{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0023401221251234047,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.925152656404256e-05,
      "grad_norm": 44.091312408447266,
      "learning_rate": 4.1999999999999995e-07,
      "loss": 9.3692,
      "step": 10
    },
    {
      "epoch": 5.850305312808512e-05,
      "grad_norm": 44.14885711669922,
      "learning_rate": 1.0199999999999998e-06,
      "loss": 9.3095,
      "step": 20
    },
    {
      "epoch": 8.775457969212769e-05,
      "grad_norm": 71.31829071044922,
      "learning_rate": 1.62e-06,
      "loss": 9.2848,
      "step": 30
    },
    {
      "epoch": 0.00011700610625617024,
      "grad_norm": 44.88148498535156,
      "learning_rate": 2.22e-06,
      "loss": 9.3214,
      "step": 40
    },
    {
      "epoch": 0.0001462576328202128,
      "grad_norm": 71.73838806152344,
      "learning_rate": 2.8199999999999997e-06,
      "loss": 9.2992,
      "step": 50
    },
    {
      "epoch": 0.00017550915938425538,
      "grad_norm": 63.85468292236328,
      "learning_rate": 3.42e-06,
      "loss": 9.3034,
      "step": 60
    },
    {
      "epoch": 0.00020476068594829793,
      "grad_norm": 47.08655548095703,
      "learning_rate": 4.02e-06,
      "loss": 9.2972,
      "step": 70
    },
    {
      "epoch": 0.00023401221251234048,
      "grad_norm": 46.76586151123047,
      "learning_rate": 4.62e-06,
      "loss": 9.282,
      "step": 80
    },
    {
      "epoch": 0.00026326373907638306,
      "grad_norm": 68.871826171875,
      "learning_rate": 5.219999999999999e-06,
      "loss": 9.2683,
      "step": 90
    },
    {
      "epoch": 0.0002925152656404256,
      "grad_norm": 40.276512145996094,
      "learning_rate": 5.819999999999999e-06,
      "loss": 9.2803,
      "step": 100
    },
    {
      "epoch": 0.00032176679220446817,
      "grad_norm": 43.05030059814453,
      "learning_rate": 6.4199999999999995e-06,
      "loss": 9.2783,
      "step": 110
    },
    {
      "epoch": 0.00035101831876851075,
      "grad_norm": 48.496463775634766,
      "learning_rate": 7.02e-06,
      "loss": 9.2473,
      "step": 120
    },
    {
      "epoch": 0.0003802698453325533,
      "grad_norm": 56.55481719970703,
      "learning_rate": 7.619999999999999e-06,
      "loss": 9.2961,
      "step": 130
    },
    {
      "epoch": 0.00040952137189659586,
      "grad_norm": 35.09830093383789,
      "learning_rate": 8.22e-06,
      "loss": 9.2656,
      "step": 140
    },
    {
      "epoch": 0.00043877289846063844,
      "grad_norm": 38.8115348815918,
      "learning_rate": 8.819999999999999e-06,
      "loss": 9.2461,
      "step": 150
    },
    {
      "epoch": 0.00046802442502468097,
      "grad_norm": 38.47610855102539,
      "learning_rate": 9.419999999999998e-06,
      "loss": 9.2523,
      "step": 160
    },
    {
      "epoch": 0.0004972759515887235,
      "grad_norm": 35.73860549926758,
      "learning_rate": 1.0019999999999999e-05,
      "loss": 9.2434,
      "step": 170
    },
    {
      "epoch": 0.0005265274781527661,
      "grad_norm": 35.11371994018555,
      "learning_rate": 1.062e-05,
      "loss": 9.2298,
      "step": 180
    },
    {
      "epoch": 0.0005557790047168087,
      "grad_norm": 41.20671844482422,
      "learning_rate": 1.122e-05,
      "loss": 9.2479,
      "step": 190
    },
    {
      "epoch": 0.0005850305312808512,
      "grad_norm": 31.733139038085938,
      "learning_rate": 1.1819999999999999e-05,
      "loss": 9.2436,
      "step": 200
    },
    {
      "epoch": 0.0006142820578448938,
      "grad_norm": 30.686948776245117,
      "learning_rate": 1.2419999999999998e-05,
      "loss": 9.2533,
      "step": 210
    },
    {
      "epoch": 0.0006435335844089363,
      "grad_norm": 28.569021224975586,
      "learning_rate": 1.3019999999999999e-05,
      "loss": 9.2148,
      "step": 220
    },
    {
      "epoch": 0.0006727851109729789,
      "grad_norm": 40.037601470947266,
      "learning_rate": 1.362e-05,
      "loss": 9.21,
      "step": 230
    },
    {
      "epoch": 0.0007020366375370215,
      "grad_norm": 63.61598587036133,
      "learning_rate": 1.4219999999999998e-05,
      "loss": 9.2505,
      "step": 240
    },
    {
      "epoch": 0.000731288164101064,
      "grad_norm": 29.98544692993164,
      "learning_rate": 1.4819999999999999e-05,
