{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.03804523578534878,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0001902261789267439,
      "grad_norm": 7.031089782714844,
      "learning_rate": 1.9022256039566294e-08,
      "loss": 2.4444,
      "step": 10
    },
    {
      "epoch": 0.0003804523578534878,
      "grad_norm": 13.060206413269043,
      "learning_rate": 3.804451207913259e-08,
      "loss": 2.614,
      "step": 20
    },
    {
      "epoch": 0.0005706785367802317,
      "grad_norm": 15.187254905700684,
      "learning_rate": 5.7066768118698886e-08,
      "loss": 3.1213,
      "step": 30
    },
    {
      "epoch": 0.0007609047157069756,
      "grad_norm": 19.703323364257812,
      "learning_rate": 7.608902415826518e-08,
      "loss": 4.1664,
      "step": 40
    },
    {
      "epoch": 0.0009511308946337195,
      "grad_norm": 32.09828186035156,
      "learning_rate": 9.511128019783147e-08,
      "loss": 5.9187,
      "step": 50
    },
    {
      "epoch": 0.0011413570735604635,
      "grad_norm": 6.527578830718994,
      "learning_rate": 1.1413353623739777e-07,
      "loss": 2.5169,
      "step": 60
    },
    {
      "epoch": 0.0013315832524872073,
      "grad_norm": 8.701460838317871,
      "learning_rate": 1.3315579227696404e-07,
      "loss": 2.6031,
      "step": 70
    },
    {
      "epoch": 0.0015218094314139513,
      "grad_norm": 19.97081184387207,
      "learning_rate": 1.5217804831653035e-07,
      "loss": 3.1127,
      "step": 80
    },
    {
      "epoch": 0.001712035610340695,
      "grad_norm": 14.284954071044922,
      "learning_rate": 1.7120030435609666e-07,
      "loss": 3.9951,
      "step": 90
    },
    {
      "epoch": 0.001902261789267439,
      "grad_norm": 25.341217041015625,
      "learning_rate": 1.9022256039566295e-07,
      "loss": 5.4591,
      "step": 100
    },
    {
      "epoch": 0.002092487968194183,
      "grad_norm": 5.924285411834717,
      "learning_rate": 2.0924481643522923e-07,
      "loss": 2.4404,
      "step": 110
    },
    {
      "epoch": 0.002282714147120927,
      "grad_norm": 11.63301944732666,
      "learning_rate": 2.2826707247479554e-07,
      "loss": 2.5059,
      "step": 120
    },
    {
      "epoch": 0.0024729403260476705,
      "grad_norm": 11.728666305541992,
      "learning_rate": 2.472893285143618e-07,
      "loss": 3.1191,
      "step": 130
    },
    {
      "epoch": 0.0026631665049744145,
      "grad_norm": 9.490508079528809,
      "learning_rate": 2.663115845539281e-07,
      "loss": 3.8237,
      "step": 140
    },
    {
      "epoch": 0.0028533926839011585,
      "grad_norm": 24.791996002197266,
      "learning_rate": 2.853338405934944e-07,
      "loss": 5.5867,
      "step": 150
    },
    {
      "epoch": 0.0030436188628279025,
      "grad_norm": 4.845604419708252,
      "learning_rate": 3.043560966330607e-07,
      "loss": 2.3633,
      "step": 160
    },
    {
      "epoch": 0.003233845041754646,
      "grad_norm": 6.938657283782959,
      "learning_rate": 3.2337835267262704e-07,
      "loss": 2.3907,
      "step": 170
    },
    {
      "epoch": 0.00342407122068139,
      "grad_norm": 8.972224235534668,
      "learning_rate": 3.424006087121933e-07,
      "loss": 2.6722,
      "step": 180
    },
    {
      "epoch": 0.003614297399608134,
      "grad_norm": 9.294358253479004,
      "learning_rate": 3.614228647517596e-07,
      "loss": 3.4523,
      "step": 190
    },
    {
      "epoch": 0.003804523578534878,
      "grad_norm": 19.425222396850586,
      "learning_rate": 3.804451207913259e-07,
      "loss": 5.2828,
      "step": 200
    },
    {
      "epoch": 0.003994749757461622,
      "grad_norm": 3.1245577335357666,
