{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 64.51612903225806,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3225806451612903,
      "grad_norm": 4.83341646194458,
      "learning_rate": 1.8e-07,
      "loss": 0.6146,
      "step": 10
    },
    {
      "epoch": 0.6451612903225806,
      "grad_norm": 4.601212978363037,
      "learning_rate": 3.8e-07,
      "loss": 0.6156,
      "step": 20
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 4.288501739501953,
      "learning_rate": 5.8e-07,
      "loss": 0.6041,
      "step": 30
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 3.9819118976593018,
      "learning_rate": 7.8e-07,
      "loss": 0.5691,
      "step": 40
    },
    {
      "epoch": 1.6129032258064515,
      "grad_norm": 3.2057559490203857,
      "learning_rate": 9.8e-07,
      "loss": 0.4867,
      "step": 50
    },
    {
      "epoch": 1.935483870967742,
      "grad_norm": 1.8752663135528564,
      "learning_rate": 1.18e-06,
      "loss": 0.4109,
      "step": 60
    },
    {
      "epoch": 2.258064516129032,
      "grad_norm": 1.5188192129135132,
      "learning_rate": 1.3800000000000001e-06,
      "loss": 0.3164,
      "step": 70
    },
    {
      "epoch": 2.5806451612903225,
      "grad_norm": 0.7869158983230591,
      "learning_rate": 1.5800000000000003e-06,
      "loss": 0.2642,
      "step": 80
    },
    {
      "epoch": 2.903225806451613,
      "grad_norm": 0.499815970659256,
      "learning_rate": 1.7800000000000001e-06,
      "loss": 0.2078,
      "step": 90
    },
    {
      "epoch": 3.225806451612903,
      "grad_norm": 0.36157462000846863,
      "learning_rate": 1.98e-06,
      "loss": 0.1876,
      "step": 100
    },
    {
      "epoch": 3.5483870967741935,
      "grad_norm": 0.4228762686252594,
      "learning_rate": 2.1800000000000003e-06,
      "loss": 0.167,
      "step": 110
    },
    {
      "epoch": 3.870967741935484,
      "grad_norm": 0.22709406912326813,
      "learning_rate": 2.38e-06,
      "loss": 0.1528,
      "step": 120
    },
    {
      "epoch": 4.193548387096774,
      "grad_norm": 0.2506711184978485,
      "learning_rate": 2.5800000000000003e-06,
      "loss": 0.141,
      "step": 130
    },
    {
      "epoch": 4.516129032258064,
      "grad_norm": 0.17586937546730042,
      "learning_rate": 2.78e-06,
      "loss": 0.1301,
      "step": 140
    },
    {
      "epoch": 4.838709677419355,
      "grad_norm": 0.2080131471157074,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.1252,
      "step": 150
    },
    {
      "epoch": 5.161290322580645,
      "grad_norm": 0.14357304573059082,
      "learning_rate": 3.1800000000000005e-06,
      "loss": 0.1174,
      "step": 160
    },
    {
      "epoch": 5.483870967741936,
      "grad_norm": 0.15768013894557953,
      "learning_rate": 3.38e-06,
      "loss": 0.1123,
      "step": 170
    },
    {
      "epoch": 5.806451612903226,
      "grad_norm": 0.1478148102760315,
      "learning_rate": 3.58e-06,
      "loss": 0.1059,
      "step": 180
    },
    {
      "epoch": 6.129032258064516,
      "grad_norm": 0.11791401356458664,
      "learning_rate": 3.7800000000000002e-06,
      "loss": 0.1028,
      "step": 190
    },
    {
      "epoch": 6.451612903225806,
      "grad_norm": 0.12674356997013092,
      "learning_rate": 3.98e-06,
      "loss": 0.1002,
      "step": 200
    },
    {
      "epoch": 6.774193548387097,
      "grad_norm": 0.10726606845855713,
      "learning_rate": 4.18e-06,
      "loss": 0.0936,
      "step": 210
    },
    {
      "epoch": 7.096774193548387,
      "grad_norm": 0.13341137766838074,
      "learning_rate": 4.38e-06,
      "loss": 0.093,
      "step": 220
    },
    {
      "epoch": 7.419354838709677,
      "grad_norm": 0.09624793380498886,
      "learning_rate": 4.58e-06,
      "loss": 0.0891,
      "step": 230
    },
    {
      "epoch": 7.741935483870968,
      "grad_norm": 0.14156094193458557,
      "learning_rate": 4.780000000000001e-06,
      "loss": 0.0861,
      "step": 240
    },
    {
      "epoch": 8.064516129032258,
      "grad_norm": 0.10326485335826874,
      "learning_rate": 4.98e-06,
      "loss": 0.0835,
      "step": 250