      "loss": 9.2188,
      "step": 250
    },
    {
      "epoch": 0.0007605396906651066,
      "grad_norm": 30.23026466369629,
      "learning_rate": 1.5419999999999998e-05,
      "loss": 9.1822,
      "step": 260
    },
    {
      "epoch": 0.0007897912172291491,
      "grad_norm": 21.082069396972656,
      "learning_rate": 1.602e-05,
      "loss": 9.2109,
      "step": 270
    },
    {
      "epoch": 0.0008190427437931917,
      "grad_norm": 34.52238464355469,
      "learning_rate": 1.6619999999999997e-05,
      "loss": 9.178,
      "step": 280
    },
    {
      "epoch": 0.0008482942703572343,
      "grad_norm": 54.66425704956055,
      "learning_rate": 1.7219999999999998e-05,
      "loss": 9.2251,
      "step": 290
    },
    {
      "epoch": 0.0008775457969212769,
      "grad_norm": 32.79227066040039,
      "learning_rate": 1.782e-05,
      "loss": 9.1966,
      "step": 300
    },
    {
      "epoch": 0.0009067973234853194,
      "grad_norm": 29.83945655822754,
      "learning_rate": 1.842e-05,
      "loss": 9.2206,
      "step": 310
    },
    {
      "epoch": 0.0009360488500493619,
      "grad_norm": 16.18548583984375,
      "learning_rate": 1.9019999999999997e-05,
      "loss": 9.2005,
      "step": 320
    },
    {
      "epoch": 0.0009653003766134045,
      "grad_norm": 27.39173698425293,
      "learning_rate": 1.962e-05,
      "loss": 9.2079,
      "step": 330
    },
    {
      "epoch": 0.000994551903177447,
      "grad_norm": 23.602752685546875,
      "learning_rate": 2.022e-05,
      "loss": 9.2022,
      "step": 340
    },
    {
      "epoch": 0.0010238034297414896,
      "grad_norm": 21.416629791259766,
      "learning_rate": 2.082e-05,
      "loss": 9.1674,
      "step": 350
    },
    {
      "epoch": 0.0010530549563055323,
      "grad_norm": Infinity,
      "learning_rate": 2.1419999999999998e-05,
      "loss": 9.1761,
      "step": 360
    },
    {
      "epoch": 0.0010823064828695747,
      "grad_norm": 27.356834411621094,
      "learning_rate": 2.196e-05,
      "loss": 9.2058,
      "step": 370
    },
    {
      "epoch": 0.0011115580094336174,
      "grad_norm": 24.82968521118164,
      "learning_rate": 2.2559999999999997e-05,
      "loss": 9.1643,
      "step": 380
    },
    {
      "epoch": 0.00114080953599766,
      "grad_norm": 19.21381187438965,
      "learning_rate": 2.316e-05,
      "loss": 9.1706,
      "step": 390
    },
    {
      "epoch": 0.0011700610625617024,
      "grad_norm": 18.351755142211914,
      "learning_rate": 2.3699999999999997e-05,
      "loss": 9.1917,
      "step": 400
    },
    {
      "epoch": 0.001199312589125745,
      "grad_norm": 17.74287223815918,
      "learning_rate": 2.4299999999999998e-05,
      "loss": 9.18,
      "step": 410
    },
    {
      "epoch": 0.0012285641156897875,
      "grad_norm": 24.708906173706055,
      "learning_rate": 2.49e-05,
      "loss": 9.2101,
      "step": 420
    },
    {
      "epoch": 0.0012578156422538302,
      "grad_norm": 22.12462043762207,
      "learning_rate": 2.55e-05,
      "loss": 9.2013,
      "step": 430
    },
    {
      "epoch": 0.0012870671688178727,
      "grad_norm": 24.500715255737305,
      "learning_rate": 2.6099999999999997e-05,
      "loss": 9.1629,
      "step": 440
    },
    {
      "epoch": 0.0013163186953819152,
      "grad_norm": 35.824031829833984,
      "learning_rate": 2.6699999999999995e-05,
      "loss": 9.1877,
      "step": 450
    },
    {
      "epoch": 0.0013455702219459578,
      "grad_norm": 28.570497512817383,
      "learning_rate": 2.7299999999999996e-05,
      "loss": 9.1961,
      "step": 460
    },
    {
      "epoch": 0.0013748217485100003,
      "grad_norm": 19.405773162841797,
      "learning_rate": 2.7899999999999997e-05,
      "loss": 9.1905,
      "step": 470
    },
    {
      "epoch": 0.001404073275074043,
      "grad_norm": 20.267793655395508,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 9.1693,
      "step": 480
    },
    {
      "epoch": 0.0014333248016380855,
      "grad_norm": 18.131561279296875,
      "learning_rate": 2.91e-05,
      "loss": 9.1895,
      "step": 490
    },
    {
      "epoch": 0.001462576328202128,
      "grad_norm": 28.323633193969727,
      "learning_rate": 2.97e-05,
      "loss": 9.1714,
      "step": 500
    },