      "learning_rate": 3.994673768308922e-07,
      "loss": 2.191,
      "step": 210
    },
    {
      "epoch": 0.004184975936388366,
      "grad_norm": 4.416078567504883,
      "learning_rate": 4.1848963287045846e-07,
      "loss": 2.4138,
      "step": 220
    },
    {
      "epoch": 0.00437520211531511,
      "grad_norm": 8.128155708312988,
      "learning_rate": 4.3751188891002475e-07,
      "loss": 2.4941,
      "step": 230
    },
    {
      "epoch": 0.004565428294241854,
      "grad_norm": 5.850177764892578,
      "learning_rate": 4.565341449495911e-07,
      "loss": 3.153,
      "step": 240
    },
    {
      "epoch": 0.004755654473168598,
      "grad_norm": 19.345319747924805,
      "learning_rate": 4.7555640098915737e-07,
      "loss": 5.0348,
      "step": 250
    },
    {
      "epoch": 0.004945880652095341,
      "grad_norm": 2.327831983566284,
      "learning_rate": 4.945786570287236e-07,
      "loss": 2.1702,
      "step": 260
    },
    {
      "epoch": 0.005136106831022085,
      "grad_norm": 2.7916853427886963,
      "learning_rate": 5.136009130682899e-07,
      "loss": 2.1161,
      "step": 270
    },
    {
      "epoch": 0.005326333009948829,
      "grad_norm": 3.9358408451080322,
      "learning_rate": 5.326231691078562e-07,
      "loss": 2.2815,
      "step": 280
    },
    {
      "epoch": 0.005516559188875573,
      "grad_norm": 4.891003131866455,
      "learning_rate": 5.516454251474225e-07,
      "loss": 2.8836,
      "step": 290
    },
    {
      "epoch": 0.005706785367802317,
      "grad_norm": 22.499956130981445,
      "learning_rate": 5.706676811869888e-07,
      "loss": 4.7026,
      "step": 300
    },
    {
      "epoch": 0.005897011546729061,
      "grad_norm": 1.9951050281524658,
      "learning_rate": 5.896899372265551e-07,
      "loss": 2.0446,
      "step": 310
    },
    {
      "epoch": 0.006087237725655805,
      "grad_norm": 3.147122859954834,
      "learning_rate": 6.087121932661214e-07,
      "loss": 2.138,
      "step": 320
    },
    {
      "epoch": 0.006277463904582548,
      "grad_norm": 3.4201011657714844,
      "learning_rate": 6.277344493056876e-07,
      "loss": 2.1024,
      "step": 330
    },
    {
      "epoch": 0.006467690083509292,
      "grad_norm": 4.823904991149902,
      "learning_rate": 6.467567053452541e-07,
      "loss": 2.7573,
      "step": 340
    },
    {
      "epoch": 0.006657916262436036,
      "grad_norm": 21.016569137573242,
      "learning_rate": 6.657789613848204e-07,
      "loss": 4.1976,
      "step": 350
    },
    {
      "epoch": 0.00684814244136278,
      "grad_norm": 1.872706413269043,
      "learning_rate": 6.848012174243867e-07,
      "loss": 1.9986,
      "step": 360
    },
    {
      "epoch": 0.007038368620289524,
      "grad_norm": 2.8600945472717285,
      "learning_rate": 7.038234734639529e-07,
      "loss": 1.9909,
      "step": 370
    },
    {
      "epoch": 0.007228594799216268,
      "grad_norm": 3.7441070079803467,
      "learning_rate": 7.228457295035192e-07,
      "loss": 2.0065,
      "step": 380
    },
    {
      "epoch": 0.007418820978143012,
      "grad_norm": 5.4876627922058105,
      "learning_rate": 7.418679855430855e-07,
      "loss": 2.7059,
      "step": 390
    },
    {
      "epoch": 0.007609047157069756,
      "grad_norm": 24.044830322265625,
      "learning_rate": 7.608902415826518e-07,
      "loss": 4.0807,
      "step": 400
    },
    {
      "epoch": 0.0077992733359964995,
      "grad_norm": 1.6859079599380493,
      "learning_rate": 7.799124976222181e-07,
      "loss": 1.974,
      "step": 410
    },
    {
      "epoch": 0.007989499514923244,
      "grad_norm": 2.8750579357147217,