    },
    {
      "epoch": 8.387096774193548,
      "grad_norm": 0.08262120187282562,
      "learning_rate": 5.18e-06,
      "loss": 0.0805,
      "step": 260
    },
    {
      "epoch": 8.709677419354838,
      "grad_norm": 0.1273198425769806,
      "learning_rate": 5.38e-06,
      "loss": 0.0779,
      "step": 270
    },
    {
      "epoch": 9.03225806451613,
      "grad_norm": 0.12214396893978119,
      "learning_rate": 5.580000000000001e-06,
      "loss": 0.0757,
      "step": 280
    },
    {
      "epoch": 9.35483870967742,
      "grad_norm": 0.1394161581993103,
      "learning_rate": 5.78e-06,
      "loss": 0.0728,
      "step": 290
    },
    {
      "epoch": 9.67741935483871,
      "grad_norm": 0.12285201996564865,
      "learning_rate": 5.98e-06,
      "loss": 0.0722,
      "step": 300
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.15410321950912476,
      "learning_rate": 6.18e-06,
      "loss": 0.0692,
      "step": 310
    },
    {
      "epoch": 10.32258064516129,
      "grad_norm": 0.07469774037599564,
      "learning_rate": 6.38e-06,
      "loss": 0.0676,
      "step": 320
    },
    {
      "epoch": 10.64516129032258,
      "grad_norm": 0.16188545525074005,
      "learning_rate": 6.58e-06,
      "loss": 0.0643,
      "step": 330
    },
    {
      "epoch": 10.967741935483872,
      "grad_norm": 0.07787156105041504,
      "learning_rate": 6.78e-06,
      "loss": 0.0632,
      "step": 340
    },
    {
      "epoch": 11.290322580645162,
      "grad_norm": 0.13963359594345093,
      "learning_rate": 6.98e-06,
      "loss": 0.0593,
      "step": 350
    },
    {
      "epoch": 11.612903225806452,
      "grad_norm": 0.12019138038158417,
      "learning_rate": 7.180000000000001e-06,
      "loss": 0.0594,
      "step": 360
    },
    {
      "epoch": 11.935483870967742,
      "grad_norm": 0.13999758660793304,
      "learning_rate": 7.3800000000000005e-06,
      "loss": 0.0578,
      "step": 370
    },
    {
      "epoch": 12.258064516129032,
      "grad_norm": 0.11083265393972397,
      "learning_rate": 7.580000000000001e-06,
      "loss": 0.0544,
      "step": 380
    },
    {
      "epoch": 12.580645161290322,
      "grad_norm": 0.10924753546714783,
      "learning_rate": 7.78e-06,
      "loss": 0.0529,
      "step": 390
    },
    {
      "epoch": 12.903225806451612,
      "grad_norm": 0.10007922351360321,
      "learning_rate": 7.98e-06,
      "loss": 0.0519,
      "step": 400
    },
    {
      "epoch": 13.225806451612904,
      "grad_norm": 0.09409403055906296,
      "learning_rate": 8.18e-06,
      "loss": 0.0499,
      "step": 410
    },
    {
      "epoch": 13.548387096774194,
      "grad_norm": 0.1096971333026886,
      "learning_rate": 8.380000000000001e-06,
      "loss": 0.049,
      "step": 420
    },
    {
      "epoch": 13.870967741935484,
      "grad_norm": 0.10034970194101334,
      "learning_rate": 8.580000000000001e-06,
      "loss": 0.0462,
      "step": 430
    },
    {
      "epoch": 14.193548387096774,
      "grad_norm": 0.13287365436553955,
      "learning_rate": 8.78e-06,
      "loss": 0.0454,
      "step": 440
    },
    {
      "epoch": 14.516129032258064,
      "grad_norm": 0.10294642299413681,
      "learning_rate": 8.98e-06,
      "loss": 0.0436,
      "step": 450
    },
    {
      "epoch": 14.838709677419354,
      "grad_norm": 0.11570604145526886,
      "learning_rate": 9.180000000000002e-06,
      "loss": 0.0404,
      "step": 460
    },
    {
      "epoch": 15.161290322580646,
      "grad_norm": 0.08866477012634277,
      "learning_rate": 9.38e-06,
      "loss": 0.0395,
      "step": 470
    },
    {
      "epoch": 15.483870967741936,
      "grad_norm": 0.1237388625741005,
      "learning_rate": 9.58e-06,
      "loss": 0.0377,
      "step": 480
    },
    {
      "epoch": 15.806451612903226,
      "grad_norm": 0.09724581241607666,
      "learning_rate": 9.78e-06,
      "loss": 0.0364,
      "step": 490
    },
    {
      "epoch": 16.129032258064516,
      "grad_norm": 0.1216067299246788,
      "learning_rate": 9.980000000000001e-06,
      "loss": 0.0369,
      "step": 500
    },
    {
      "epoch": 16.451612903225808,