    {
      "epoch": 0.0014918278547661706,
      "grad_norm": 20.251134872436523,
      "learning_rate": 3.0299999999999998e-05,
      "loss": 9.1529,
      "step": 510
    },
    {
      "epoch": 0.0015210793813302131,
      "grad_norm": 17.09373664855957,
      "learning_rate": 3.0839999999999996e-05,
      "loss": 9.1546,
      "step": 520
    },
    {
      "epoch": 0.0015503309078942558,
      "grad_norm": 16.57909393310547,
      "learning_rate": 3.144e-05,
      "loss": 9.1673,
      "step": 530
    },
    {
      "epoch": 0.0015795824344582983,
      "grad_norm": 10.633030891418457,
      "learning_rate": 3.204e-05,
      "loss": 9.1716,
      "step": 540
    },
    {
      "epoch": 0.0016088339610223408,
      "grad_norm": 14.826539039611816,
      "learning_rate": 3.263999999999999e-05,
      "loss": 9.1649,
      "step": 550
    },
    {
      "epoch": 0.0016380854875863834,
      "grad_norm": 17.31394386291504,
      "learning_rate": 3.3239999999999993e-05,
      "loss": 9.173,
      "step": 560
    },
    {
      "epoch": 0.001667337014150426,
      "grad_norm": 13.405190467834473,
      "learning_rate": 3.3839999999999994e-05,
      "loss": 9.1724,
      "step": 570
    },
    {
      "epoch": 0.0016965885407144686,
      "grad_norm": 16.752643585205078,
      "learning_rate": 3.4439999999999996e-05,
      "loss": 9.1688,
      "step": 580
    },
    {
      "epoch": 0.001725840067278511,
      "grad_norm": 18.50242042541504,
      "learning_rate": 3.5039999999999997e-05,
      "loss": 9.1638,
      "step": 590
    },
    {
      "epoch": 0.0017550915938425538,
      "grad_norm": 14.997364044189453,
      "learning_rate": 3.564e-05,
      "loss": 9.1565,
      "step": 600
    },
    {
      "epoch": 0.0017843431204065962,
      "grad_norm": 11.956835746765137,
      "learning_rate": 3.624e-05,
      "loss": 9.1598,
      "step": 610
    },
    {
      "epoch": 0.0018135946469706387,
      "grad_norm": 15.849851608276367,
      "learning_rate": 3.684e-05,
      "loss": 9.1541,
      "step": 620
    },
    {
      "epoch": 0.0018428461735346814,
      "grad_norm": 17.481584548950195,
      "learning_rate": 3.7439999999999994e-05,
      "loss": 9.1533,
      "step": 630
    },
    {
      "epoch": 0.0018720977000987239,
      "grad_norm": 15.944308280944824,
      "learning_rate": 3.8039999999999995e-05,
      "loss": 9.1647,
      "step": 640
    },
    {
      "epoch": 0.0019013492266627666,
      "grad_norm": 15.674567222595215,
      "learning_rate": 3.8639999999999996e-05,
      "loss": 9.161,
      "step": 650
    },
    {
      "epoch": 0.001930600753226809,
      "grad_norm": 13.785255432128906,
      "learning_rate": 3.924e-05,
      "loss": 9.1601,
      "step": 660
    },
    {
      "epoch": 0.0019598522797908517,
      "grad_norm": 15.747488975524902,
      "learning_rate": 3.984e-05,
      "loss": 9.128,
      "step": 670
    },
    {
      "epoch": 0.001989103806354894,
      "grad_norm": 11.030411720275879,
      "learning_rate": 4.044e-05,
      "loss": 9.1476,
      "step": 680
    },
    {
      "epoch": 0.0020183553329189367,
      "grad_norm": 15.800025939941406,
      "learning_rate": 4.104e-05,
      "loss": 9.1129,
      "step": 690
    },
    {
      "epoch": 0.002047606859482979,
      "grad_norm": 11.431413650512695,
      "learning_rate": 4.164e-05,
      "loss": 9.1712,
      "step": 700
    },
    {
      "epoch": 0.002076858386047022,
      "grad_norm": 18.95769500732422,
      "learning_rate": 4.224e-05,
      "loss": 9.1428,
      "step": 710
    },
    {
      "epoch": 0.0021061099126110645,
      "grad_norm": 26.438623428344727,
      "learning_rate": 4.2839999999999996e-05,
      "loss": 9.1454,
      "step": 720
    },
    {
      "epoch": 0.002135361439175107,
      "grad_norm": 19.888408660888672,
      "learning_rate": 4.344e-05,
      "loss": 9.1551,
      "step": 730
    },
    {
      "epoch": 0.0021646129657391495,
      "grad_norm": 13.31981372833252,
      "learning_rate": 4.404e-05,
      "loss": 9.1835,
      "step": 740
    },
    {
      "epoch": 0.002193864492303192,
      "grad_norm": 18.036945343017578,
      "learning_rate": 4.463999999999999e-05,
      "loss": 9.1492,
      "step": 750
    },
    {
      "epoch": 0.002223116018867235,