      "learning_rate": 7.989347536617844e-07,
      "loss": 1.835,
      "step": 420
    },
    {
      "epoch": 0.008179725693849988,
      "grad_norm": 4.05927038192749,
      "learning_rate": 8.179570097013506e-07,
      "loss": 1.9796,
      "step": 430
    },
    {
      "epoch": 0.008369951872776732,
      "grad_norm": 6.277343273162842,
      "learning_rate": 8.369792657409169e-07,
      "loss": 2.474,
      "step": 440
    },
    {
      "epoch": 0.008560178051703475,
      "grad_norm": 32.71267318725586,
      "learning_rate": 8.560015217804832e-07,
      "loss": 3.7252,
      "step": 450
    },
    {
      "epoch": 0.00875040423063022,
      "grad_norm": 1.6604670286178589,
      "learning_rate": 8.750237778200495e-07,
      "loss": 1.851,
      "step": 460
    },
    {
      "epoch": 0.008940630409556963,
      "grad_norm": 2.756211280822754,
      "learning_rate": 8.940460338596159e-07,
      "loss": 1.6797,
      "step": 470
    },
    {
      "epoch": 0.009130856588483708,
      "grad_norm": 4.476970195770264,
      "learning_rate": 9.130682898991822e-07,
      "loss": 1.8404,
      "step": 480
    },
    {
      "epoch": 0.00932108276741045,
      "grad_norm": 6.676766395568848,
      "learning_rate": 9.320905459387485e-07,
      "loss": 2.2254,
      "step": 490
    },
    {
      "epoch": 0.009511308946337196,
      "grad_norm": 33.66838836669922,
      "learning_rate": 9.511128019783147e-07,
      "loss": 3.3597,
      "step": 500
    },
    {
      "epoch": 0.009701535125263939,
      "grad_norm": 1.771323561668396,
      "learning_rate": 9.70135058017881e-07,
      "loss": 1.7942,
      "step": 510
    },
    {
      "epoch": 0.009891761304190682,
      "grad_norm": 3.0909101963043213,
      "learning_rate": 9.891573140574472e-07,
      "loss": 1.7758,
      "step": 520
    },
    {
      "epoch": 0.010081987483117427,
      "grad_norm": 4.743190765380859,
      "learning_rate": 1.0081795700970136e-06,
      "loss": 1.7564,
      "step": 530
    },
    {
      "epoch": 0.01027221366204417,
      "grad_norm": 7.06041145324707,
      "learning_rate": 1.0272018261365798e-06,
      "loss": 2.1511,
      "step": 540
    },
    {
      "epoch": 0.010462439840970915,
      "grad_norm": 26.422157287597656,
      "learning_rate": 1.0462240821761462e-06,
      "loss": 2.8421,
      "step": 550
    },
    {
      "epoch": 0.010652666019897658,
      "grad_norm": 1.8234245777130127,
      "learning_rate": 1.0652463382157123e-06,
      "loss": 1.7149,
      "step": 560
    },
    {
      "epoch": 0.010842892198824403,
      "grad_norm": 3.059401273727417,
      "learning_rate": 1.0842685942552787e-06,
      "loss": 1.647,
      "step": 570
    },
    {
      "epoch": 0.011033118377751146,
      "grad_norm": 4.818949222564697,
      "learning_rate": 1.103290850294845e-06,
      "loss": 1.5044,
      "step": 580
    },
    {
      "epoch": 0.01122334455667789,
      "grad_norm": 6.994847774505615,
      "learning_rate": 1.1223131063344113e-06,
      "loss": 1.8919,
      "step": 590
    },
    {
      "epoch": 0.011413570735604634,
      "grad_norm": 38.19530487060547,
      "learning_rate": 1.1413353623739777e-06,
      "loss": 2.4963,
      "step": 600
    },
    {
      "epoch": 0.011603796914531377,
      "grad_norm": 2.0052971839904785,
      "learning_rate": 1.160357618413544e-06,
      "loss": 1.6709,
      "step": 610
    },
    {
      "epoch": 0.011794023093458122,
      "grad_norm": 1.6200284957885742,
      "learning_rate": 1.1793798744531103e-06,
      "loss": 1.5749,
      "step": 620
    },
    {
      "epoch": 0.011984249272384865,
      "grad_norm": 2.2375597953796387,
      "learning_rate": 1.1984021304926766e-06,