      "grad_norm": 0.14183978736400604,
      "learning_rate": 1.018e-05,
      "loss": 0.0335,
      "step": 510
    },
    {
      "epoch": 16.774193548387096,
      "grad_norm": 0.10971418768167496,
      "learning_rate": 1.038e-05,
      "loss": 0.0337,
      "step": 520
    },
    {
      "epoch": 17.096774193548388,
      "grad_norm": 0.10205938667058945,
      "learning_rate": 1.058e-05,
      "loss": 0.0314,
      "step": 530
    },
    {
      "epoch": 17.419354838709676,
      "grad_norm": 0.08642183989286423,
      "learning_rate": 1.0780000000000002e-05,
      "loss": 0.0309,
      "step": 540
    },
    {
      "epoch": 17.741935483870968,
      "grad_norm": 0.11878576874732971,
      "learning_rate": 1.098e-05,
      "loss": 0.0313,
      "step": 550
    },
    {
      "epoch": 18.06451612903226,
      "grad_norm": 0.11662951111793518,
      "learning_rate": 1.118e-05,
      "loss": 0.03,
      "step": 560
    },
    {
      "epoch": 18.387096774193548,
      "grad_norm": 0.13768725097179413,
      "learning_rate": 1.1380000000000001e-05,
      "loss": 0.0285,
      "step": 570
    },
    {
      "epoch": 18.70967741935484,
      "grad_norm": 0.1288958489894867,
      "learning_rate": 1.1580000000000001e-05,
      "loss": 0.0288,
      "step": 580
    },
    {
      "epoch": 19.032258064516128,
      "grad_norm": 0.11418294161558151,
      "learning_rate": 1.178e-05,
      "loss": 0.0289,
      "step": 590
    },
    {
      "epoch": 19.35483870967742,
      "grad_norm": 0.12079351395368576,
      "learning_rate": 1.198e-05,
      "loss": 0.0287,
      "step": 600
    },
    {
      "epoch": 19.677419354838708,
      "grad_norm": 0.2023964524269104,
      "learning_rate": 1.2180000000000002e-05,
      "loss": 0.0279,
      "step": 610
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.19349494576454163,
      "learning_rate": 1.238e-05,
      "loss": 0.0278,
      "step": 620
    },
    {
      "epoch": 20.322580645161292,
      "grad_norm": 0.17760895192623138,
      "learning_rate": 1.258e-05,
      "loss": 0.0273,
      "step": 630
    },
    {
      "epoch": 20.64516129032258,
      "grad_norm": 0.1515321135520935,
      "learning_rate": 1.278e-05,
      "loss": 0.0257,
      "step": 640
    },
    {
      "epoch": 20.967741935483872,
      "grad_norm": 0.10156595706939697,
      "learning_rate": 1.2980000000000001e-05,
      "loss": 0.0257,
      "step": 650
    },
    {
      "epoch": 21.29032258064516,
      "grad_norm": 0.12676751613616943,
      "learning_rate": 1.3180000000000001e-05,
      "loss": 0.0249,
      "step": 660
    },
    {
      "epoch": 21.612903225806452,
      "grad_norm": 0.13134433329105377,
      "learning_rate": 1.338e-05,
      "loss": 0.0239,
      "step": 670
    },
    {
      "epoch": 21.93548387096774,
      "grad_norm": 0.13195613026618958,
      "learning_rate": 1.358e-05,
      "loss": 0.0242,
      "step": 680
    },
    {
      "epoch": 22.258064516129032,
      "grad_norm": 0.12307468056678772,
      "learning_rate": 1.3780000000000002e-05,
      "loss": 0.0238,
      "step": 690
    },
    {
      "epoch": 22.580645161290324,
      "grad_norm": 0.1110624298453331,
      "learning_rate": 1.3980000000000002e-05,
      "loss": 0.0253,
      "step": 700
    },
    {
      "epoch": 22.903225806451612,
      "grad_norm": 0.13264206051826477,
      "learning_rate": 1.4180000000000001e-05,
      "loss": 0.0239,
      "step": 710
    },
    {
      "epoch": 23.225806451612904,
      "grad_norm": 0.11219354718923569,
      "learning_rate": 1.4380000000000001e-05,
      "loss": 0.0234,
      "step": 720
    },
    {
      "epoch": 23.548387096774192,
      "grad_norm": 0.14356958866119385,
      "learning_rate": 1.4580000000000003e-05,
      "loss": 0.0225,
      "step": 730
    },
    {
      "epoch": 23.870967741935484,
      "grad_norm": 0.13088181614875793,
      "learning_rate": 1.4779999999999999e-05,
      "loss": 0.0221,
      "step": 740
    },
    {
      "epoch": 24.193548387096776,
      "grad_norm": 0.15026481449604034,
      "learning_rate": 1.4979999999999999e-05,
      "loss": 0.0225,
      "step": 750
    },
    {
