      "learning_rate": 4.43912812333865e-05,
      "loss": 0.3236,
      "step": 840
    },
    {
      "epoch": 1.0033492822966508,
      "grad_norm": 0.7264971733093262,
      "learning_rate": 4.425837320574163e-05,
      "loss": 0.3854,
      "step": 850
    },
    {
      "epoch": 1.0057416267942583,
      "grad_norm": 0.12801001965999603,
      "learning_rate": 4.412546517809676e-05,
      "loss": 0.2788,
      "step": 860
    },
    {
      "epoch": 1.008133971291866,
      "grad_norm": 29.689430236816406,
      "learning_rate": 4.399255715045189e-05,
      "loss": 0.5362,
      "step": 870
    },
    {
      "epoch": 1.0105263157894737,
      "grad_norm": 3.776660442352295,
      "learning_rate": 4.3859649122807014e-05,
      "loss": 0.2869,
      "step": 880
    },
    {
      "epoch": 1.0129186602870814,
      "grad_norm": 0.310320645570755,
      "learning_rate": 4.372674109516215e-05,
      "loss": 0.3163,
      "step": 890
    },
    {
      "epoch": 1.015311004784689,
      "grad_norm": 25.65765953063965,
      "learning_rate": 4.359383306751728e-05,
      "loss": 0.1512,
      "step": 900
    },
    {
      "epoch": 1.0177033492822967,
      "grad_norm": 68.732666015625,
      "learning_rate": 4.346092503987241e-05,
      "loss": 0.7086,
      "step": 910
    },
    {
      "epoch": 1.0200956937799044,
      "grad_norm": 18.452295303344727,
      "learning_rate": 4.332801701222754e-05,
      "loss": 0.3176,
      "step": 920
    },
    {
      "epoch": 1.022488038277512,
      "grad_norm": 1.5882939100265503,
      "learning_rate": 4.319510898458267e-05,
      "loss": 0.3374,
      "step": 930
    },
    {
      "epoch": 1.0248803827751196,
      "grad_norm": 22.042537689208984,
      "learning_rate": 4.3062200956937806e-05,
      "loss": 0.1728,
      "step": 940
    },
    {
      "epoch": 1.0272727272727273,
      "grad_norm": 1.0320727825164795,
      "learning_rate": 4.292929292929293e-05,
      "loss": 0.2954,
      "step": 950
    },
    {
      "epoch": 1.0296650717703348,
      "grad_norm": 27.908615112304688,
      "learning_rate": 4.2796384901648065e-05,
      "loss": 0.1296,
      "step": 960
    },
    {
      "epoch": 1.0320574162679426,
      "grad_norm": 39.791996002197266,
      "learning_rate": 4.266347687400319e-05,
      "loss": 0.0841,
      "step": 970
    },
    {
      "epoch": 1.0344497607655503,
      "grad_norm": 1.0737611055374146,
      "learning_rate": 4.253056884635832e-05,
      "loss": 0.4375,
      "step": 980
    },
    {
      "epoch": 1.0368421052631578,
      "grad_norm": 40.09695053100586,
      "learning_rate": 4.239766081871345e-05,
      "loss": 0.2464,
      "step": 990
    },
    {
      "epoch": 1.0392344497607655,
      "grad_norm": 5.695016384124756,
      "learning_rate": 4.2264752791068584e-05,
      "loss": 0.3341,
      "step": 1000
    },
    {
      "epoch": 1.0416267942583732,
      "grad_norm": 12.370247840881348,
      "learning_rate": 4.213184476342371e-05,
      "loss": 0.3143,
      "step": 1010
    },
    {
      "epoch": 1.044019138755981,
      "grad_norm": 0.9093998670578003,
      "learning_rate": 4.199893673577884e-05,
      "loss": 0.2729,
      "step": 1020
    },
    {
      "epoch": 1.0464114832535885,
      "grad_norm": 7.610515117645264,
      "learning_rate": 4.1866028708133976e-05,
      "loss": 0.1896,
      "step": 1030
    },
    {
      "epoch": 1.0488038277511962,
      "grad_norm": 7.180859088897705,
      "learning_rate": 4.17331206804891e-05,
      "loss": 0.4265,
      "step": 1040
    },
    {
      "epoch": 1.051196172248804,
      "grad_norm": 0.1555383950471878,
      "learning_rate": 4.1600212652844235e-05,
      "loss": 0.0456,
      "step": 1050
    },
    {
      "epoch": 1.0535885167464114,
      "grad_norm": 0.6277172565460205,
      "learning_rate": 4.146730462519937e-05,
      "loss": 0.3013,
      "step": 1060
    },
    {
      "epoch": 1.0559808612440191,
      "grad_norm": 26.010774612426758,
      "learning_rate": 4.1334396597554494e-05,
      "loss": 0.1184,
      "step": 1070
    },
    {
      "epoch": 1.0583732057416269,
      "grad_norm": 20.830724716186523,
      "learning_rate": 4.120148856990963e-05,
      "loss": 0.2435,
      "step": 1080
    },
    {
      "epoch": 1.0607655502392344,
      "grad_norm": 43.08734893798828,
      "learning_rate": 4.1068580542264754e-05,
      "loss": 0.1626,
      "step": 1090
    },
    {
      "epoch": 1.063157894736842,
      "grad_norm": 1.0129386186599731,
      "learning_rate": 4.093567251461988e-05,
      "loss": 0.1131,
      "step": 1100
    },
    {
      "epoch": 1.0655502392344498,
      "grad_norm": 74.8167495727539,
      "learning_rate": 4.080276448697501e-05,
      "loss": 0.3853,
      "step": 1110
    },
    {
      "epoch": 1.0679425837320573,
      "grad_norm": 0.1025879979133606,
      "learning_rate": 4.0669856459330146e-05,
      "loss": 0.1161,
      "step": 1120
    },
    {
      "epoch": 1.070334928229665,
      "grad_norm": 37.73963165283203,
      "learning_rate": 4.053694843168527e-05,
      "loss": 0.4214,
      "step": 1130
    },
    {
      "epoch": 1.0727272727272728,
      "grad_norm": 0.028422629460692406,
      "learning_rate": 4.0404040404040405e-05,
      "loss": 0.2753,
      "step": 1140
    },
    {
      "epoch": 1.0751196172248805,
      "grad_norm": 71.2274398803711,
      "learning_rate": 4.027113237639554e-05,
      "loss": 0.2009,
      "step": 1150
    },
    {
      "epoch": 1.077511961722488,
      "grad_norm": 39.377342224121094,
      "learning_rate": 4.0138224348750665e-05,
      "loss": 0.3117,
      "step": 1160
    },
    {
      "epoch": 1.0799043062200957,
      "grad_norm": 0.112571582198143,
      "learning_rate": 4.00053163211058e-05,
      "loss": 0.1175,
      "step": 1170
    },
    {
      "epoch": 1.0822966507177034,
      "grad_norm": 25.220449447631836,
      "learning_rate": 3.987240829346093e-05,
      "loss": 0.3254,
      "step": 1180
    },
    {
      "epoch": 1.084688995215311,
      "grad_norm": 50.95032501220703,
      "learning_rate": 3.973950026581606e-05,
      "loss": 0.2824,
      "step": 1190
    },
    {
      "epoch": 1.0870813397129186,
      "grad_norm": 24.743955612182617,
      "learning_rate": 3.960659223817118e-05,
      "loss": 0.2875,
      "step": 1200
    },
    {
      "epoch": 1.0894736842105264,
      "grad_norm": 57.72587966918945,
      "learning_rate": 3.9473684210526316e-05,
      "loss": 0.2018,
      "step": 1210
    },
    {
      "epoch": 1.0918660287081339,
      "grad_norm": 0.6570279598236084,
      "learning_rate": 3.934077618288145e-05,
      "loss": 0.4795,
      "step": 1220
    },
    {
      "epoch": 1.0942583732057416,
      "grad_norm": 2.577521324157715,
      "learning_rate": 3.9207868155236576e-05,
      "loss": 0.117,
      "step": 1230
    },
    {
      "epoch": 1.0966507177033493,
      "grad_norm": 0.06854643672704697,
      "learning_rate": 3.907496012759171e-05,
      "loss": 0.209,
      "step": 1240
    },
    {
      "epoch": 1.0990430622009568,
      "grad_norm": 0.17914560437202454,
      "learning_rate": 3.894205209994684e-05,
      "loss": 0.5758,
      "step": 1250
    },
    {
      "epoch": 1.1014354066985645,
      "grad_norm": 6.280195236206055,
      "learning_rate": 3.880914407230197e-05,
      "loss": 0.1779,
      "step": 1260
    },
    {
      "epoch": 1.1038277511961723,
      "grad_norm": 1.3312019109725952,
      "learning_rate": 3.86762360446571e-05,
      "loss": 0.1311,
      "step": 1270
    },