      "loss": 1.4411,
      "step": 630
    },
    {
      "epoch": 0.01217447545131161,
      "grad_norm": 5.535067558288574,
      "learning_rate": 1.2174243865322428e-06,
      "loss": 1.845,
      "step": 640
    },
    {
      "epoch": 0.012364701630238353,
      "grad_norm": 9.7356595993042,
      "learning_rate": 1.2364466425718092e-06,
      "loss": 2.524,
      "step": 650
    },
    {
      "epoch": 0.012554927809165096,
      "grad_norm": 1.4515656232833862,
      "learning_rate": 1.2554688986113752e-06,
      "loss": 1.6263,
      "step": 660
    },
    {
      "epoch": 0.012745153988091841,
      "grad_norm": 1.6473021507263184,
      "learning_rate": 1.2744911546509418e-06,
      "loss": 1.5024,
      "step": 670
    },
    {
      "epoch": 0.012935380167018585,
      "grad_norm": 2.3517134189605713,
      "learning_rate": 1.2935134106905082e-06,
      "loss": 1.4407,
      "step": 680
    },
    {
      "epoch": 0.01312560634594533,
      "grad_norm": 2.8996622562408447,
      "learning_rate": 1.3125356667300743e-06,
      "loss": 1.6358,
      "step": 690
    },
    {
      "epoch": 0.013315832524872073,
      "grad_norm": 11.34484577178955,
      "learning_rate": 1.3315579227696407e-06,
      "loss": 2.2334,
      "step": 700
    },
    {
      "epoch": 0.013506058703798817,
      "grad_norm": 1.1996574401855469,
      "learning_rate": 1.350580178809207e-06,
      "loss": 1.4753,
      "step": 710
    },
    {
      "epoch": 0.01369628488272556,
      "grad_norm": 1.6772912740707397,
      "learning_rate": 1.3696024348487733e-06,
      "loss": 1.5256,
      "step": 720
    },
    {
      "epoch": 0.013886511061652304,
      "grad_norm": 2.156625986099243,
      "learning_rate": 1.3886246908883395e-06,
      "loss": 1.4419,
      "step": 730
    },
    {
      "epoch": 0.014076737240579049,
      "grad_norm": 3.710660219192505,
      "learning_rate": 1.4076469469279059e-06,
      "loss": 1.6054,
      "step": 740
    },
    {
      "epoch": 0.014266963419505792,
      "grad_norm": 7.250610828399658,
      "learning_rate": 1.426669202967472e-06,
      "loss": 2.1039,
      "step": 750
    },
    {
      "epoch": 0.014457189598432537,
      "grad_norm": 1.4629435539245605,
      "learning_rate": 1.4456914590070384e-06,
      "loss": 1.5061,
      "step": 760
    },
    {
      "epoch": 0.01464741577735928,
      "grad_norm": 1.7025142908096313,
      "learning_rate": 1.4647137150466046e-06,
      "loss": 1.4471,
      "step": 770
    },
    {
      "epoch": 0.014837641956286025,
      "grad_norm": 2.418815851211548,
      "learning_rate": 1.483735971086171e-06,
      "loss": 1.3384,
      "step": 780
    },
    {
      "epoch": 0.015027868135212768,
      "grad_norm": 4.666192531585693,
      "learning_rate": 1.5027582271257372e-06,
      "loss": 1.599,
      "step": 790
    },
    {
      "epoch": 0.015218094314139513,
      "grad_norm": 8.951377868652344,
      "learning_rate": 1.5217804831653036e-06,
      "loss": 1.9391,
      "step": 800
    },
    {
      "epoch": 0.015408320493066256,
      "grad_norm": 1.383745551109314,
      "learning_rate": 1.54080273920487e-06,
      "loss": 1.5076,
      "step": 810
    },
    {
      "epoch": 0.015598546671992999,
      "grad_norm": 1.4020535945892334,
      "learning_rate": 1.5598249952444361e-06,
      "loss": 1.3977,
      "step": 820
    },
    {
      "epoch": 0.015788772850919744,
      "grad_norm": 2.2491354942321777,
      "learning_rate": 1.5788472512840025e-06,
      "loss": 1.3356,
      "step": 830
    },
    {
      "epoch": 0.015978999029846487,
      "grad_norm": 3.5490710735321045,
      "learning_rate": 1.5978695073235687e-06,