      "epoch": 24.516129032258064,
      "grad_norm": 0.20722658932209015,
      "learning_rate": 1.518e-05,
      "loss": 0.0225,
      "step": 760
    },
    {
      "epoch": 24.838709677419356,
      "grad_norm": 0.13866043090820312,
      "learning_rate": 1.538e-05,
      "loss": 0.0221,
      "step": 770
    },
    {
      "epoch": 25.161290322580644,
      "grad_norm": 0.11805257201194763,
      "learning_rate": 1.558e-05,
      "loss": 0.0214,
      "step": 780
    },
    {
      "epoch": 25.483870967741936,
      "grad_norm": 0.1360214352607727,
      "learning_rate": 1.578e-05,
      "loss": 0.0205,
      "step": 790
    },
    {
      "epoch": 25.806451612903224,
      "grad_norm": 0.15285663306713104,
      "learning_rate": 1.598e-05,
      "loss": 0.0215,
      "step": 800
    },
    {
      "epoch": 26.129032258064516,
      "grad_norm": 0.13831627368927002,
      "learning_rate": 1.618e-05,
      "loss": 0.021,
      "step": 810
    },
    {
      "epoch": 26.451612903225808,
      "grad_norm": 0.1219727173447609,
      "learning_rate": 1.6380000000000002e-05,
      "loss": 0.0208,
      "step": 820
    },
    {
      "epoch": 26.774193548387096,
      "grad_norm": 0.14780209958553314,
      "learning_rate": 1.658e-05,
      "loss": 0.0212,
      "step": 830
    },
    {
      "epoch": 27.096774193548388,
      "grad_norm": 0.13057461380958557,
      "learning_rate": 1.6780000000000002e-05,
      "loss": 0.0206,
      "step": 840
    },
    {
      "epoch": 27.419354838709676,
      "grad_norm": 0.13308848440647125,
      "learning_rate": 1.698e-05,
      "loss": 0.021,
      "step": 850
    },
    {
      "epoch": 27.741935483870968,
      "grad_norm": 0.1406048834323883,
      "learning_rate": 1.718e-05,
      "loss": 0.0204,
      "step": 860
    },
    {
      "epoch": 28.06451612903226,
      "grad_norm": 0.1681479662656784,
      "learning_rate": 1.7380000000000003e-05,
      "loss": 0.0203,
      "step": 870
    },
    {
      "epoch": 28.387096774193548,
      "grad_norm": 0.15317517518997192,
      "learning_rate": 1.758e-05,
      "loss": 0.0192,
      "step": 880
    },
    {
      "epoch": 28.70967741935484,
      "grad_norm": 0.17327739298343658,
      "learning_rate": 1.7780000000000003e-05,
      "loss": 0.0191,
      "step": 890
    },
    {
      "epoch": 29.032258064516128,
      "grad_norm": 0.13825540244579315,
      "learning_rate": 1.798e-05,
      "loss": 0.0194,
      "step": 900
    },
    {
      "epoch": 29.35483870967742,
      "grad_norm": 0.14294251799583435,
      "learning_rate": 1.818e-05,
      "loss": 0.0191,
      "step": 910
    },
    {
      "epoch": 29.677419354838708,
      "grad_norm": 0.1148597002029419,
      "learning_rate": 1.838e-05,
      "loss": 0.0194,
      "step": 920
    },
    {
      "epoch": 30.0,
      "grad_norm": 0.10468383878469467,
      "learning_rate": 1.858e-05,
      "loss": 0.019,
      "step": 930
    },
    {
      "epoch": 30.322580645161292,
      "grad_norm": 0.16523204743862152,
      "learning_rate": 1.878e-05,
      "loss": 0.0183,
      "step": 940
    },
    {
      "epoch": 30.64516129032258,
      "grad_norm": 0.1483301818370819,
      "learning_rate": 1.898e-05,
      "loss": 0.0183,
      "step": 950
    },
    {
      "epoch": 30.967741935483872,
      "grad_norm": 0.19092631340026855,
      "learning_rate": 1.918e-05,
      "loss": 0.0188,
      "step": 960
    },
    {
      "epoch": 31.29032258064516,
      "grad_norm": 0.19763915240764618,
      "learning_rate": 1.938e-05,
      "loss": 0.0181,
      "step": 970
    },
    {
      "epoch": 31.612903225806452,
      "grad_norm": 0.18357335031032562,
      "learning_rate": 1.9580000000000002e-05,
      "loss": 0.0179,
      "step": 980
    },
    {
      "epoch": 31.93548387096774,
      "grad_norm": 0.1674022376537323,
      "learning_rate": 1.978e-05,
      "loss": 0.0181,
      "step": 990
    },
    {
      "epoch": 32.25806451612903,
      "grad_norm": 0.1773853302001953,
      "learning_rate": 1.9980000000000002e-05,
      "loss": 0.0187,
      "step": 1000
    },
    {
      "epoch": 32.58064516129032,
      "grad_norm": 0.17372146248817444,