    {
      "epoch": 1.10622009569378,
      "grad_norm": 9.676103591918945,
      "learning_rate": 3.8543328017012234e-05,
      "loss": 0.1207,
      "step": 1280
    },
    {
      "epoch": 1.1086124401913875,
      "grad_norm": 104.29061889648438,
      "learning_rate": 3.841041998936736e-05,
      "loss": 0.3504,
      "step": 1290
    },
    {
      "epoch": 1.1110047846889952,
      "grad_norm": 0.15878012776374817,
      "learning_rate": 3.8277511961722486e-05,
      "loss": 0.1365,
      "step": 1300
    },
    {
      "epoch": 1.113397129186603,
      "grad_norm": 0.07822709530591965,
      "learning_rate": 3.814460393407762e-05,
      "loss": 0.4698,
      "step": 1310
    },
    {
      "epoch": 1.1157894736842104,
      "grad_norm": 0.5035278797149658,
      "learning_rate": 3.8011695906432746e-05,
      "loss": 0.2533,
      "step": 1320
    },
    {
      "epoch": 1.1181818181818182,
      "grad_norm": 35.283233642578125,
      "learning_rate": 3.787878787878788e-05,
      "loss": 0.3578,
      "step": 1330
    },
    {
      "epoch": 1.120574162679426,
      "grad_norm": 55.772911071777344,
      "learning_rate": 3.774587985114301e-05,
      "loss": 0.3235,
      "step": 1340
    },
    {
      "epoch": 1.1229665071770334,
      "grad_norm": 0.025184158235788345,
      "learning_rate": 3.761297182349814e-05,
      "loss": 0.0543,
      "step": 1350
    },
    {
      "epoch": 1.1253588516746411,
      "grad_norm": 0.04470227286219597,
      "learning_rate": 3.748006379585327e-05,
      "loss": 0.0378,
      "step": 1360
    },
    {
      "epoch": 1.1277511961722488,
      "grad_norm": 0.10474701225757599,
      "learning_rate": 3.7347155768208404e-05,
      "loss": 0.2598,
      "step": 1370
    },
    {
      "epoch": 1.1301435406698563,
      "grad_norm": 0.1666097193956375,
      "learning_rate": 3.721424774056353e-05,
      "loss": 0.3193,
      "step": 1380
    },
    {
      "epoch": 1.132535885167464,
      "grad_norm": 0.2347351312637329,
      "learning_rate": 3.7081339712918663e-05,
      "loss": 0.1546,
      "step": 1390
    },
    {
      "epoch": 1.1349282296650718,
      "grad_norm": 2.922492027282715,
      "learning_rate": 3.6948431685273796e-05,
      "loss": 0.0459,
      "step": 1400
    },
    {
      "epoch": 1.1373205741626795,
      "grad_norm": 90.11852264404297,
      "learning_rate": 3.681552365762892e-05,
      "loss": 0.1136,
      "step": 1410
    },
    {
      "epoch": 1.139712918660287,
      "grad_norm": 0.0664602667093277,
      "learning_rate": 3.668261562998405e-05,
      "loss": 0.1917,
      "step": 1420
    },
    {
      "epoch": 1.1421052631578947,
      "grad_norm": 28.271459579467773,
      "learning_rate": 3.654970760233918e-05,
      "loss": 0.0633,
      "step": 1430
    },
    {
      "epoch": 1.1444976076555025,
      "grad_norm": 48.338539123535156,
      "learning_rate": 3.6416799574694315e-05,
      "loss": 0.1383,
      "step": 1440
    },
    {
      "epoch": 1.14688995215311,
      "grad_norm": 1.0220551490783691,
      "learning_rate": 3.628389154704944e-05,
      "loss": 0.2675,
      "step": 1450
    },
    {
      "epoch": 1.1492822966507177,
      "grad_norm": 0.07435596734285355,
      "learning_rate": 3.6150983519404574e-05,
      "loss": 0.0058,
      "step": 1460
    },
    {
      "epoch": 1.1516746411483254,
      "grad_norm": 0.025288909673690796,
      "learning_rate": 3.601807549175971e-05,
      "loss": 0.3105,
      "step": 1470
    },
    {
      "epoch": 1.1540669856459331,
      "grad_norm": 3.0553793907165527,
      "learning_rate": 3.5885167464114834e-05,
      "loss": 0.1359,
      "step": 1480
    },
    {
      "epoch": 1.1564593301435406,
      "grad_norm": 12.173036575317383,
      "learning_rate": 3.5752259436469967e-05,
      "loss": 0.1923,
      "step": 1490
    },
    {
      "epoch": 1.1588516746411484,
      "grad_norm": 0.060364119708538055,
      "learning_rate": 3.56193514088251e-05,
      "loss": 0.0117,
      "step": 1500
    },
    {
      "epoch": 1.1612440191387559,
      "grad_norm": 0.03194255381822586,
      "learning_rate": 3.5486443381180226e-05,
      "loss": 0.004,
      "step": 1510
    },
    {
      "epoch": 1.1636363636363636,
      "grad_norm": 81.4512710571289,
      "learning_rate": 3.535353535353535e-05,
      "loss": 0.0776,
      "step": 1520
    },
    {
      "epoch": 1.1660287081339713,
      "grad_norm": 0.016529429703950882,
      "learning_rate": 3.5220627325890485e-05,
      "loss": 0.3877,
      "step": 1530
    },
    {
      "epoch": 1.168421052631579,
      "grad_norm": 0.6734187602996826,
      "learning_rate": 3.508771929824561e-05,
      "loss": 0.1606,
      "step": 1540
    },
    {
      "epoch": 1.1708133971291865,
      "grad_norm": 0.019191304221749306,
      "learning_rate": 3.4954811270600744e-05,
      "loss": 0.1287,
      "step": 1550
    },
    {
      "epoch": 1.1732057416267943,
      "grad_norm": 53.84104919433594,
      "learning_rate": 3.482190324295588e-05,
      "loss": 0.1832,
      "step": 1560
    },
    {
      "epoch": 1.175598086124402,
      "grad_norm": 30.083900451660156,
      "learning_rate": 3.4688995215311004e-05,
      "loss": 0.3451,
      "step": 1570
    },
    {
      "epoch": 1.1779904306220095,
      "grad_norm": 9.459929466247559,
      "learning_rate": 3.455608718766614e-05,
      "loss": 0.5572,
      "step": 1580
    },
    {
      "epoch": 1.1803827751196172,
      "grad_norm": 0.8345646262168884,
      "learning_rate": 3.442317916002127e-05,
      "loss": 0.0222,
      "step": 1590
    },
    {
      "epoch": 1.182775119617225,
      "grad_norm": 29.698089599609375,
      "learning_rate": 3.4290271132376396e-05,
      "loss": 0.1476,
      "step": 1600
    },
    {
      "epoch": 1.1851674641148326,
      "grad_norm": 40.125308990478516,
      "learning_rate": 3.415736310473153e-05,
      "loss": 0.1169,
      "step": 1610
    },
    {
      "epoch": 1.1875598086124401,
      "grad_norm": 0.015308327041566372,
      "learning_rate": 3.402445507708666e-05,
      "loss": 0.1322,
      "step": 1620
    },
    {
      "epoch": 1.1899521531100479,
      "grad_norm": 57.31826400756836,
      "learning_rate": 3.389154704944179e-05,
      "loss": 0.2188,
      "step": 1630
    },
    {
      "epoch": 1.1923444976076556,
      "grad_norm": 2.116442918777466,
      "learning_rate": 3.3758639021796915e-05,
      "loss": 0.0865,
      "step": 1640
    },
    {
      "epoch": 1.194736842105263,
      "grad_norm": 2.048318862915039,
      "learning_rate": 3.362573099415205e-05,
      "loss": 0.0313,
      "step": 1650
    },
    {
      "epoch": 1.1971291866028708,
      "grad_norm": 1.6264055967330933,
      "learning_rate": 3.349282296650718e-05,
      "loss": 0.0748,
      "step": 1660
    },
    {
      "epoch": 1.1995215311004785,
      "grad_norm": 0.020111849531531334,
      "learning_rate": 3.335991493886231e-05,
      "loss": 0.1242,
      "step": 1670
    },
    {
      "epoch": 1.2,
      "eval_accuracy": 0.9681620839363242,
      "eval_loss": 0.11400265246629715,
      "eval_runtime": 833.2657,
      "eval_samples_per_second": 0.829,
      "eval_steps_per_second": 0.208,
      "step": 1672
    },
    {
      "epoch": 2.001913875598086,
      "grad_norm": 0.04623245820403099,
      "learning_rate": 3.322700691121744e-05,
      "loss": 0.2269,
      "step": 1680
    },
    {
      "epoch": 2.004306220095694,
      "grad_norm": 0.04105829820036888,
      "learning_rate": 3.309409888357257e-05,
      "loss": 0.0314,