      "grad_norm": 16.109806060791016,
      "learning_rate": 4.5239999999999994e-05,
      "loss": 9.1337,
      "step": 760
    },
    {
      "epoch": 0.0022523675454312773,
      "grad_norm": 13.018072128295898,
      "learning_rate": 4.5839999999999995e-05,
      "loss": 9.1411,
      "step": 770
    },
    {
      "epoch": 0.00228161907199532,
      "grad_norm": 19.20261573791504,
      "learning_rate": 4.6439999999999996e-05,
      "loss": 9.1204,
      "step": 780
    },
    {
      "epoch": 0.0023108705985593623,
      "grad_norm": 28.50835609436035,
      "learning_rate": 4.704e-05,
      "loss": 9.1516,
      "step": 790
    },
    {
      "epoch": 0.0023401221251234047,
      "grad_norm": 12.84471321105957,
      "learning_rate": 4.764e-05,
      "loss": 9.139,
      "step": 800
    },
    {
      "epoch": 0.0023693736516874476,
      "grad_norm": 11.8615083694458,
      "learning_rate": 4.823999999999999e-05,
      "loss": 9.1208,
      "step": 810
    },
    {
      "epoch": 0.00239862517825149,
      "grad_norm": 11.587111473083496,
      "learning_rate": 4.883999999999999e-05,
      "loss": 9.1362,
      "step": 820
    },
    {
      "epoch": 0.0024278767048155326,
      "grad_norm": 13.630415916442871,
      "learning_rate": 4.9439999999999994e-05,
      "loss": 9.1454,
      "step": 830
    },
    {
      "epoch": 0.002457128231379575,
      "grad_norm": 11.879384994506836,
      "learning_rate": 5.0039999999999995e-05,
      "loss": 9.1438,
      "step": 840
    },
    {
      "epoch": 0.0024863797579436175,
      "grad_norm": 16.579822540283203,
      "learning_rate": 5.0639999999999996e-05,
      "loss": 9.1477,
      "step": 850
    },
    {
      "epoch": 0.0025156312845076604,
      "grad_norm": 31.36379623413086,
      "learning_rate": 5.124e-05,
      "loss": 9.1349,
      "step": 860
    },
    {
      "epoch": 0.002544882811071703,
      "grad_norm": 17.268699645996094,
      "learning_rate": 5.184e-05,
      "loss": 9.127,
      "step": 870
    },
    {
      "epoch": 0.0025741343376357454,
      "grad_norm": 13.554245948791504,
      "learning_rate": 5.244e-05,
      "loss": 9.1293,
      "step": 880
    },
    {
      "epoch": 0.002603385864199788,
      "grad_norm": 16.45536231994629,
      "learning_rate": 5.304e-05,
      "loss": 9.1015,
      "step": 890
    },
    {
      "epoch": 0.0026326373907638303,
      "grad_norm": 14.396760940551758,
      "learning_rate": 5.3639999999999995e-05,
      "loss": 9.1368,
      "step": 900
    },
    {
      "epoch": 0.0026618889173278732,
      "grad_norm": 8.299905776977539,
      "learning_rate": 5.423999999999999e-05,
      "loss": 9.1264,
      "step": 910
    },
    {
      "epoch": 0.0026911404438919157,
      "grad_norm": 13.574129104614258,
      "learning_rate": 5.483999999999999e-05,
      "loss": 9.1403,
      "step": 920
    },
    {
      "epoch": 0.002720391970455958,
      "grad_norm": 32.47224044799805,
      "learning_rate": 5.543999999999999e-05,
      "loss": 9.1285,
      "step": 930
    },
    {
      "epoch": 0.0027496434970200006,
      "grad_norm": 13.24270248413086,
      "learning_rate": 5.603999999999999e-05,
      "loss": 9.1363,
      "step": 940
    },
    {
      "epoch": 0.002778895023584043,
      "grad_norm": 12.032825469970703,
      "learning_rate": 5.663999999999999e-05,
      "loss": 9.0825,
      "step": 950
    },
    {
      "epoch": 0.002808146550148086,
      "grad_norm": 20.875722885131836,
      "learning_rate": 5.7239999999999994e-05,
      "loss": 9.1188,
      "step": 960
    },
    {
      "epoch": 0.0028373980767121285,
      "grad_norm": 19.95583152770996,
      "learning_rate": 5.7839999999999995e-05,
      "loss": 9.1286,
      "step": 970
    },
    {
      "epoch": 0.002866649603276171,
      "grad_norm": 11.723488807678223,
      "learning_rate": 5.8439999999999996e-05,
      "loss": 9.1152,
      "step": 980
    },
    {
      "epoch": 0.0028959011298402134,
      "grad_norm": 9.709324836730957,
      "learning_rate": 5.904e-05,
      "loss": 9.1629,
      "step": 990
    },
    {
      "epoch": 0.002925152656404256,
      "grad_norm": 13.559165954589844,
      "learning_rate": 5.964e-05,
      "loss": 9.1292,
      "step": 1000
    },
    {
      "epoch": 0.002954404182968299,
      "grad_norm": 13.783267974853516,