      "learning_rate": 2.0180000000000003e-05,
      "loss": 0.0175,
      "step": 1010
    },
    {
      "epoch": 32.903225806451616,
      "grad_norm": 0.16912385821342468,
      "learning_rate": 2.038e-05,
      "loss": 0.0177,
      "step": 1020
    },
    {
      "epoch": 33.225806451612904,
      "grad_norm": 0.1924871802330017,
      "learning_rate": 2.0580000000000003e-05,
      "loss": 0.017,
      "step": 1030
    },
    {
      "epoch": 33.54838709677419,
      "grad_norm": 0.1980140060186386,
      "learning_rate": 2.078e-05,
      "loss": 0.0174,
      "step": 1040
    },
    {
      "epoch": 33.87096774193548,
      "grad_norm": 0.15361692011356354,
      "learning_rate": 2.098e-05,
      "loss": 0.0169,
      "step": 1050
    },
    {
      "epoch": 34.193548387096776,
      "grad_norm": 0.15744690597057343,
      "learning_rate": 2.118e-05,
      "loss": 0.0174,
      "step": 1060
    },
    {
      "epoch": 34.516129032258064,
      "grad_norm": 0.18941748142242432,
      "learning_rate": 2.138e-05,
      "loss": 0.0173,
      "step": 1070
    },
    {
      "epoch": 34.83870967741935,
      "grad_norm": 0.18401823937892914,
      "learning_rate": 2.158e-05,
      "loss": 0.0176,
      "step": 1080
    },
    {
      "epoch": 35.16129032258065,
      "grad_norm": 0.1347324103116989,
      "learning_rate": 2.178e-05,
      "loss": 0.017,
      "step": 1090
    },
    {
      "epoch": 35.483870967741936,
      "grad_norm": 0.13733318448066711,
      "learning_rate": 2.198e-05,
      "loss": 0.0163,
      "step": 1100
    },
    {
      "epoch": 35.806451612903224,
      "grad_norm": 0.17490868270397186,
      "learning_rate": 2.218e-05,
      "loss": 0.0167,
      "step": 1110
    },
    {
      "epoch": 36.12903225806452,
      "grad_norm": 0.17697739601135254,
      "learning_rate": 2.2380000000000003e-05,
      "loss": 0.0173,
      "step": 1120
    },
    {
      "epoch": 36.45161290322581,
      "grad_norm": 0.19351543486118317,
      "learning_rate": 2.258e-05,
      "loss": 0.0174,
      "step": 1130
    },
    {
      "epoch": 36.774193548387096,
      "grad_norm": 0.17104071378707886,
      "learning_rate": 2.2780000000000002e-05,
      "loss": 0.0159,
      "step": 1140
    },
    {
      "epoch": 37.096774193548384,
      "grad_norm": 0.1536385416984558,
      "learning_rate": 2.298e-05,
      "loss": 0.016,
      "step": 1150
    },
    {
      "epoch": 37.41935483870968,
      "grad_norm": 0.14215905964374542,
      "learning_rate": 2.318e-05,
      "loss": 0.0167,
      "step": 1160
    },
    {
      "epoch": 37.74193548387097,
      "grad_norm": 0.17754510045051575,
      "learning_rate": 2.3380000000000003e-05,
      "loss": 0.0163,
      "step": 1170
    },
    {
      "epoch": 38.064516129032256,
      "grad_norm": 0.15013834834098816,
      "learning_rate": 2.358e-05,
      "loss": 0.0152,
      "step": 1180
    },
    {
      "epoch": 38.38709677419355,
      "grad_norm": 0.17872834205627441,
      "learning_rate": 2.3780000000000003e-05,
      "loss": 0.0152,
      "step": 1190
    },
    {
      "epoch": 38.70967741935484,
      "grad_norm": 0.16451333463191986,
      "learning_rate": 2.398e-05,
      "loss": 0.015,
      "step": 1200
    },
    {
      "epoch": 39.03225806451613,
      "grad_norm": 0.17044350504875183,
      "learning_rate": 2.418e-05,
      "loss": 0.0149,
      "step": 1210
    },
    {
      "epoch": 39.354838709677416,
      "grad_norm": 0.16590650379657745,
      "learning_rate": 2.438e-05,
      "loss": 0.0155,
      "step": 1220
    },
    {
      "epoch": 39.67741935483871,
      "grad_norm": 0.1371670961380005,
      "learning_rate": 2.4580000000000002e-05,
      "loss": 0.015,
      "step": 1230
    },
    {
      "epoch": 40.0,
      "grad_norm": 0.11834592372179031,
      "learning_rate": 2.478e-05,
      "loss": 0.0158,
      "step": 1240
    },
    {
      "epoch": 40.32258064516129,
      "grad_norm": 0.17337927222251892,
      "learning_rate": 2.498e-05,
      "loss": 0.0148,
      "step": 1250
    },
    {
      "epoch": 40.645161290322584,
      "grad_norm": 0.17919541895389557,