      "step": 1690
    },
    {
      "epoch": 2.0066985645933015,
      "grad_norm": 0.012316164560616016,
      "learning_rate": 3.29611908559277e-05,
      "loss": 0.0812,
      "step": 1700
    },
    {
      "epoch": 2.0090909090909093,
      "grad_norm": 0.014452405273914337,
      "learning_rate": 3.282828282828283e-05,
      "loss": 0.095,
      "step": 1710
    },
    {
      "epoch": 2.0114832535885165,
      "grad_norm": 0.02369452267885208,
      "learning_rate": 3.2695374800637965e-05,
      "loss": 0.0393,
      "step": 1720
    },
    {
      "epoch": 2.0138755980861243,
      "grad_norm": 0.05732015147805214,
      "learning_rate": 3.256246677299309e-05,
      "loss": 0.0019,
      "step": 1730
    },
    {
      "epoch": 2.016267942583732,
      "grad_norm": 0.23258601129055023,
      "learning_rate": 3.242955874534822e-05,
      "loss": 0.1842,
      "step": 1740
    },
    {
      "epoch": 2.0186602870813397,
      "grad_norm": 0.09864845871925354,
      "learning_rate": 3.229665071770335e-05,
      "loss": 0.0921,
      "step": 1750
    },
    {
      "epoch": 2.0210526315789474,
      "grad_norm": 472.58856201171875,
      "learning_rate": 3.216374269005848e-05,
      "loss": 0.1973,
      "step": 1760
    },
    {
      "epoch": 2.023444976076555,
      "grad_norm": 0.010467225685715675,
      "learning_rate": 3.203083466241361e-05,
      "loss": 0.0142,
      "step": 1770
    },
    {
      "epoch": 2.025837320574163,
      "grad_norm": 34.93497848510742,
      "learning_rate": 3.189792663476874e-05,
      "loss": 0.1638,
      "step": 1780
    },
    {
      "epoch": 2.02822966507177,
      "grad_norm": 108.92484283447266,
      "learning_rate": 3.176501860712387e-05,
      "loss": 0.2049,
      "step": 1790
    },
    {
      "epoch": 2.030622009569378,
      "grad_norm": 0.12743550539016724,
      "learning_rate": 3.1632110579479e-05,
      "loss": 0.0541,
      "step": 1800
    },
    {
      "epoch": 2.0330143540669856,
      "grad_norm": 0.01830146089196205,
      "learning_rate": 3.1499202551834136e-05,
      "loss": 0.0266,
      "step": 1810
    },
    {
      "epoch": 2.0354066985645933,
      "grad_norm": 0.07744389772415161,
      "learning_rate": 3.136629452418926e-05,
      "loss": 0.2081,
      "step": 1820
    },
    {
      "epoch": 2.037799043062201,
      "grad_norm": 0.09580251574516296,
      "learning_rate": 3.1233386496544395e-05,
      "loss": 0.1809,
      "step": 1830
    },
    {
      "epoch": 2.0401913875598088,
      "grad_norm": 1.139581561088562,
      "learning_rate": 3.110047846889952e-05,
      "loss": 0.0047,
      "step": 1840
    },
    {
      "epoch": 2.042583732057416,
      "grad_norm": 109.73816680908203,
      "learning_rate": 3.0967570441254654e-05,
      "loss": 0.1755,
      "step": 1850
    },
    {
      "epoch": 2.044976076555024,
      "grad_norm": 0.021538248285651207,
      "learning_rate": 3.083466241360978e-05,
      "loss": 0.146,
      "step": 1860
    },
    {
      "epoch": 2.0473684210526315,
      "grad_norm": 0.06147049739956856,
      "learning_rate": 3.0701754385964913e-05,
      "loss": 0.1163,
      "step": 1870
    },
    {
      "epoch": 2.0497607655502392,
      "grad_norm": 0.05214022099971771,
      "learning_rate": 3.056884635832004e-05,
      "loss": 0.0021,
      "step": 1880
    },
    {
      "epoch": 2.052153110047847,
      "grad_norm": 0.006269617471843958,
      "learning_rate": 3.0435938330675173e-05,
      "loss": 0.0012,
      "step": 1890
    },
    {
      "epoch": 2.0545454545454547,
      "grad_norm": 3.528353691101074,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 0.0097,
      "step": 1900
    },
    {
      "epoch": 2.0569377990430624,
      "grad_norm": 0.43716534972190857,
      "learning_rate": 3.017012227538544e-05,
      "loss": 0.0138,
      "step": 1910
    },
    {
      "epoch": 2.0593301435406697,
      "grad_norm": 0.008898725733160973,
      "learning_rate": 3.0037214247740565e-05,
      "loss": 0.1143,
      "step": 1920
    },
    {
      "epoch": 2.0617224880382774,
      "grad_norm": 18.777921676635742,
      "learning_rate": 2.9904306220095695e-05,
      "loss": 0.0059,
      "step": 1930
    },
    {
      "epoch": 2.064114832535885,
      "grad_norm": 0.010017622262239456,
      "learning_rate": 2.9771398192450828e-05,
      "loss": 0.0021,
      "step": 1940
    },
    {
      "epoch": 2.066507177033493,
      "grad_norm": 0.028644366189837456,
      "learning_rate": 2.9638490164805954e-05,
      "loss": 0.0024,
      "step": 1950
    },
    {
      "epoch": 2.0688995215311006,
      "grad_norm": 0.018241288140416145,
      "learning_rate": 2.9505582137161087e-05,
      "loss": 0.2653,
      "step": 1960
    },
    {
      "epoch": 2.0712918660287083,
      "grad_norm": 0.027866262942552567,
      "learning_rate": 2.937267410951622e-05,
      "loss": 0.0659,
      "step": 1970
    },
    {
      "epoch": 2.0736842105263156,
      "grad_norm": 0.0228541512042284,
      "learning_rate": 2.9239766081871346e-05,
      "loss": 0.0024,
      "step": 1980
    },
    {
      "epoch": 2.0760765550239233,
      "grad_norm": 1.270308494567871,
      "learning_rate": 2.9106858054226476e-05,
      "loss": 0.1811,
      "step": 1990
    },
    {
      "epoch": 2.078468899521531,
      "grad_norm": 0.03892623633146286,
      "learning_rate": 2.897395002658161e-05,
      "loss": 0.2084,
      "step": 2000
    },
    {
      "epoch": 2.0808612440191387,
      "grad_norm": 145.8848876953125,
      "learning_rate": 2.8841041998936735e-05,
      "loss": 0.064,
      "step": 2010
    },
    {
      "epoch": 2.0832535885167465,
      "grad_norm": 0.01369253359735012,
      "learning_rate": 2.8708133971291868e-05,
      "loss": 0.0026,
      "step": 2020
    },
    {
      "epoch": 2.085645933014354,
      "grad_norm": 0.06184624508023262,
      "learning_rate": 2.8575225943646998e-05,
      "loss": 0.0449,
      "step": 2030
    },
    {
      "epoch": 2.088038277511962,
      "grad_norm": 0.018334228545427322,
      "learning_rate": 2.8442317916002124e-05,
      "loss": 0.039,
      "step": 2040
    },
    {
      "epoch": 2.090430622009569,
      "grad_norm": 0.01396853942424059,
      "learning_rate": 2.8309409888357257e-05,
      "loss": 0.0009,
      "step": 2050
    },
    {
      "epoch": 2.092822966507177,
      "grad_norm": 9.39311408996582,
      "learning_rate": 2.817650186071239e-05,
      "loss": 0.0502,
      "step": 2060
    },
    {
      "epoch": 2.0952153110047846,
      "grad_norm": 0.01102653332054615,
      "learning_rate": 2.8043593833067516e-05,
      "loss": 0.0142,
      "step": 2070
    },
    {
      "epoch": 2.0976076555023924,
      "grad_norm": 0.030104391276836395,
      "learning_rate": 2.791068580542265e-05,
      "loss": 0.0014,
      "step": 2080
    },
    {
      "epoch": 2.1,
      "grad_norm": 2.8127565383911133,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.0858,
      "step": 2090
    },
    {
      "epoch": 2.102392344497608,
      "grad_norm": 0.022678395733237267,
      "learning_rate": 2.7644869750132905e-05,
      "loss": 0.0855,
      "step": 2100
    },
    {
      "epoch": 2.104784688995215,
      "grad_norm": 0.011784915812313557,
      "learning_rate": 2.751196172248804e-05,
      "loss": 0.2841,
      "step": 2110
    },
    {
      "epoch": 2.107177033492823,
      "grad_norm": 0.01566784456372261,
      "learning_rate": 2.737905369484317e-05,
      "loss": 0.4591,
      "step": 2120
    },
    {
      "epoch": 2.1095693779904305,