      "loss": 1.5511,
      "step": 840
    },
    {
      "epoch": 0.01616922520877323,
      "grad_norm": 11.946330070495605,
      "learning_rate": 1.616891763363135e-06,
      "loss": 1.8743,
      "step": 850
    },
    {
      "epoch": 0.016359451387699977,
      "grad_norm": 1.4720561504364014,
      "learning_rate": 1.6359140194027013e-06,
      "loss": 1.4478,
      "step": 860
    },
    {
      "epoch": 0.01654967756662672,
      "grad_norm": 1.6156224012374878,
      "learning_rate": 1.6549362754422677e-06,
      "loss": 1.4235,
      "step": 870
    },
    {
      "epoch": 0.016739903745553463,
      "grad_norm": 2.14105224609375,
      "learning_rate": 1.6739585314818339e-06,
      "loss": 1.3183,
      "step": 880
    },
    {
      "epoch": 0.016930129924480206,
      "grad_norm": 4.404923915863037,
      "learning_rate": 1.6929807875214002e-06,
      "loss": 1.3946,
      "step": 890
    },
    {
      "epoch": 0.01712035610340695,
      "grad_norm": 7.182434558868408,
      "learning_rate": 1.7120030435609664e-06,
      "loss": 1.8036,
      "step": 900
    },
    {
      "epoch": 0.017310582282333696,
      "grad_norm": 1.494486689567566,
      "learning_rate": 1.7310252996005328e-06,
      "loss": 1.4478,
      "step": 910
    },
    {
      "epoch": 0.01750080846126044,
      "grad_norm": 1.6941497325897217,
      "learning_rate": 1.750047555640099e-06,
      "loss": 1.3241,
      "step": 920
    },
    {
      "epoch": 0.017691034640187182,
      "grad_norm": 2.1194875240325928,
      "learning_rate": 1.7690698116796654e-06,
      "loss": 1.189,
      "step": 930
    },
    {
      "epoch": 0.017881260819113926,
      "grad_norm": 2.8642826080322266,
      "learning_rate": 1.7880920677192318e-06,
      "loss": 1.3617,
      "step": 940
    },
    {
      "epoch": 0.01807148699804067,
      "grad_norm": 9.907108306884766,
      "learning_rate": 1.807114323758798e-06,
      "loss": 1.7392,
      "step": 950
    },
    {
      "epoch": 0.018261713176967415,
      "grad_norm": 1.627024531364441,
      "learning_rate": 1.8261365797983643e-06,
      "loss": 1.4536,
      "step": 960
    },
    {
      "epoch": 0.01845193935589416,
      "grad_norm": 1.7732610702514648,
      "learning_rate": 1.8451588358379305e-06,
      "loss": 1.3351,
      "step": 970
    },
    {
      "epoch": 0.0186421655348209,
      "grad_norm": 2.412036180496216,
      "learning_rate": 1.864181091877497e-06,
      "loss": 1.2028,
      "step": 980
    },
    {
      "epoch": 0.018832391713747645,
      "grad_norm": 4.754890441894531,
      "learning_rate": 1.883203347917063e-06,
      "loss": 1.3517,
      "step": 990
    },
    {
      "epoch": 0.01902261789267439,
      "grad_norm": 8.927350997924805,
      "learning_rate": 1.9022256039566295e-06,
      "loss": 1.6531,
      "step": 1000
    },
    {
      "epoch": 0.019212844071601134,
      "grad_norm": 1.6348631381988525,
      "learning_rate": 1.9212478599961954e-06,
      "loss": 1.4934,
      "step": 1010
    },
    {
      "epoch": 0.019403070250527878,
      "grad_norm": 1.9770840406417847,
      "learning_rate": 1.940270116035762e-06,
      "loss": 1.3571,
      "step": 1020
    },
    {
      "epoch": 0.01959329642945462,
      "grad_norm": 2.404418468475342,
      "learning_rate": 1.9592923720753282e-06,
      "loss": 1.1291,
      "step": 1030
    },
    {
      "epoch": 0.019783522608381364,
      "grad_norm": 3.4404196739196777,
      "learning_rate": 1.9783146281148944e-06,
      "loss": 1.2498,
      "step": 1040
    },
    {
      "epoch": 0.01997374878730811,
      "grad_norm": 9.02733325958252,
      "learning_rate": 1.9973368841544606e-06,
      "loss": 1.5646,