      "learning_rate": 2.5180000000000003e-05,
      "loss": 0.0146,
      "step": 1260
    },
    {
      "epoch": 40.96774193548387,
      "grad_norm": 0.1446114331483841,
      "learning_rate": 2.5380000000000004e-05,
      "loss": 0.0144,
      "step": 1270
    },
    {
      "epoch": 41.29032258064516,
      "grad_norm": 0.12958332896232605,
      "learning_rate": 2.5580000000000002e-05,
      "loss": 0.0143,
      "step": 1280
    },
    {
      "epoch": 41.61290322580645,
      "grad_norm": 0.208790123462677,
      "learning_rate": 2.5779999999999997e-05,
      "loss": 0.0145,
      "step": 1290
    },
    {
      "epoch": 41.935483870967744,
      "grad_norm": 0.13380534946918488,
      "learning_rate": 2.598e-05,
      "loss": 0.0144,
      "step": 1300
    },
    {
      "epoch": 42.25806451612903,
      "grad_norm": 0.202678382396698,
      "learning_rate": 2.618e-05,
      "loss": 0.0137,
      "step": 1310
    },
    {
      "epoch": 42.58064516129032,
      "grad_norm": 0.19495125114917755,
      "learning_rate": 2.6379999999999998e-05,
      "loss": 0.0147,
      "step": 1320
    },
    {
      "epoch": 42.903225806451616,
      "grad_norm": 0.17326615750789642,
      "learning_rate": 2.658e-05,
      "loss": 0.0143,
      "step": 1330
    },
    {
      "epoch": 43.225806451612904,
      "grad_norm": 0.17888951301574707,
      "learning_rate": 2.678e-05,
      "loss": 0.0143,
      "step": 1340
    },
    {
      "epoch": 43.54838709677419,
      "grad_norm": 0.1571645885705948,
      "learning_rate": 2.698e-05,
      "loss": 0.0136,
      "step": 1350
    },
    {
      "epoch": 43.87096774193548,
      "grad_norm": 0.13882234692573547,
      "learning_rate": 2.718e-05,
      "loss": 0.0136,
      "step": 1360
    },
    {
      "epoch": 44.193548387096776,
      "grad_norm": 0.12413440644741058,
      "learning_rate": 2.738e-05,
      "loss": 0.0132,
      "step": 1370
    },
    {
      "epoch": 44.516129032258064,
      "grad_norm": 0.13164107501506805,
      "learning_rate": 2.758e-05,
      "loss": 0.0138,
      "step": 1380
    },
    {
      "epoch": 44.83870967741935,
      "grad_norm": 0.14705562591552734,
      "learning_rate": 2.778e-05,
      "loss": 0.0135,
      "step": 1390
    },
    {
      "epoch": 45.16129032258065,
      "grad_norm": 0.17040109634399414,
      "learning_rate": 2.798e-05,
      "loss": 0.0136,
      "step": 1400
    },
    {
      "epoch": 45.483870967741936,
      "grad_norm": 0.17892764508724213,
      "learning_rate": 2.818e-05,
      "loss": 0.0135,
      "step": 1410
    },
    {
      "epoch": 45.806451612903224,
      "grad_norm": 0.14825959503650665,
      "learning_rate": 2.8380000000000003e-05,
      "loss": 0.013,
      "step": 1420
    },
    {
      "epoch": 46.12903225806452,
      "grad_norm": 0.11769016087055206,
      "learning_rate": 2.858e-05,
      "loss": 0.0134,
      "step": 1430
    },
    {
      "epoch": 46.45161290322581,
      "grad_norm": 0.16785204410552979,
      "learning_rate": 2.8780000000000002e-05,
      "loss": 0.0131,
      "step": 1440
    },
    {
      "epoch": 46.774193548387096,
      "grad_norm": 0.12980397045612335,
      "learning_rate": 2.898e-05,
      "loss": 0.0141,
      "step": 1450
    },
    {
      "epoch": 47.096774193548384,
      "grad_norm": 0.19282320141792297,
      "learning_rate": 2.9180000000000002e-05,
      "loss": 0.014,
      "step": 1460
    },
    {
      "epoch": 47.41935483870968,
      "grad_norm": 0.17180287837982178,
      "learning_rate": 2.9380000000000003e-05,
      "loss": 0.0133,
      "step": 1470
    },
    {
      "epoch": 47.74193548387097,
      "grad_norm": 0.1480427086353302,
      "learning_rate": 2.958e-05,
      "loss": 0.0133,
      "step": 1480
    },
    {
      "epoch": 48.064516129032256,
      "grad_norm": 0.18842412531375885,
      "learning_rate": 2.9780000000000003e-05,
      "loss": 0.0135,
      "step": 1490
    },
    {
      "epoch": 48.38709677419355,
      "grad_norm": 0.2088242620229721,
      "learning_rate": 2.998e-05,
      "loss": 0.0142,
      "step": 1500
    },
    {
      "epoch": 48.70967741935484,
      "grad_norm": 0.1986250877380371,