      "grad_norm": 0.22584687173366547,
      "learning_rate": 2.7246145667198298e-05,
      "loss": 0.1164,
      "step": 2130
    },
    {
      "epoch": 2.1119617224880383,
      "grad_norm": 0.293266624212265,
      "learning_rate": 2.711323763955343e-05,
      "loss": 0.0014,
      "step": 2140
    },
    {
      "epoch": 2.114354066985646,
      "grad_norm": 0.04173046350479126,
      "learning_rate": 2.698032961190856e-05,
      "loss": 0.0024,
      "step": 2150
    },
    {
      "epoch": 2.1167464114832537,
      "grad_norm": 0.014709273353219032,
      "learning_rate": 2.6847421584263693e-05,
      "loss": 0.0234,
      "step": 2160
    },
    {
      "epoch": 2.1191387559808614,
      "grad_norm": 0.060059864073991776,
      "learning_rate": 2.671451355661882e-05,
      "loss": 0.0406,
      "step": 2170
    },
    {
      "epoch": 2.1215311004784687,
      "grad_norm": 0.025863196700811386,
      "learning_rate": 2.6581605528973953e-05,
      "loss": 0.1082,
      "step": 2180
    },
    {
      "epoch": 2.1239234449760764,
      "grad_norm": 0.011012042872607708,
      "learning_rate": 2.6448697501329082e-05,
      "loss": 0.1153,
      "step": 2190
    },
    {
      "epoch": 2.126315789473684,
      "grad_norm": 0.009336967021226883,
      "learning_rate": 2.6315789473684212e-05,
      "loss": 0.1899,
      "step": 2200
    },
    {
      "epoch": 2.128708133971292,
      "grad_norm": 0.022180695086717606,
      "learning_rate": 2.618288144603934e-05,
      "loss": 0.0947,
      "step": 2210
    },
    {
      "epoch": 2.1311004784688996,
      "grad_norm": 0.06437038630247116,
      "learning_rate": 2.6049973418394475e-05,
      "loss": 0.046,
      "step": 2220
    },
    {
      "epoch": 2.1334928229665073,
      "grad_norm": 0.023490458726882935,
      "learning_rate": 2.59170653907496e-05,
      "loss": 0.0028,
      "step": 2230
    },
    {
      "epoch": 2.1358851674641146,
      "grad_norm": 0.010341249406337738,
      "learning_rate": 2.5784157363104734e-05,
      "loss": 0.0011,
      "step": 2240
    },
    {
      "epoch": 2.1382775119617223,
      "grad_norm": 0.019244810566306114,
      "learning_rate": 2.5651249335459864e-05,
      "loss": 0.001,
      "step": 2250
    },
    {
      "epoch": 2.14066985645933,
      "grad_norm": 2.1345505714416504,
      "learning_rate": 2.551834130781499e-05,
      "loss": 0.1981,
      "step": 2260
    },
    {
      "epoch": 2.143062200956938,
      "grad_norm": 0.06913293898105621,
      "learning_rate": 2.5385433280170123e-05,
      "loss": 0.0117,
      "step": 2270
    },
    {
      "epoch": 2.1454545454545455,
      "grad_norm": 17.53659439086914,
      "learning_rate": 2.5252525252525256e-05,
      "loss": 0.0048,
      "step": 2280
    },
    {
      "epoch": 2.1478468899521532,
      "grad_norm": 1.950246810913086,
      "learning_rate": 2.5119617224880382e-05,
      "loss": 0.0045,
      "step": 2290
    },
    {
      "epoch": 2.150239234449761,
      "grad_norm": 0.15338365733623505,
      "learning_rate": 2.4986709197235515e-05,
      "loss": 0.0287,
      "step": 2300
    },
    {
      "epoch": 2.1526315789473682,
      "grad_norm": 0.08566514402627945,
      "learning_rate": 2.485380116959064e-05,
      "loss": 0.0009,
      "step": 2310
    },
    {
      "epoch": 2.155023923444976,
      "grad_norm": 0.005700881592929363,
      "learning_rate": 2.4720893141945774e-05,
      "loss": 0.0013,
      "step": 2320
    },
    {
      "epoch": 2.1574162679425837,
      "grad_norm": 0.010471828281879425,
      "learning_rate": 2.4587985114300904e-05,
      "loss": 0.04,
      "step": 2330
    },
    {
      "epoch": 2.1598086124401914,
      "grad_norm": 0.15442727506160736,
      "learning_rate": 2.4455077086656037e-05,
      "loss": 0.0032,
      "step": 2340
    },
    {
      "epoch": 2.162200956937799,
      "grad_norm": 0.013687992468476295,
      "learning_rate": 2.4322169059011167e-05,
      "loss": 0.0006,
      "step": 2350
    },
    {
      "epoch": 2.164593301435407,
      "grad_norm": 15.276226043701172,
      "learning_rate": 2.4189261031366296e-05,
      "loss": 0.1538,
      "step": 2360
    },
    {
      "epoch": 2.166985645933014,
      "grad_norm": 0.011922297067940235,
      "learning_rate": 2.4056353003721426e-05,
      "loss": 0.004,
      "step": 2370
    },
    {
      "epoch": 2.169377990430622,
      "grad_norm": 0.012630579993128777,
      "learning_rate": 2.3923444976076556e-05,
      "loss": 0.0029,
      "step": 2380
    },
    {
      "epoch": 2.1717703349282296,
      "grad_norm": 2.478867530822754,
      "learning_rate": 2.3790536948431685e-05,
      "loss": 0.0054,
      "step": 2390
    },
    {
      "epoch": 2.1741626794258373,
      "grad_norm": 0.032478444278240204,
      "learning_rate": 2.365762892078682e-05,
      "loss": 0.0009,
      "step": 2400
    },
    {
      "epoch": 2.176555023923445,
      "grad_norm": 1.2129992246627808,
      "learning_rate": 2.3524720893141948e-05,
      "loss": 0.0015,
      "step": 2410
    },
    {
      "epoch": 2.1789473684210527,
      "grad_norm": 68.73790740966797,
      "learning_rate": 2.3391812865497074e-05,
      "loss": 0.0066,
      "step": 2420
    },
    {
      "epoch": 2.1813397129186605,
      "grad_norm": 0.03823688626289368,
      "learning_rate": 2.3258904837852207e-05,
      "loss": 0.0034,
      "step": 2430
    },
    {
      "epoch": 2.1837320574162677,
      "grad_norm": 0.010348823852837086,
      "learning_rate": 2.3125996810207337e-05,
      "loss": 0.0034,
      "step": 2440
    },
    {
      "epoch": 2.1861244019138755,
      "grad_norm": 0.03327915072441101,
      "learning_rate": 2.299308878256247e-05,
      "loss": 0.0044,
      "step": 2450
    },
    {
      "epoch": 2.188516746411483,
      "grad_norm": 0.0061185345984995365,
      "learning_rate": 2.28601807549176e-05,
      "loss": 0.001,
      "step": 2460
    },
    {
      "epoch": 2.190909090909091,
      "grad_norm": 0.019545989111065865,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.0545,
      "step": 2470
    },
    {
      "epoch": 2.1933014354066986,
      "grad_norm": 0.027101656422019005,
      "learning_rate": 2.259436469962786e-05,
      "loss": 0.0021,
      "step": 2480
    },
    {
      "epoch": 2.1956937799043064,
      "grad_norm": 99.50630950927734,
      "learning_rate": 2.246145667198299e-05,
      "loss": 0.0218,
      "step": 2490
    },
    {
      "epoch": 2.1980861244019136,
      "grad_norm": 0.009054621681571007,
      "learning_rate": 2.2328548644338118e-05,
      "loss": 0.0008,
      "step": 2500
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.9797395079594791,
      "eval_loss": 0.09082657098770142,
      "eval_runtime": 794.713,
      "eval_samples_per_second": 0.869,
      "eval_steps_per_second": 0.218,
      "step": 2508
    },
    {
      "epoch": 3.0004784688995216,
      "grad_norm": 0.006668519228696823,
      "learning_rate": 2.219564061669325e-05,
      "loss": 0.2009,
      "step": 2510
    },
    {
      "epoch": 3.0028708133971294,
      "grad_norm": 38.766822814941406,
      "learning_rate": 2.206273258904838e-05,
      "loss": 0.2188,
      "step": 2520
    },
    {
      "epoch": 3.0052631578947366,
      "grad_norm": 0.01620459370315075,
      "learning_rate": 2.1929824561403507e-05,
      "loss": 0.0007,
      "step": 2530
    },
    {
      "epoch": 3.0076555023923444,
      "grad_norm": 0.010571425780653954,
      "learning_rate": 2.179691653375864e-05,
      "loss": 0.0011,
      "step": 2540
    },