      "learning_rate": 6.024e-05,
      "loss": 9.1223,
      "step": 1010
    },
    {
      "epoch": 0.0029836557095323413,
      "grad_norm": 10.38874626159668,
      "learning_rate": 6.084e-05,
      "loss": 9.1176,
      "step": 1020
    },
    {
      "epoch": 0.0030129072360963838,
      "grad_norm": 13.770158767700195,
      "learning_rate": 6.144e-05,
      "loss": 9.1132,
      "step": 1030
    },
    {
      "epoch": 0.0030421587626604262,
      "grad_norm": 10.09971809387207,
      "learning_rate": 6.204e-05,
      "loss": 9.1004,
      "step": 1040
    },
    {
      "epoch": 0.0030714102892244687,
      "grad_norm": 10.593066215515137,
      "learning_rate": 6.264e-05,
      "loss": 9.0929,
      "step": 1050
    },
    {
      "epoch": 0.0031006618157885116,
      "grad_norm": 12.87540340423584,
      "learning_rate": 6.323999999999998e-05,
      "loss": 9.1417,
      "step": 1060
    },
    {
      "epoch": 0.003129913342352554,
      "grad_norm": 15.385080337524414,
      "learning_rate": 6.383999999999999e-05,
      "loss": 9.1001,
      "step": 1070
    },
    {
      "epoch": 0.0031591648689165966,
      "grad_norm": 10.08580493927002,
      "learning_rate": 6.443999999999999e-05,
      "loss": 9.1269,
      "step": 1080
    },
    {
      "epoch": 0.003188416395480639,
      "grad_norm": 9.87668514251709,
      "learning_rate": 6.503999999999999e-05,
      "loss": 9.1051,
      "step": 1090
    },
    {
      "epoch": 0.0032176679220446815,
      "grad_norm": 8.900529861450195,
      "learning_rate": 6.563999999999999e-05,
      "loss": 9.076,
      "step": 1100
    },
    {
      "epoch": 0.0032469194486087244,
      "grad_norm": 11.309889793395996,
      "learning_rate": 6.623999999999999e-05,
      "loss": 9.1034,
      "step": 1110
    },
    {
      "epoch": 0.003276170975172767,
      "grad_norm": 21.489351272583008,
      "learning_rate": 6.683999999999999e-05,
      "loss": 9.1436,
      "step": 1120
    },
    {
      "epoch": 0.0033054225017368094,
      "grad_norm": 9.117624282836914,
      "learning_rate": 6.743999999999999e-05,
      "loss": 9.0917,
      "step": 1130
    },
    {
      "epoch": 0.003334674028300852,
      "grad_norm": 10.040019035339355,
      "learning_rate": 6.803999999999999e-05,
      "loss": 9.0955,
      "step": 1140
    },
    {
      "epoch": 0.0033639255548648943,
      "grad_norm": 8.548238754272461,
      "learning_rate": 6.864e-05,
      "loss": 9.1113,
      "step": 1150
    },
    {
      "epoch": 0.003393177081428937,
      "grad_norm": 10.410948753356934,
      "learning_rate": 6.924e-05,
      "loss": 9.0923,
      "step": 1160
    },
    {
      "epoch": 0.0034224286079929797,
      "grad_norm": 7.203856945037842,
      "learning_rate": 6.984e-05,
      "loss": 9.0896,
      "step": 1170
    },
    {
      "epoch": 0.003451680134557022,
      "grad_norm": 9.361257553100586,
      "learning_rate": 7.044e-05,
      "loss": 9.1132,
      "step": 1180
    },
    {
      "epoch": 0.0034809316611210646,
      "grad_norm": 9.721388816833496,
      "learning_rate": 7.104e-05,
      "loss": 9.1222,
      "step": 1190
    },
    {
      "epoch": 0.0035101831876851075,
      "grad_norm": 8.46793270111084,
      "learning_rate": 7.164e-05,
      "loss": 9.0961,
      "step": 1200
    },
    {
      "epoch": 2.925152656404256e-05,
      "grad_norm": 9.392881393432617,
      "learning_rate": 7.223999999999999e-05,
      "loss": 9.1174,
      "step": 1210
    },
    {
      "epoch": 5.850305312808512e-05,
      "grad_norm": 9.305628776550293,
      "learning_rate": 7.283999999999999e-05,
      "loss": 9.0925,
      "step": 1220
    },
    {
      "epoch": 8.775457969212769e-05,
      "grad_norm": 11.655509948730469,
      "learning_rate": 7.343999999999999e-05,
      "loss": 9.0674,
      "step": 1230
    },
    {
      "epoch": 0.00011700610625617024,
      "grad_norm": 9.204011917114258,
      "learning_rate": 7.403999999999999e-05,
      "loss": 9.1135,
      "step": 1240
    },
    {
      "epoch": 0.0001462576328202128,
      "grad_norm": 10.369616508483887,
      "learning_rate": 7.463999999999999e-05,
      "loss": 9.0955,
      "step": 1250
    },
    {
      "epoch": 0.00017550915938425538,
      "grad_norm": 11.895001411437988,