      "learning_rate": 3.0180000000000002e-05,
      "loss": 0.0128,
      "step": 1510
    },
    {
      "epoch": 49.03225806451613,
      "grad_norm": 0.1687200963497162,
      "learning_rate": 3.0380000000000004e-05,
      "loss": 0.0127,
      "step": 1520
    },
    {
      "epoch": 49.354838709677416,
      "grad_norm": 0.1558302789926529,
      "learning_rate": 3.058e-05,
      "loss": 0.0131,
      "step": 1530
    },
    {
      "epoch": 49.67741935483871,
      "grad_norm": 0.17784525454044342,
      "learning_rate": 3.078e-05,
      "loss": 0.0125,
      "step": 1540
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.22291983664035797,
      "learning_rate": 3.0980000000000005e-05,
      "loss": 0.0121,
      "step": 1550
    },
    {
      "epoch": 50.32258064516129,
      "grad_norm": 0.21824175119400024,
      "learning_rate": 3.118e-05,
      "loss": 0.012,
      "step": 1560
    },
    {
      "epoch": 50.645161290322584,
      "grad_norm": 0.2027992457151413,
      "learning_rate": 3.138e-05,
      "loss": 0.0127,
      "step": 1570
    },
    {
      "epoch": 50.96774193548387,
      "grad_norm": 0.2533424496650696,
      "learning_rate": 3.1580000000000006e-05,
      "loss": 0.0121,
      "step": 1580
    },
    {
      "epoch": 51.29032258064516,
      "grad_norm": 0.15423865616321564,
      "learning_rate": 3.1780000000000004e-05,
      "loss": 0.0125,
      "step": 1590
    },
    {
      "epoch": 51.61290322580645,
      "grad_norm": 0.13885457813739777,
      "learning_rate": 3.198e-05,
      "loss": 0.013,
      "step": 1600
    },
    {
      "epoch": 51.935483870967744,
      "grad_norm": 0.1417354941368103,
      "learning_rate": 3.218e-05,
      "loss": 0.0119,
      "step": 1610
    },
    {
      "epoch": 52.25806451612903,
      "grad_norm": 0.17335128784179688,
      "learning_rate": 3.238e-05,
      "loss": 0.012,
      "step": 1620
    },
    {
      "epoch": 52.58064516129032,
      "grad_norm": 0.19661465287208557,
      "learning_rate": 3.2579999999999996e-05,
      "loss": 0.0119,
      "step": 1630
    },
    {
      "epoch": 52.903225806451616,
      "grad_norm": 0.2602691650390625,
      "learning_rate": 3.278e-05,
      "loss": 0.0122,
      "step": 1640
    },
    {
      "epoch": 53.225806451612904,
      "grad_norm": 0.21529579162597656,
      "learning_rate": 3.298e-05,
      "loss": 0.0117,
      "step": 1650
    },
    {
      "epoch": 53.54838709677419,
      "grad_norm": 0.1529514640569687,
      "learning_rate": 3.318e-05,
      "loss": 0.0116,
      "step": 1660
    },
    {
      "epoch": 53.87096774193548,
      "grad_norm": 0.1834433674812317,
      "learning_rate": 3.338e-05,
      "loss": 0.0125,
      "step": 1670
    },
    {
      "epoch": 54.193548387096776,
      "grad_norm": 0.24152186512947083,
      "learning_rate": 3.358e-05,
      "loss": 0.0122,
      "step": 1680
    },
    {
      "epoch": 54.516129032258064,
      "grad_norm": 0.1319815069437027,
      "learning_rate": 3.378e-05,
      "loss": 0.0116,
      "step": 1690
    },
    {
      "epoch": 54.83870967741935,
      "grad_norm": 0.16622133553028107,
      "learning_rate": 3.398e-05,
      "loss": 0.0123,
      "step": 1700
    },
    {
      "epoch": 55.16129032258065,
      "grad_norm": 0.12881875038146973,
      "learning_rate": 3.418e-05,
      "loss": 0.0115,
      "step": 1710
    },
    {
      "epoch": 55.483870967741936,
      "grad_norm": 0.16083110868930817,
      "learning_rate": 3.438e-05,
      "loss": 0.0121,
      "step": 1720
    },
    {
      "epoch": 55.806451612903224,
      "grad_norm": 0.17213866114616394,
      "learning_rate": 3.4580000000000004e-05,
      "loss": 0.0111,
      "step": 1730
    },
    {
      "epoch": 56.12903225806452,
      "grad_norm": 0.1291566640138626,
      "learning_rate": 3.478e-05,
      "loss": 0.0113,
      "step": 1740
    },
    {
      "epoch": 56.45161290322581,
      "grad_norm": 0.1637270748615265,
      "learning_rate": 3.498e-05,
      "loss": 0.011,
      "step": 1750
    },
    {
      "epoch": 56.774193548387096,
      "grad_norm": 0.1748403012752533,
      "learning_rate": 3.518e-05,