    {
      "epoch": 3.010047846889952,
      "grad_norm": 0.007847270928323269,
      "learning_rate": 2.166400850611377e-05,
      "loss": 0.0019,
      "step": 2550
    },
    {
      "epoch": 3.01244019138756,
      "grad_norm": 2.9406144618988037,
      "learning_rate": 2.1531100478468903e-05,
      "loss": 0.008,
      "step": 2560
    },
    {
      "epoch": 3.0148325358851675,
      "grad_norm": 0.16233409941196442,
      "learning_rate": 2.1398192450824032e-05,
      "loss": 0.001,
      "step": 2570
    },
    {
      "epoch": 3.0172248803827753,
      "grad_norm": 0.034926705062389374,
      "learning_rate": 2.126528442317916e-05,
      "loss": 0.0007,
      "step": 2580
    },
    {
      "epoch": 3.019617224880383,
      "grad_norm": 0.007242594845592976,
      "learning_rate": 2.1132376395534292e-05,
      "loss": 0.2104,
      "step": 2590
    },
    {
      "epoch": 3.0220095693779903,
      "grad_norm": 0.007837713696062565,
      "learning_rate": 2.099946836788942e-05,
      "loss": 0.2077,
      "step": 2600
    },
    {
      "epoch": 3.024401913875598,
      "grad_norm": 0.006022488698363304,
      "learning_rate": 2.086656034024455e-05,
      "loss": 0.0011,
      "step": 2610
    },
    {
      "epoch": 3.0267942583732057,
      "grad_norm": 0.08888078480958939,
      "learning_rate": 2.0733652312599684e-05,
      "loss": 0.0017,
      "step": 2620
    },
    {
      "epoch": 3.0291866028708134,
      "grad_norm": 0.05022843927145004,
      "learning_rate": 2.0600744284954814e-05,
      "loss": 0.1537,
      "step": 2630
    },
    {
      "epoch": 3.031578947368421,
      "grad_norm": 0.020734012126922607,
      "learning_rate": 2.046783625730994e-05,
      "loss": 0.0015,
      "step": 2640
    },
    {
      "epoch": 3.033971291866029,
      "grad_norm": 97.93255615234375,
      "learning_rate": 2.0334928229665073e-05,
      "loss": 0.0167,
      "step": 2650
    },
    {
      "epoch": 3.036363636363636,
      "grad_norm": 0.018566004931926727,
      "learning_rate": 2.0202020202020203e-05,
      "loss": 0.0012,
      "step": 2660
    },
    {
      "epoch": 3.038755980861244,
      "grad_norm": 0.005064102355390787,
      "learning_rate": 2.0069112174375332e-05,
      "loss": 0.0005,
      "step": 2670
    },
    {
      "epoch": 3.0411483253588516,
      "grad_norm": 0.019417481496930122,
      "learning_rate": 1.9936204146730465e-05,
      "loss": 0.0049,
      "step": 2680
    },
    {
      "epoch": 3.0435406698564593,
      "grad_norm": 0.007396561559289694,
      "learning_rate": 1.980329611908559e-05,
      "loss": 0.0025,
      "step": 2690
    },
    {
      "epoch": 3.045933014354067,
      "grad_norm": 61.74516677856445,
      "learning_rate": 1.9670388091440725e-05,
      "loss": 0.0991,
      "step": 2700
    },
    {
      "epoch": 3.0483253588516748,
      "grad_norm": 0.021895471960306168,
      "learning_rate": 1.9537480063795854e-05,
      "loss": 0.0007,
      "step": 2710
    },
    {
      "epoch": 3.0507177033492825,
      "grad_norm": 0.011149121448397636,
      "learning_rate": 1.9404572036150984e-05,
      "loss": 0.0006,
      "step": 2720
    },
    {
      "epoch": 3.0531100478468898,
      "grad_norm": 0.004441165365278721,
      "learning_rate": 1.9271664008506117e-05,
      "loss": 0.1103,
      "step": 2730
    },
    {
      "epoch": 3.0555023923444975,
      "grad_norm": 0.01931576617062092,
      "learning_rate": 1.9138755980861243e-05,
      "loss": 0.0029,
      "step": 2740
    },
    {
      "epoch": 3.057894736842105,
      "grad_norm": 0.005823149345815182,
      "learning_rate": 1.9005847953216373e-05,
      "loss": 0.1682,
      "step": 2750
    },
    {
      "epoch": 3.060287081339713,
      "grad_norm": 0.00678264070302248,
      "learning_rate": 1.8872939925571506e-05,
      "loss": 0.1651,
      "step": 2760
    },
    {
      "epoch": 3.0626794258373207,
      "grad_norm": 0.010020887479186058,
      "learning_rate": 1.8740031897926636e-05,
      "loss": 0.0011,
      "step": 2770
    },
    {
      "epoch": 3.0650717703349284,
      "grad_norm": 0.005996669642627239,
      "learning_rate": 1.8607123870281765e-05,
      "loss": 0.0018,
      "step": 2780
    },
    {
      "epoch": 3.0674641148325357,
      "grad_norm": 0.09342297166585922,
      "learning_rate": 1.8474215842636898e-05,
      "loss": 0.0008,
      "step": 2790
    },
    {
      "epoch": 3.0698564593301434,
      "grad_norm": 0.00897641945630312,
      "learning_rate": 1.8341307814992024e-05,
      "loss": 0.0108,
      "step": 2800
    },
    {
      "epoch": 3.072248803827751,
      "grad_norm": 0.1280699372291565,
      "learning_rate": 1.8208399787347157e-05,
      "loss": 0.0066,
      "step": 2810
    },
    {
      "epoch": 3.074641148325359,
      "grad_norm": 0.0044610267505049706,
      "learning_rate": 1.8075491759702287e-05,
      "loss": 0.0006,
      "step": 2820
    },
    {
      "epoch": 3.0770334928229666,
      "grad_norm": 0.006019377615302801,
      "learning_rate": 1.7942583732057417e-05,
      "loss": 0.0005,
      "step": 2830
    },
    {
      "epoch": 3.0794258373205743,
      "grad_norm": 0.0070389206521213055,
      "learning_rate": 1.780967570441255e-05,
      "loss": 0.006,
      "step": 2840
    },
    {
      "epoch": 3.081818181818182,
      "grad_norm": 0.010742398910224438,
      "learning_rate": 1.7676767676767676e-05,
      "loss": 0.0012,
      "step": 2850
    },
    {
      "epoch": 3.0842105263157893,
      "grad_norm": 0.019535677507519722,
      "learning_rate": 1.7543859649122806e-05,
      "loss": 0.0005,
      "step": 2860
    },
    {
      "epoch": 3.086602870813397,
      "grad_norm": 0.004338996019214392,
      "learning_rate": 1.741095162147794e-05,
      "loss": 0.0006,
      "step": 2870
    },
    {
      "epoch": 3.0889952153110047,
      "grad_norm": 0.004541581030935049,
      "learning_rate": 1.727804359383307e-05,
      "loss": 0.0873,
      "step": 2880
    },
    {
      "epoch": 3.0913875598086125,
      "grad_norm": 0.006278311833739281,
      "learning_rate": 1.7145135566188198e-05,
      "loss": 0.0012,
      "step": 2890
    },
    {
      "epoch": 3.09377990430622,
      "grad_norm": 1.023903250694275,
      "learning_rate": 1.701222753854333e-05,
      "loss": 0.0007,
      "step": 2900
    },
    {
      "epoch": 3.096172248803828,
      "grad_norm": 0.005157362204045057,
      "learning_rate": 1.6879319510898457e-05,
      "loss": 0.0004,
      "step": 2910
    },
    {
      "epoch": 3.098564593301435,
      "grad_norm": 0.006036865524947643,
      "learning_rate": 1.674641148325359e-05,
      "loss": 0.0041,
      "step": 2920
    },
    {
      "epoch": 3.100956937799043,
      "grad_norm": 0.7861006259918213,
      "learning_rate": 1.661350345560872e-05,
      "loss": 0.003,
      "step": 2930
    },
    {
      "epoch": 3.1033492822966506,
      "grad_norm": 0.4006275534629822,
      "learning_rate": 1.648059542796385e-05,
      "loss": 0.0253,
      "step": 2940
    },
    {
      "epoch": 3.1057416267942584,
      "grad_norm": 0.13688111305236816,
      "learning_rate": 1.6347687400318983e-05,
      "loss": 0.0006,
      "step": 2950
    },
    {
      "epoch": 3.108133971291866,
      "grad_norm": 0.007709797937422991,
      "learning_rate": 1.621477937267411e-05,
      "loss": 0.0006,
      "step": 2960
    },
    {
      "epoch": 3.110526315789474,
      "grad_norm": 0.004617704544216394,
      "learning_rate": 1.608187134502924e-05,
      "loss": 0.0008,
      "step": 2970
    },