      "step": 1050
    },
    {
      "epoch": 0.020163974966234854,
      "grad_norm": 1.9403765201568604,
      "learning_rate": 2.016359140194027e-06,
      "loss": 1.3972,
      "step": 1060
    },
    {
      "epoch": 0.020354201145161597,
      "grad_norm": 2.009138584136963,
      "learning_rate": 2.0353813962335938e-06,
      "loss": 1.303,
      "step": 1070
    },
    {
      "epoch": 0.02054442732408834,
      "grad_norm": 2.1750669479370117,
      "learning_rate": 2.0544036522731595e-06,
      "loss": 1.1875,
      "step": 1080
    },
    {
      "epoch": 0.020734653503015087,
      "grad_norm": 3.3346753120422363,
      "learning_rate": 2.073425908312726e-06,
      "loss": 1.262,
      "step": 1090
    },
    {
      "epoch": 0.02092487968194183,
      "grad_norm": 8.939334869384766,
      "learning_rate": 2.0924481643522923e-06,
      "loss": 1.547,
      "step": 1100
    },
    {
      "epoch": 0.021115105860868573,
      "grad_norm": 1.614304780960083,
      "learning_rate": 2.111470420391859e-06,
      "loss": 1.3818,
      "step": 1110
    },
    {
      "epoch": 0.021305332039795316,
      "grad_norm": 1.98068106174469,
      "learning_rate": 2.1304926764314247e-06,
      "loss": 1.2253,
      "step": 1120
    },
    {
      "epoch": 0.02149555821872206,
      "grad_norm": 2.3504743576049805,
      "learning_rate": 2.1495149324709913e-06,
      "loss": 1.1185,
      "step": 1130
    },
    {
      "epoch": 0.021685784397648806,
      "grad_norm": 3.188448667526245,
      "learning_rate": 2.1685371885105575e-06,
      "loss": 1.2133,
      "step": 1140
    },
    {
      "epoch": 0.02187601057657555,
      "grad_norm": 11.752523422241211,
      "learning_rate": 2.187559444550124e-06,
      "loss": 1.5604,
      "step": 1150
    },
    {
      "epoch": 0.022066236755502292,
      "grad_norm": 1.4586597681045532,
      "learning_rate": 2.20658170058969e-06,
      "loss": 1.2922,
      "step": 1160
    },
    {
      "epoch": 0.022256462934429035,
      "grad_norm": 1.83343505859375,
      "learning_rate": 2.2256039566292564e-06,
      "loss": 1.2738,
      "step": 1170
    },
    {
      "epoch": 0.02244668911335578,
      "grad_norm": 2.553067684173584,
      "learning_rate": 2.2446262126688226e-06,
      "loss": 1.096,
      "step": 1180
    },
    {
      "epoch": 0.022636915292282525,
      "grad_norm": 4.024831771850586,
      "learning_rate": 2.263648468708389e-06,
      "loss": 1.1661,
      "step": 1190
    },
    {
      "epoch": 0.022827141471209268,
      "grad_norm": 11.252798080444336,
      "learning_rate": 2.2826707247479554e-06,
      "loss": 1.4848,
      "step": 1200
    },
    {
      "epoch": 0.02301736765013601,
      "grad_norm": 2.0187082290649414,
      "learning_rate": 2.3016929807875215e-06,
      "loss": 1.2914,
      "step": 1210
    },
    {
      "epoch": 0.023207593829062755,
      "grad_norm": 2.2791354656219482,
      "learning_rate": 2.320715236827088e-06,
      "loss": 1.2912,
      "step": 1220
    },
    {
      "epoch": 0.0233978200079895,
      "grad_norm": 2.755721092224121,
      "learning_rate": 2.3397374928666543e-06,
      "loss": 1.1446,
      "step": 1230
    },
    {
      "epoch": 0.023588046186916244,
      "grad_norm": 4.635268688201904,
      "learning_rate": 2.3587597489062205e-06,
      "loss": 1.156,
      "step": 1240
    },
    {
      "epoch": 0.023778272365842987,
      "grad_norm": 11.180182456970215,
      "learning_rate": 2.3777820049457867e-06,
      "loss": 1.5128,
      "step": 1250
    },
    {
      "epoch": 0.02396849854476973,
      "grad_norm": 2.6687514781951904,
      "learning_rate": 2.3968042609853533e-06,
      "loss": 1.3078,