      "loss": 0.0108,
      "step": 1760
    },
    {
      "epoch": 57.096774193548384,
      "grad_norm": 0.18834678828716278,
      "learning_rate": 3.5380000000000003e-05,
      "loss": 0.0109,
      "step": 1770
    },
    {
      "epoch": 57.41935483870968,
      "grad_norm": 0.1650463044643402,
      "learning_rate": 3.558e-05,
      "loss": 0.0115,
      "step": 1780
    },
    {
      "epoch": 57.74193548387097,
      "grad_norm": 0.17039620876312256,
      "learning_rate": 3.578e-05,
      "loss": 0.0114,
      "step": 1790
    },
    {
      "epoch": 58.064516129032256,
      "grad_norm": 0.13908374309539795,
      "learning_rate": 3.5980000000000004e-05,
      "loss": 0.0107,
      "step": 1800
    },
    {
      "epoch": 58.38709677419355,
      "grad_norm": 0.1755029857158661,
      "learning_rate": 3.618e-05,
      "loss": 0.0104,
      "step": 1810
    },
    {
      "epoch": 58.70967741935484,
      "grad_norm": 0.16456303000450134,
      "learning_rate": 3.638e-05,
      "loss": 0.011,
      "step": 1820
    },
    {
      "epoch": 59.03225806451613,
      "grad_norm": 0.1215883269906044,
      "learning_rate": 3.6580000000000006e-05,
      "loss": 0.0106,
      "step": 1830
    },
    {
      "epoch": 59.354838709677416,
      "grad_norm": 0.13105925917625427,
      "learning_rate": 3.6780000000000004e-05,
      "loss": 0.0104,
      "step": 1840
    },
    {
      "epoch": 59.67741935483871,
      "grad_norm": 0.14026036858558655,
      "learning_rate": 3.698e-05,
      "loss": 0.0105,
      "step": 1850
    },
    {
      "epoch": 60.0,
      "grad_norm": 0.15687642991542816,
      "learning_rate": 3.7180000000000007e-05,
      "loss": 0.0105,
      "step": 1860
    },
    {
      "epoch": 60.32258064516129,
      "grad_norm": 0.1946769654750824,
      "learning_rate": 3.7380000000000005e-05,
      "loss": 0.011,
      "step": 1870
    },
    {
      "epoch": 60.645161290322584,
      "grad_norm": 0.2352798581123352,
      "learning_rate": 3.758e-05,
      "loss": 0.0108,
      "step": 1880
    },
    {
      "epoch": 60.96774193548387,
      "grad_norm": 0.2311507761478424,
      "learning_rate": 3.778000000000001e-05,
      "loss": 0.0119,
      "step": 1890
    },
    {
      "epoch": 61.29032258064516,
      "grad_norm": 0.24302849173545837,
      "learning_rate": 3.7980000000000006e-05,
      "loss": 0.0104,
      "step": 1900
    },
    {
      "epoch": 61.61290322580645,
      "grad_norm": 0.20787030458450317,
      "learning_rate": 3.818e-05,
      "loss": 0.011,
      "step": 1910
    },
    {
      "epoch": 61.935483870967744,
      "grad_norm": 0.21964043378829956,
      "learning_rate": 3.838e-05,
      "loss": 0.0105,
      "step": 1920
    },
    {
      "epoch": 62.25806451612903,
      "grad_norm": 0.17259149253368378,
      "learning_rate": 3.858e-05,
      "loss": 0.011,
      "step": 1930
    },
    {
      "epoch": 62.58064516129032,
      "grad_norm": 0.2735665440559387,
      "learning_rate": 3.878e-05,
      "loss": 0.0108,
      "step": 1940
    },
    {
      "epoch": 62.903225806451616,
      "grad_norm": 0.19282570481300354,
      "learning_rate": 3.898e-05,
      "loss": 0.0108,
      "step": 1950
    },
    {
      "epoch": 63.225806451612904,
      "grad_norm": 0.21735602617263794,
      "learning_rate": 3.918e-05,
      "loss": 0.0098,
      "step": 1960
    },
    {
      "epoch": 63.54838709677419,
      "grad_norm": 0.192997008562088,
      "learning_rate": 3.938e-05,
      "loss": 0.0101,
      "step": 1970
    },
    {
      "epoch": 63.87096774193548,
      "grad_norm": 0.21661442518234253,
      "learning_rate": 3.958e-05,
      "loss": 0.0107,
      "step": 1980
    },
    {
      "epoch": 64.19354838709677,
      "grad_norm": 0.23013737797737122,
      "learning_rate": 3.978e-05,
      "loss": 0.0106,
      "step": 1990
    },
    {
      "epoch": 64.51612903225806,
      "grad_norm": 0.17058345675468445,
      "learning_rate": 3.998e-05,
      "loss": 0.0101,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3226,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 512,
  "trial_name": null,
  "trial_params": null
}