      "learning_rate": 7.524e-05,
      "loss": 9.1122,
      "step": 1260
    },
    {
      "epoch": 0.00020476068594829793,
      "grad_norm": 9.267058372497559,
      "learning_rate": 7.584e-05,
      "loss": 9.109,
      "step": 1270
    },
    {
      "epoch": 0.00023401221251234048,
      "grad_norm": 17.40606689453125,
      "learning_rate": 7.644e-05,
      "loss": 9.1109,
      "step": 1280
    },
    {
      "epoch": 0.00026326373907638306,
      "grad_norm": 9.98201847076416,
      "learning_rate": 7.703999999999998e-05,
      "loss": 9.0955,
      "step": 1290
    },
    {
      "epoch": 0.0002925152656404256,
      "grad_norm": 18.65011215209961,
      "learning_rate": 7.763999999999998e-05,
      "loss": 9.1246,
      "step": 1300
    },
    {
      "epoch": 0.00032176679220446817,
      "grad_norm": 7.038486480712891,
      "learning_rate": 7.823999999999998e-05,
      "loss": 9.1277,
      "step": 1310
    },
    {
      "epoch": 0.00035101831876851075,
      "grad_norm": 7.008432865142822,
      "learning_rate": 7.883999999999998e-05,
      "loss": 9.106,
      "step": 1320
    },
    {
      "epoch": 0.0003802698453325533,
      "grad_norm": 11.14470100402832,
      "learning_rate": 7.943999999999998e-05,
      "loss": 9.156,
      "step": 1330
    },
    {
      "epoch": 0.00040952137189659586,
      "grad_norm": 9.248774528503418,
      "learning_rate": 8.003999999999999e-05,
      "loss": 9.119,
      "step": 1340
    },
    {
      "epoch": 0.00043877289846063844,
      "grad_norm": 10.61188793182373,
      "learning_rate": 8.063999999999999e-05,
      "loss": 9.1022,
      "step": 1350
    },
    {
      "epoch": 0.00046802442502468097,
      "grad_norm": 19.044048309326172,
      "learning_rate": 8.123999999999999e-05,
      "loss": 9.0931,
      "step": 1360
    },
    {
      "epoch": 0.0004972759515887235,
      "grad_norm": 7.324682235717773,
      "learning_rate": 8.183999999999999e-05,
      "loss": 9.0959,
      "step": 1370
    },
    {
      "epoch": 0.0005265274781527661,
      "grad_norm": 13.514555931091309,
      "learning_rate": 8.243999999999999e-05,
      "loss": 9.0835,
      "step": 1380
    },
    {
      "epoch": 0.0005557790047168087,
      "grad_norm": 11.843216896057129,
      "learning_rate": 8.303999999999999e-05,
      "loss": 9.1067,
      "step": 1390
    },
    {
      "epoch": 0.0005850305312808512,
      "grad_norm": 10.607686042785645,
      "learning_rate": 8.363999999999999e-05,
      "loss": 9.1045,
      "step": 1400
    },
    {
      "epoch": 0.0006142820578448938,
      "grad_norm": 9.441575050354004,
      "learning_rate": 8.423999999999999e-05,
      "loss": 9.1112,
      "step": 1410
    },
    {
      "epoch": 0.0006435335844089363,
      "grad_norm": 9.14914608001709,
      "learning_rate": 8.484e-05,
      "loss": 9.0828,
      "step": 1420
    },
    {
      "epoch": 0.0006727851109729789,
      "grad_norm": 8.015144348144531,
      "learning_rate": 8.544e-05,
      "loss": 9.075,
      "step": 1430
    },
    {
      "epoch": 0.0007020366375370215,
      "grad_norm": 8.468117713928223,
      "learning_rate": 8.604e-05,
      "loss": 9.1258,
      "step": 1440
    },
    {
      "epoch": 0.000731288164101064,
      "grad_norm": 6.242001533508301,
      "learning_rate": 8.664e-05,
      "loss": 9.0907,
      "step": 1450
    },
    {
      "epoch": 0.0007605396906651066,
      "grad_norm": 9.129263877868652,
      "learning_rate": 8.724e-05,
      "loss": 9.0592,
      "step": 1460
    },
    {
      "epoch": 0.0007897912172291491,
      "grad_norm": 5.5744404792785645,
      "learning_rate": 8.784e-05,
      "loss": 9.0972,
      "step": 1470
    },
    {
      "epoch": 0.0008190427437931917,
      "grad_norm": 6.7295451164245605,
      "learning_rate": 8.844e-05,
      "loss": 9.0438,
      "step": 1480
    },
    {
      "epoch": 0.0008482942703572343,
      "grad_norm": 9.726662635803223,
      "learning_rate": 8.904e-05,
      "loss": 9.0978,
      "step": 1490
    },
    {
      "epoch": 0.0008775457969212769,
      "grad_norm": 12.181228637695312,
      "learning_rate": 8.963999999999999e-05,
      "loss": 9.0741,
      "step": 1500
    },
    {
      "epoch": 0.0009067973234853194,
      "grad_norm": 9.468052864074707,
      "learning_rate": 9.023999999999999e-05,