    {
      "epoch": 3.1129186602870815,
      "grad_norm": 0.011701893992722034,
      "learning_rate": 1.594896331738437e-05,
      "loss": 0.0582,
      "step": 2980
    },
    {
      "epoch": 3.115311004784689,
      "grad_norm": 0.20932549238204956,
      "learning_rate": 1.58160552897395e-05,
      "loss": 0.0012,
      "step": 2990
    },
    {
      "epoch": 3.1177033492822965,
      "grad_norm": 0.008711748756468296,
      "learning_rate": 1.568314726209463e-05,
      "loss": 0.0938,
      "step": 3000
    },
    {
      "epoch": 3.1200956937799043,
      "grad_norm": 0.004468224477022886,
      "learning_rate": 1.555023923444976e-05,
      "loss": 0.0006,
      "step": 3010
    },
    {
      "epoch": 3.122488038277512,
      "grad_norm": 0.061067596077919006,
      "learning_rate": 1.541733120680489e-05,
      "loss": 0.0007,
      "step": 3020
    },
    {
      "epoch": 3.1248803827751197,
      "grad_norm": 0.025924433022737503,
      "learning_rate": 1.528442317916002e-05,
      "loss": 0.0147,
      "step": 3030
    },
    {
      "epoch": 3.1272727272727274,
      "grad_norm": 0.33229324221611023,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.0006,
      "step": 3040
    },
    {
      "epoch": 3.1296650717703347,
      "grad_norm": 0.019822843372821808,
      "learning_rate": 1.5018607123870283e-05,
      "loss": 0.0017,
      "step": 3050
    },
    {
      "epoch": 3.1320574162679424,
      "grad_norm": 0.009158506989479065,
      "learning_rate": 1.4885699096225414e-05,
      "loss": 0.0179,
      "step": 3060
    },
    {
      "epoch": 3.13444976076555,
      "grad_norm": 104.4133071899414,
      "learning_rate": 1.4752791068580543e-05,
      "loss": 0.1185,
      "step": 3070
    },
    {
      "epoch": 3.136842105263158,
      "grad_norm": 0.013615977019071579,
      "learning_rate": 1.4619883040935673e-05,
      "loss": 0.0012,
      "step": 3080
    },
    {
      "epoch": 3.1392344497607656,
      "grad_norm": 0.007452609483152628,
      "learning_rate": 1.4486975013290804e-05,
      "loss": 0.0006,
      "step": 3090
    },
    {
      "epoch": 3.1416267942583733,
      "grad_norm": 0.017694944515824318,
      "learning_rate": 1.4354066985645934e-05,
      "loss": 0.0889,
      "step": 3100
    },
    {
      "epoch": 3.144019138755981,
      "grad_norm": 0.0034980145283043385,
      "learning_rate": 1.4221158958001062e-05,
      "loss": 0.0137,
      "step": 3110
    },
    {
      "epoch": 3.1464114832535883,
      "grad_norm": 0.006997366435825825,
      "learning_rate": 1.4088250930356195e-05,
      "loss": 0.1826,
      "step": 3120
    },
    {
      "epoch": 3.148803827751196,
      "grad_norm": 0.20292145013809204,
      "learning_rate": 1.3955342902711325e-05,
      "loss": 0.0007,
      "step": 3130
    },
    {
      "epoch": 3.1511961722488038,
      "grad_norm": 0.009792889468371868,
      "learning_rate": 1.3822434875066453e-05,
      "loss": 0.0338,
      "step": 3140
    },
    {
      "epoch": 3.1535885167464115,
      "grad_norm": 0.006899695377796888,
      "learning_rate": 1.3689526847421586e-05,
      "loss": 0.0006,
      "step": 3150
    },
    {
      "epoch": 3.155980861244019,
      "grad_norm": 0.004146776627749205,
      "learning_rate": 1.3556618819776715e-05,
      "loss": 0.0019,
      "step": 3160
    },
    {
      "epoch": 3.158373205741627,
      "grad_norm": 0.003076452063396573,
      "learning_rate": 1.3423710792131847e-05,
      "loss": 0.0004,
      "step": 3170
    },
    {
      "epoch": 3.1607655502392342,
      "grad_norm": 0.022380296140909195,
      "learning_rate": 1.3290802764486976e-05,
      "loss": 0.0011,
      "step": 3180
    },
    {
      "epoch": 3.163157894736842,
      "grad_norm": 0.005187951028347015,
      "learning_rate": 1.3157894736842106e-05,
      "loss": 0.0005,
      "step": 3190
    },
    {
      "epoch": 3.1655502392344497,
      "grad_norm": 0.004998629912734032,
      "learning_rate": 1.3024986709197237e-05,
      "loss": 0.0006,
      "step": 3200
    },
    {
      "epoch": 3.1679425837320574,
      "grad_norm": 0.016898494213819504,
      "learning_rate": 1.2892078681552367e-05,
      "loss": 0.0007,
      "step": 3210
    },
    {
      "epoch": 3.170334928229665,
      "grad_norm": 0.43881699442863464,
      "learning_rate": 1.2759170653907495e-05,
      "loss": 0.0041,
      "step": 3220
    },
    {
      "epoch": 3.172727272727273,
      "grad_norm": 0.10537323355674744,
      "learning_rate": 1.2626262626262628e-05,
      "loss": 0.0009,
      "step": 3230
    },
    {
      "epoch": 3.1751196172248806,
      "grad_norm": 0.00541999377310276,
      "learning_rate": 1.2493354598617758e-05,
      "loss": 0.0004,
      "step": 3240
    },
    {
      "epoch": 3.177511961722488,
      "grad_norm": 0.0035200186539441347,
      "learning_rate": 1.2360446570972887e-05,
      "loss": 0.0007,
      "step": 3250
    },
    {
      "epoch": 3.1799043062200956,
      "grad_norm": 0.003413601079955697,
      "learning_rate": 1.2227538543328019e-05,
      "loss": 0.0004,
      "step": 3260
    },
    {
      "epoch": 3.1822966507177033,
      "grad_norm": 0.003805672749876976,
      "learning_rate": 1.2094630515683148e-05,
      "loss": 0.0004,
      "step": 3270
    },
    {
      "epoch": 3.184688995215311,
      "grad_norm": 0.008552059531211853,
      "learning_rate": 1.1961722488038278e-05,
      "loss": 0.0048,
      "step": 3280
    },
    {
      "epoch": 3.1870813397129187,
      "grad_norm": 0.003063785843551159,
      "learning_rate": 1.182881446039341e-05,
      "loss": 0.0003,
      "step": 3290
    },
    {
      "epoch": 3.1894736842105265,
      "grad_norm": 0.41841933131217957,
      "learning_rate": 1.1695906432748537e-05,
      "loss": 0.0005,
      "step": 3300
    },
    {
      "epoch": 3.1918660287081337,
      "grad_norm": 0.003419842105358839,
      "learning_rate": 1.1562998405103668e-05,
      "loss": 0.0048,
      "step": 3310
    },
    {
      "epoch": 3.1942583732057415,
      "grad_norm": 0.007587193511426449,
      "learning_rate": 1.14300903774588e-05,
      "loss": 0.0004,
      "step": 3320
    },
    {
      "epoch": 3.196650717703349,
      "grad_norm": 0.014273949898779392,
      "learning_rate": 1.129718234981393e-05,
      "loss": 0.0004,
      "step": 3330
    },
    {
      "epoch": 3.199043062200957,
      "grad_norm": 0.003971030004322529,
      "learning_rate": 1.1164274322169059e-05,
      "loss": 0.0004,
      "step": 3340
    },
    {
      "epoch": 3.2,
      "eval_accuracy": 0.9956584659913169,
      "eval_loss": 0.011571008712053299,
      "eval_runtime": 802.1767,
      "eval_samples_per_second": 0.861,
      "eval_steps_per_second": 0.216,
      "step": 3344
    },
    {
      "epoch": 4.0014354066985645,
      "grad_norm": 0.0065171197056770325,
      "learning_rate": 1.103136629452419e-05,
      "loss": 0.0004,
      "step": 3350
    },
    {
      "epoch": 4.003827751196172,
      "grad_norm": 0.005604379344731569,
      "learning_rate": 1.089845826687932e-05,
      "loss": 0.0013,
      "step": 3360
    },
    {
      "epoch": 4.00622009569378,
      "grad_norm": 0.009186330251395702,
      "learning_rate": 1.0765550239234451e-05,
      "loss": 0.0004,
      "step": 3370
    },
    {
      "epoch": 4.008612440191388,
      "grad_norm": 0.004106089938431978,
      "learning_rate": 1.063264221158958e-05,
      "loss": 0.0013,
      "step": 3380
    },
    {
      "epoch": 4.011004784688995,
      "grad_norm": 0.01519398856908083,