      "step": 1260
    },
    {
      "epoch": 0.024158724723696474,
      "grad_norm": 2.1617326736450195,
      "learning_rate": 2.4158265170249195e-06,
      "loss": 1.187,
      "step": 1270
    },
    {
      "epoch": 0.02434895090262322,
      "grad_norm": 2.7917706966400146,
      "learning_rate": 2.4348487730644856e-06,
      "loss": 1.1008,
      "step": 1280
    },
    {
      "epoch": 0.024539177081549964,
      "grad_norm": 3.9902455806732178,
      "learning_rate": 2.453871029104052e-06,
      "loss": 1.1694,
      "step": 1290
    },
    {
      "epoch": 0.024729403260476707,
      "grad_norm": 10.19775104522705,
      "learning_rate": 2.4728932851436184e-06,
      "loss": 1.4882,
      "step": 1300
    },
    {
      "epoch": 0.02491962943940345,
      "grad_norm": 1.815943956375122,
      "learning_rate": 2.4919155411831846e-06,
      "loss": 1.3296,
      "step": 1310
    },
    {
      "epoch": 0.025109855618330193,
      "grad_norm": 1.7802958488464355,
      "learning_rate": 2.5109377972227504e-06,
      "loss": 1.1588,
      "step": 1320
    },
    {
      "epoch": 0.02530008179725694,
      "grad_norm": 2.8108973503112793,
      "learning_rate": 2.529960053262317e-06,
      "loss": 1.0912,
      "step": 1330
    },
    {
      "epoch": 0.025490307976183683,
      "grad_norm": 3.822803497314453,
      "learning_rate": 2.5489823093018836e-06,
      "loss": 1.1713,
      "step": 1340
    },
    {
      "epoch": 0.025680534155110426,
      "grad_norm": 10.885619163513184,
      "learning_rate": 2.5680045653414497e-06,
      "loss": 1.4048,
      "step": 1350
    },
    {
      "epoch": 0.02587076033403717,
      "grad_norm": 1.5321264266967773,
      "learning_rate": 2.5870268213810163e-06,
      "loss": 1.2555,
      "step": 1360
    },
    {
      "epoch": 0.026060986512963916,
      "grad_norm": 2.115370512008667,
      "learning_rate": 2.606049077420582e-06,
      "loss": 1.1987,
      "step": 1370
    },
    {
      "epoch": 0.02625121269189066,
      "grad_norm": 2.6233675479888916,
      "learning_rate": 2.6250713334601487e-06,
      "loss": 1.0758,
      "step": 1380
    },
    {
      "epoch": 0.026441438870817402,
      "grad_norm": 3.8407323360443115,
      "learning_rate": 2.644093589499715e-06,
      "loss": 1.1887,
      "step": 1390
    },
    {
      "epoch": 0.026631665049744145,
      "grad_norm": 12.603764533996582,
      "learning_rate": 2.6631158455392815e-06,
      "loss": 1.4082,
      "step": 1400
    },
    {
      "epoch": 0.026821891228670888,
      "grad_norm": 1.6481189727783203,
      "learning_rate": 2.6821381015788472e-06,
      "loss": 1.2527,
      "step": 1410
    },
    {
      "epoch": 0.027012117407597635,
      "grad_norm": 2.1282284259796143,
      "learning_rate": 2.701160357618414e-06,
      "loss": 1.2139,
      "step": 1420
    },
    {
      "epoch": 0.027202343586524378,
      "grad_norm": 2.9524528980255127,
      "learning_rate": 2.72018261365798e-06,
      "loss": 1.09,
      "step": 1430
    },
    {
      "epoch": 0.02739256976545112,
      "grad_norm": 4.275539398193359,
      "learning_rate": 2.7392048696975466e-06,
      "loss": 1.1149,
      "step": 1440
    },
    {
      "epoch": 0.027582795944377864,
      "grad_norm": 12.224852561950684,
      "learning_rate": 2.7582271257371124e-06,
      "loss": 1.3974,
      "step": 1450
    },
    {
      "epoch": 0.027773022123304607,
      "grad_norm": 2.3314363956451416,
      "learning_rate": 2.777249381776679e-06,
      "loss": 1.3061,
      "step": 1460
    },
    {
      "epoch": 0.027963248302231354,
      "grad_norm": 2.170639991760254,
      "learning_rate": 2.796271637816245e-06,