      "loss": 9.1054,
      "step": 1510
    },
    {
      "epoch": 0.0009360488500493619,
      "grad_norm": 5.643945217132568,
      "learning_rate": 9.083999999999999e-05,
      "loss": 9.0783,
      "step": 1520
    },
    {
      "epoch": 0.0009653003766134045,
      "grad_norm": 8.345160484313965,
      "learning_rate": 9.143999999999999e-05,
      "loss": 9.0883,
      "step": 1530
    },
    {
      "epoch": 0.000994551903177447,
      "grad_norm": 7.018232345581055,
      "learning_rate": 9.203999999999999e-05,
      "loss": 9.0856,
      "step": 1540
    },
    {
      "epoch": 0.0010238034297414896,
      "grad_norm": 9.320673942565918,
      "learning_rate": 9.264e-05,
      "loss": 9.0597,
      "step": 1550
    },
    {
      "epoch": 0.0010530549563055323,
      "grad_norm": 8.91433048248291,
      "learning_rate": 9.324e-05,
      "loss": 9.0651,
      "step": 1560
    },
    {
      "epoch": 0.0010823064828695747,
      "grad_norm": 6.868046283721924,
      "learning_rate": 9.384e-05,
      "loss": 9.099,
      "step": 1570
    },
    {
      "epoch": 0.0011115580094336174,
      "grad_norm": 9.723581314086914,
      "learning_rate": 9.444e-05,
      "loss": 9.0521,
      "step": 1580
    },
    {
      "epoch": 0.00114080953599766,
      "grad_norm": 10.664285659790039,
      "learning_rate": 9.504e-05,
      "loss": 9.0714,
      "step": 1590
    },
    {
      "epoch": 0.0011700610625617024,
      "grad_norm": 6.757033348083496,
      "learning_rate": 9.563999999999999e-05,
      "loss": 9.0717,
      "step": 1600
    },
    {
      "epoch": 0.001199312589125745,
      "grad_norm": 31.045188903808594,
      "learning_rate": 9.623999999999999e-05,
      "loss": 9.0689,
      "step": 1610
    },
    {
      "epoch": 0.0012285641156897875,
      "grad_norm": 3.5348541736602783,
      "learning_rate": 9.683999999999999e-05,
      "loss": 9.1034,
      "step": 1620
    },
    {
      "epoch": 0.0012578156422538302,
      "grad_norm": 11.082318305969238,
      "learning_rate": 9.743999999999999e-05,
      "loss": 9.0926,
      "step": 1630
    },
    {
      "epoch": 0.0012870671688178727,
      "grad_norm": 7.686779975891113,
      "learning_rate": 9.803999999999999e-05,
      "loss": 9.0571,
      "step": 1640
    },
    {
      "epoch": 0.0013163186953819152,
      "grad_norm": 5.0025763511657715,
      "learning_rate": 9.863999999999999e-05,
      "loss": 9.0827,
      "step": 1650
    },
    {
      "epoch": 0.0013455702219459578,
      "grad_norm": 13.137277603149414,
      "learning_rate": 9.923999999999999e-05,
      "loss": 9.0888,
      "step": 1660
    },
    {
      "epoch": 0.0013748217485100003,
      "grad_norm": 7.32058572769165,
      "learning_rate": 9.983999999999999e-05,
      "loss": 9.0778,
      "step": 1670
    },
    {
      "epoch": 0.001404073275074043,
      "grad_norm": 8.926586151123047,
      "learning_rate": 0.00010044,
      "loss": 9.0649,
      "step": 1680
    },
    {
      "epoch": 0.0014333248016380855,
      "grad_norm": 10.003230094909668,
      "learning_rate": 0.00010103999999999998,
      "loss": 9.0763,
      "step": 1690
    },
    {
      "epoch": 0.001462576328202128,
      "grad_norm": 4.709443092346191,
      "learning_rate": 0.00010163999999999998,
      "loss": 9.0613,
      "step": 1700
    },
    {
      "epoch": 0.0014918278547661706,
      "grad_norm": 9.15395450592041,
      "learning_rate": 0.00010223999999999998,
      "loss": 9.046,
      "step": 1710
    },
    {
      "epoch": 0.0015210793813302131,
      "grad_norm": 6.152580738067627,
      "learning_rate": 0.00010283999999999998,
      "loss": 9.0487,
      "step": 1720
    },
    {
      "epoch": 0.0015503309078942558,
      "grad_norm": 11.63197135925293,
      "learning_rate": 0.00010343999999999999,
      "loss": 9.0683,
      "step": 1730
    },
    {
      "epoch": 0.0015795824344582983,
      "grad_norm": 10.146611213684082,
      "learning_rate": 0.00010403999999999999,
      "loss": 9.0738,
      "step": 1740
    },
    {
      "epoch": 0.0016088339610223408,
      "grad_norm": 6.11122989654541,
      "learning_rate": 0.00010463999999999999,
      "loss": 9.0757,
      "step": 1750
    },
    {
      "epoch": 0.0016380854875863834,
      "grad_norm": 6.611430644989014,