      "learning_rate": 1.049973418394471e-05,
      "loss": 0.0004,
      "step": 3390
    },
    {
      "epoch": 4.013397129186603,
      "grad_norm": 0.025932341814041138,
      "learning_rate": 1.0366826156299842e-05,
      "loss": 0.0005,
      "step": 3400
    },
    {
      "epoch": 4.015789473684211,
      "grad_norm": 0.003742667380720377,
      "learning_rate": 1.023391812865497e-05,
      "loss": 0.0015,
      "step": 3410
    },
    {
      "epoch": 4.0181818181818185,
      "grad_norm": 0.011091819033026695,
      "learning_rate": 1.0101010101010101e-05,
      "loss": 0.0005,
      "step": 3420
    },
    {
      "epoch": 4.020574162679426,
      "grad_norm": 0.14429621398448944,
      "learning_rate": 9.968102073365233e-06,
      "loss": 0.0008,
      "step": 3430
    },
    {
      "epoch": 4.022966507177033,
      "grad_norm": 0.006444338243454695,
      "learning_rate": 9.835194045720362e-06,
      "loss": 0.0004,
      "step": 3440
    },
    {
      "epoch": 4.025358851674641,
      "grad_norm": 0.003463227301836014,
      "learning_rate": 9.702286018075492e-06,
      "loss": 0.0008,
      "step": 3450
    },
    {
      "epoch": 4.0277511961722485,
      "grad_norm": 0.003970850259065628,
      "learning_rate": 9.569377990430622e-06,
      "loss": 0.0003,
      "step": 3460
    },
    {
      "epoch": 4.030143540669856,
      "grad_norm": 0.004473558627068996,
      "learning_rate": 9.436469962785753e-06,
      "loss": 0.0325,
      "step": 3470
    },
    {
      "epoch": 4.032535885167464,
      "grad_norm": 0.0038171063642948866,
      "learning_rate": 9.303561935140883e-06,
      "loss": 0.0004,
      "step": 3480
    },
    {
      "epoch": 4.034928229665072,
      "grad_norm": 0.0045017702504992485,
      "learning_rate": 9.170653907496012e-06,
      "loss": 0.0003,
      "step": 3490
    },
    {
      "epoch": 4.037320574162679,
      "grad_norm": 0.0030059998389333487,
      "learning_rate": 9.037745879851144e-06,
      "loss": 0.0004,
      "step": 3500
    },
    {
      "epoch": 4.039712918660287,
      "grad_norm": 0.03430689498782158,
      "learning_rate": 8.904837852206275e-06,
      "loss": 0.0005,
      "step": 3510
    },
    {
      "epoch": 4.042105263157895,
      "grad_norm": 0.0038995530921965837,
      "learning_rate": 8.771929824561403e-06,
      "loss": 0.0004,
      "step": 3520
    },
    {
      "epoch": 4.044497607655503,
      "grad_norm": 0.005899826996028423,
      "learning_rate": 8.639021796916534e-06,
      "loss": 0.0004,
      "step": 3530
    },
    {
      "epoch": 4.04688995215311,
      "grad_norm": 0.01342977024614811,
      "learning_rate": 8.506113769271666e-06,
      "loss": 0.0003,
      "step": 3540
    },
    {
      "epoch": 4.049282296650718,
      "grad_norm": 0.0030721963848918676,
      "learning_rate": 8.373205741626795e-06,
      "loss": 0.0005,
      "step": 3550
    },
    {
      "epoch": 4.051674641148326,
      "grad_norm": 0.017302511259913445,
      "learning_rate": 8.240297713981925e-06,
      "loss": 0.0005,
      "step": 3560
    },
    {
      "epoch": 4.054066985645933,
      "grad_norm": 0.018140502274036407,
      "learning_rate": 8.107389686337054e-06,
      "loss": 0.0003,
      "step": 3570
    },
    {
      "epoch": 4.05645933014354,
      "grad_norm": 0.0029472303576767445,
      "learning_rate": 7.974481658692186e-06,
      "loss": 0.0003,
      "step": 3580
    },
    {
      "epoch": 4.058851674641148,
      "grad_norm": 0.005908080376684666,
      "learning_rate": 7.841573631047315e-06,
      "loss": 0.0003,
      "step": 3590
    },
    {
      "epoch": 4.061244019138756,
      "grad_norm": 0.004262011032551527,
      "learning_rate": 7.708665603402445e-06,
      "loss": 0.0005,
      "step": 3600
    },
    {
      "epoch": 4.0636363636363635,
      "grad_norm": 0.004099706653505564,
      "learning_rate": 7.5757575757575764e-06,
      "loss": 0.0003,
      "step": 3610
    },
    {
      "epoch": 4.066028708133971,
      "grad_norm": 0.00331715471111238,
      "learning_rate": 7.442849548112707e-06,
      "loss": 0.0003,
      "step": 3620
    },
    {
      "epoch": 4.068421052631579,
      "grad_norm": 0.0059309545904397964,
      "learning_rate": 7.3099415204678366e-06,
      "loss": 0.0003,
      "step": 3630
    },
    {
      "epoch": 4.070813397129187,
      "grad_norm": 0.004388094414025545,
      "learning_rate": 7.177033492822967e-06,
      "loss": 0.0015,
      "step": 3640
    },
    {
      "epoch": 4.073205741626794,
      "grad_norm": 0.029676349833607674,
      "learning_rate": 7.0441254651780975e-06,
      "loss": 0.0004,
      "step": 3650
    },
    {
      "epoch": 4.075598086124402,
      "grad_norm": 0.003281336510553956,
      "learning_rate": 6.911217437533226e-06,
      "loss": 0.0003,
      "step": 3660
    },
    {
      "epoch": 4.07799043062201,
      "grad_norm": 0.02144763059914112,
      "learning_rate": 6.778309409888358e-06,
      "loss": 0.001,
      "step": 3670
    },
    {
      "epoch": 4.0803827751196176,
      "grad_norm": 0.0047360495664179325,
      "learning_rate": 6.645401382243488e-06,
      "loss": 0.0004,
      "step": 3680
    },
    {
      "epoch": 4.082775119617225,
      "grad_norm": 0.004007202573120594,
      "learning_rate": 6.512493354598619e-06,
      "loss": 0.0003,
      "step": 3690
    },
    {
      "epoch": 4.085167464114832,
      "grad_norm": 0.01407600287348032,
      "learning_rate": 6.3795853269537475e-06,
      "loss": 0.083,
      "step": 3700
    },
    {
      "epoch": 4.08755980861244,
      "grad_norm": 0.004998557735234499,
      "learning_rate": 6.246677299308879e-06,
      "loss": 0.0004,
      "step": 3710
    },
    {
      "epoch": 4.089952153110048,
      "grad_norm": 0.003679374000057578,
      "learning_rate": 6.113769271664009e-06,
      "loss": 0.0003,
      "step": 3720
    },
    {
      "epoch": 4.092344497607655,
      "grad_norm": 0.004649085458368063,
      "learning_rate": 5.980861244019139e-06,
      "loss": 0.0003,
      "step": 3730
    },
    {
      "epoch": 4.094736842105263,
      "grad_norm": 0.009950079955160618,
      "learning_rate": 5.8479532163742686e-06,
      "loss": 0.0003,
      "step": 3740
    },
    {
      "epoch": 4.097129186602871,
      "grad_norm": 0.002897342899814248,
      "learning_rate": 5.7150451887294e-06,
      "loss": 0.0004,
      "step": 3750
    },
    {
      "epoch": 4.0995215311004785,
      "grad_norm": 0.002834454644471407,
      "learning_rate": 5.5821371610845296e-06,
      "loss": 0.0004,
      "step": 3760
    },
    {
      "epoch": 4.101913875598086,
      "grad_norm": 0.0034763123840093613,
      "learning_rate": 5.44922913343966e-06,
      "loss": 0.0003,
      "step": 3770
    },
    {
      "epoch": 4.104306220095694,
      "grad_norm": 0.01105444598942995,
      "learning_rate": 5.31632110579479e-06,
      "loss": 0.0004,
      "step": 3780
    },
    {
      "epoch": 4.106698564593302,
      "grad_norm": 0.0030907553154975176,
      "learning_rate": 5.183413078149921e-06,
      "loss": 0.0019,
      "step": 3790
    },
    {
      "epoch": 4.109090909090909,
      "grad_norm": 0.00505461310967803,
      "learning_rate": 5.050505050505051e-06,
      "loss": 0.0004,
      "step": 3800
    },
    {
      "epoch": 4.111483253588517,
      "grad_norm": 0.0030872768256813288,
      "learning_rate": 4.917597022860181e-06,
      "loss": 0.0003,
      "step": 3810
    },
    {
      "epoch": 4.113875598086125,
      "grad_norm": 0.0036562816239893436,
      "learning_rate": 4.784688995215311e-06,
      "loss": 0.0004,
      "step": 3820
    },
    {
      "epoch": 4.116267942583732,
      "grad_norm": 0.00387106416746974,
      "learning_rate": 4.651780967570441e-06,
      "loss": 0.0004,
      "step": 3830
    },
    {
      "epoch": 4.118660287081339,
      "grad_norm": 0.00350612192414701,
      "learning_rate": 4.518872939925572e-06,
      "loss": 0.0004,
      "step": 3840
    },
    {
      "epoch": 4.121052631578947,
      "grad_norm": 0.015147331170737743,
      "learning_rate": 4.3859649122807014e-06,
      "loss": 0.0003,
      "step": 3850
    },
    {
      "epoch": 4.123444976076555,
      "grad_norm": 0.003980391658842564,
      "learning_rate": 4.253056884635833e-06,
      "loss": 0.0003,
      "step": 3860
    },
    {
      "epoch": 4.1258373205741625,
      "grad_norm": 0.00292576034553349,
      "learning_rate": 4.120148856990962e-06,
      "loss": 0.0003,
      "step": 3870
    },
    {
      "epoch": 4.12822966507177,
      "grad_norm": 0.003524401690810919,
      "learning_rate": 3.987240829346093e-06,
      "loss": 0.0003,
      "step": 3880
    },
    {
      "epoch": 4.130622009569378,
      "grad_norm": 0.003340751165524125,
      "learning_rate": 3.8543328017012225e-06,
      "loss": 0.0003,
      "step": 3890
    },
    {
      "epoch": 4.133014354066986,
      "grad_norm": 0.00597594678401947,
      "learning_rate": 3.7214247740563535e-06,
      "loss": 0.0003,
      "step": 3900
    },
    {
      "epoch": 4.135406698564593,
      "grad_norm": 0.40342918038368225,
      "learning_rate": 3.5885167464114835e-06,
      "loss": 0.0004,
      "step": 3910
    },
    {
      "epoch": 4.137799043062201,
      "grad_norm": 0.0027185771614313126,
      "learning_rate": 3.455608718766613e-06,
      "loss": 0.0003,
      "step": 3920
    },
    {
      "epoch": 4.140191387559809,
      "grad_norm": 0.003001193981617689,
      "learning_rate": 3.322700691121744e-06,
      "loss": 0.0003,
      "step": 3930
    },
    {
      "epoch": 4.142583732057417,
      "grad_norm": 0.011745316907763481,
      "learning_rate": 3.1897926634768737e-06,
      "loss": 0.0004,
      "step": 3940
    },
    {
      "epoch": 4.144976076555024,
      "grad_norm": 0.004574726335704327,
      "learning_rate": 3.0568846358320046e-06,
      "loss": 0.0003,
      "step": 3950
    },
    {
      "epoch": 4.147368421052631,
      "grad_norm": 0.00624841870740056,
      "learning_rate": 2.9239766081871343e-06,
      "loss": 0.0003,
      "step": 3960
    },
    {
      "epoch": 4.149760765550239,
      "grad_norm": 0.005558318924158812,
      "learning_rate": 2.7910685805422648e-06,
      "loss": 0.0003,
      "step": 3970
    },
    {
      "epoch": 4.152153110047847,
      "grad_norm": 0.0035703121684491634,
      "learning_rate": 2.658160552897395e-06,
      "loss": 0.0003,
      "step": 3980
    },
    {
      "epoch": 4.154545454545454,
      "grad_norm": 0.0034527035895735025,
      "learning_rate": 2.5252525252525253e-06,
      "loss": 0.0003,
      "step": 3990
    },
    {
      "epoch": 4.156937799043062,
      "grad_norm": 0.0036090209614485502,
      "learning_rate": 2.3923444976076554e-06,
      "loss": 0.0003,
      "step": 4000
    },
    {
      "epoch": 4.15933014354067,
      "grad_norm": 0.0026591310743242502,
      "learning_rate": 2.259436469962786e-06,
      "loss": 0.004,
      "step": 4010
    },
    {
      "epoch": 4.1617224880382775,
      "grad_norm": 0.0031338431872427464,
      "learning_rate": 2.1265284423179164e-06,
      "loss": 0.0003,
      "step": 4020
    },
    {
      "epoch": 4.164114832535885,
      "grad_norm": 0.006377640645951033,
      "learning_rate": 1.9936204146730465e-06,
      "loss": 0.0003,
      "step": 4030
    },
    {
      "epoch": 4.166507177033493,
      "grad_norm": 0.003080443711951375,
      "learning_rate": 1.8607123870281767e-06,
      "loss": 0.0003,
      "step": 4040
    },
    {
      "epoch": 4.168899521531101,
      "grad_norm": 0.1149815171957016,
      "learning_rate": 1.7278043593833066e-06,
      "loss": 0.0004,
      "step": 4050
    },
    {
      "epoch": 4.171291866028708,
      "grad_norm": 0.0024256748147308826,
      "learning_rate": 1.5948963317384369e-06,
      "loss": 0.1869,
      "step": 4060
    },
    {
      "epoch": 4.173684210526316,
      "grad_norm": 0.007068783510476351,
      "learning_rate": 1.4619883040935671e-06,
      "loss": 0.0003,
      "step": 4070
    },
    {
      "epoch": 4.176076555023924,
      "grad_norm": 0.034253306686878204,
      "learning_rate": 1.3290802764486974e-06,
      "loss": 0.0004,
      "step": 4080
    },
    {
      "epoch": 4.1784688995215316,
      "grad_norm": 0.003701072186231613,
      "learning_rate": 1.1961722488038277e-06,
      "loss": 0.0003,
      "step": 4090
    },
    {
      "epoch": 4.180861244019138,
      "grad_norm": 0.0036447285674512386,
      "learning_rate": 1.0632642211589582e-06,
      "loss": 0.0003,
      "step": 4100
    },
    {
      "epoch": 4.183253588516746,
      "grad_norm": 0.011764708906412125,
      "learning_rate": 9.303561935140884e-07,
      "loss": 0.0003,
      "step": 4110
    },
    {
      "epoch": 4.185645933014354,
      "grad_norm": 0.005143681075423956,
      "learning_rate": 7.974481658692184e-07,
      "loss": 0.0025,
      "step": 4120
    },
    {
      "epoch": 4.188038277511962,
      "grad_norm": 0.005259453784674406,
      "learning_rate": 6.645401382243487e-07,
      "loss": 0.0003,
      "step": 4130
    },
    {
      "epoch": 4.190430622009569,
      "grad_norm": 0.00764945475384593,
      "learning_rate": 5.316321105794791e-07,
      "loss": 0.0003,
      "step": 4140
    },
    {
      "epoch": 4.192822966507177,
      "grad_norm": 0.004675131756812334,
      "learning_rate": 3.987240829346092e-07,
      "loss": 0.0003,
      "step": 4150
    },
    {
      "epoch": 4.195215311004785,
      "grad_norm": 0.003127172589302063,
      "learning_rate": 2.6581605528973955e-07,
      "loss": 0.0003,
      "step": 4160
    },
    {
      "epoch": 4.1976076555023925,
      "grad_norm": 0.003473317250609398,
      "learning_rate": 1.3290802764486977e-07,
      "loss": 0.0009,
      "step": 4170
    },
    {
      "epoch": 4.2,
      "grad_norm": 0.006573684047907591,
      "learning_rate": 0.0,
      "loss": 0.0004,
      "step": 4180
    },
    {
      "epoch": 4.2,
      "eval_accuracy": 0.9985528219971056,
      "eval_loss": 0.004572847858071327,
      "eval_runtime": 792.1261,
      "eval_samples_per_second": 0.872,
      "eval_steps_per_second": 0.218,
      "step": 4180
    },
    {
      "epoch": 4.2,
      "step": 4180,
      "total_flos": 2.0835872156866314e+19,
      "train_loss": 0.2077333271433265,
      "train_runtime": 25357.3809,
      "train_samples_per_second": 0.659,
      "train_steps_per_second": 0.165
    },
    {
      "epoch": 4.2,
      "eval_accuracy": 0.9929988331388565,
      "eval_loss": 0.025693388655781746,
      "eval_runtime": 985.7172,
      "eval_samples_per_second": 0.869,
      "eval_steps_per_second": 0.218,
      "step": 4180
    },
    {
      "epoch": 4.2,
      "eval_accuracy": 0.9929988331388565,
      "eval_loss": 0.025693388655781746,
      "eval_runtime": 981.9678,
      "eval_samples_per_second": 0.873,
      "eval_steps_per_second": 0.219,
      "step": 4180
    }
  ],
  "logging_steps": 10,
  "max_steps": 4180,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 500,
  "total_flos": 2.0835872156866314e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}