      "learning_rate": 0.00010523999999999999,
      "loss": 9.0853,
      "step": 1760
    },
    {
      "epoch": 0.001667337014150426,
      "grad_norm": 6.685217380523682,
      "learning_rate": 0.00010583999999999999,
      "loss": 9.0756,
      "step": 1770
    },
    {
      "epoch": 0.0016965885407144686,
      "grad_norm": 8.16845417022705,
      "learning_rate": 0.00010643999999999999,
      "loss": 9.0701,
      "step": 1780
    },
    {
      "epoch": 0.001725840067278511,
      "grad_norm": 6.97514533996582,
      "learning_rate": 0.00010703999999999999,
      "loss": 9.0742,
      "step": 1790
    },
    {
      "epoch": 0.0017550915938425538,
      "grad_norm": 7.153378963470459,
      "learning_rate": 0.00010763999999999999,
      "loss": 9.0725,
      "step": 1800
    },
    {
      "epoch": 0.0017843431204065962,
      "grad_norm": 5.415036201477051,
      "learning_rate": 0.00010824,
      "loss": 9.0684,
      "step": 1810
    },
    {
      "epoch": 0.0018135946469706387,
      "grad_norm": 5.254737377166748,
      "learning_rate": 0.00010884,
      "loss": 9.058,
      "step": 1820
    },
    {
      "epoch": 0.0018428461735346814,
      "grad_norm": 12.102179527282715,
      "learning_rate": 0.00010944,
      "loss": 9.0707,
      "step": 1830
    },
    {
      "epoch": 0.0018720977000987239,
      "grad_norm": 6.1733174324035645,
      "learning_rate": 0.00011004,
      "loss": 9.0647,
      "step": 1840
    },
    {
      "epoch": 0.0019013492266627666,
      "grad_norm": 7.331373691558838,
      "learning_rate": 0.00011064,
      "loss": 9.0775,
      "step": 1850
    },
    {
      "epoch": 0.001930600753226809,
      "grad_norm": 12.588440895080566,
      "learning_rate": 0.00011124,
      "loss": 9.0727,
      "step": 1860
    },
    {
      "epoch": 0.0019598522797908517,
      "grad_norm": 7.4883036613464355,
      "learning_rate": 0.00011184,
      "loss": 9.0397,
      "step": 1870
    },
    {
      "epoch": 0.001989103806354894,
      "grad_norm": 4.863422870635986,
      "learning_rate": 0.00011244,
      "loss": 9.0588,
      "step": 1880
    },
    {
      "epoch": 0.0020183553329189367,
      "grad_norm": 10.97378158569336,
      "learning_rate": 0.00011304,
      "loss": 9.0248,
      "step": 1890
    },
    {
      "epoch": 0.002047606859482979,
      "grad_norm": 10.34542179107666,
      "learning_rate": 0.00011364,
      "loss": 9.0844,
      "step": 1900
    },
    {
      "epoch": 0.002076858386047022,
      "grad_norm": 8.598560333251953,
      "learning_rate": 0.00011424,
      "loss": 9.0572,
      "step": 1910
    },
    {
      "epoch": 0.0021061099126110645,
      "grad_norm": 7.107445240020752,
      "learning_rate": 0.00011483999999999998,
      "loss": 9.0532,
      "step": 1920
    },
    {
      "epoch": 0.002135361439175107,
      "grad_norm": 14.190967559814453,
      "learning_rate": 0.00011543999999999998,
      "loss": 9.045,
      "step": 1930
    },
    {
      "epoch": 0.0021646129657391495,
      "grad_norm": 31.4570369720459,
      "learning_rate": 0.00011603999999999998,
      "loss": 9.0697,
      "step": 1940
    },
    {
      "epoch": 0.002193864492303192,
      "grad_norm": 10.80720043182373,
      "learning_rate": 0.00011663999999999998,
      "loss": 9.0477,
      "step": 1950
    },
    {
      "epoch": 0.002223116018867235,
      "grad_norm": 11.1956205368042,
      "learning_rate": 0.00011723999999999998,
      "loss": 9.0341,
      "step": 1960
    },
    {
      "epoch": 0.0022523675454312773,
      "grad_norm": 6.121568202972412,
      "learning_rate": 0.00011783999999999998,
      "loss": 9.0552,
      "step": 1970
    },
    {
      "epoch": 0.00228161907199532,
      "grad_norm": 6.542274475097656,
      "learning_rate": 0.00011843999999999998,
      "loss": 9.0369,
      "step": 1980
    },
    {
      "epoch": 0.0023108705985593623,
      "grad_norm": 5.4809794425964355,
      "learning_rate": 0.00011903999999999999,
      "loss": 9.0542,
      "step": 1990
    },
    {
      "epoch": 0.0023401221251234047,
      "grad_norm": 8.081892013549805,
      "learning_rate": 0.00011963999999999999,
      "loss": 9.0496,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5813095090176000.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}