      "loss": 1.2254,
      "step": 1470
    },
    {
      "epoch": 0.028153474481158097,
      "grad_norm": 2.8608880043029785,
      "learning_rate": 2.8152938938558117e-06,
      "loss": 1.0807,
      "step": 1480
    },
    {
      "epoch": 0.02834370066008484,
      "grad_norm": 3.440014600753784,
      "learning_rate": 2.834316149895378e-06,
      "loss": 1.1255,
      "step": 1490
    },
    {
      "epoch": 0.028533926839011584,
      "grad_norm": 10.175577163696289,
      "learning_rate": 2.853338405934944e-06,
      "loss": 1.3387,
      "step": 1500
    },
    {
      "epoch": 0.02872415301793833,
      "grad_norm": 2.22709321975708,
      "learning_rate": 2.8723606619745103e-06,
      "loss": 1.2654,
      "step": 1510
    },
    {
      "epoch": 0.028914379196865073,
      "grad_norm": 1.8235417604446411,
      "learning_rate": 2.891382918014077e-06,
      "loss": 1.2287,
      "step": 1520
    },
    {
      "epoch": 0.029104605375791816,
      "grad_norm": 2.5544168949127197,
      "learning_rate": 2.910405174053643e-06,
      "loss": 0.9776,
      "step": 1530
    },
    {
      "epoch": 0.02929483155471856,
      "grad_norm": 4.163671016693115,
      "learning_rate": 2.9294274300932092e-06,
      "loss": 1.0837,
      "step": 1540
    },
    {
      "epoch": 0.029485057733645303,
      "grad_norm": 10.698822021484375,
      "learning_rate": 2.9484496861327754e-06,
      "loss": 1.3228,
      "step": 1550
    },
    {
      "epoch": 0.02967528391257205,
      "grad_norm": 1.6194937229156494,
      "learning_rate": 2.967471942172342e-06,
      "loss": 1.2534,
      "step": 1560
    },
    {
      "epoch": 0.029865510091498793,
      "grad_norm": 1.9834601879119873,
      "learning_rate": 2.986494198211908e-06,
      "loss": 1.201,
      "step": 1570
    },
    {
      "epoch": 0.030055736270425536,
      "grad_norm": 2.591292142868042,
      "learning_rate": 3.0055164542514744e-06,
      "loss": 1.0589,
      "step": 1580
    },
    {
      "epoch": 0.03024596244935228,
      "grad_norm": 4.2067670822143555,
      "learning_rate": 3.0245387102910406e-06,
      "loss": 1.0603,
      "step": 1590
    },
    {
      "epoch": 0.030436188628279025,
      "grad_norm": 9.938916206359863,
      "learning_rate": 3.043560966330607e-06,
      "loss": 1.2619,
      "step": 1600
    },
    {
      "epoch": 0.03062641480720577,
      "grad_norm": 2.7832484245300293,
      "learning_rate": 3.0625832223701733e-06,
      "loss": 1.2533,
      "step": 1610
    },
    {
      "epoch": 0.030816640986132512,
      "grad_norm": 2.117129325866699,
      "learning_rate": 3.08160547840974e-06,
      "loss": 1.1825,
      "step": 1620
    },
    {
      "epoch": 0.031006867165059255,
      "grad_norm": 3.0919687747955322,
      "learning_rate": 3.1006277344493057e-06,
      "loss": 1.095,
      "step": 1630
    },
    {
      "epoch": 0.031197093343985998,
      "grad_norm": 4.166520595550537,
      "learning_rate": 3.1196499904888723e-06,
      "loss": 1.047,
      "step": 1640
    },
    {
      "epoch": 0.031387319522912745,
      "grad_norm": 12.0921630859375,
      "learning_rate": 3.1386722465284385e-06,
      "loss": 1.3526,
      "step": 1650
    },
    {
      "epoch": 0.03157754570183949,
      "grad_norm": 1.9751317501068115,
      "learning_rate": 3.157694502568005e-06,
      "loss": 1.2864,
      "step": 1660
    },
    {
      "epoch": 0.03176777188076623,
      "grad_norm": 2.2880234718322754,
      "learning_rate": 3.176716758607571e-06,
      "loss": 1.1079,
      "step": 1670
    },
    {
      "epoch": 0.031957998059692974,
      "grad_norm": 2.804194688796997,
      "learning_rate": 3.1957390146471374e-06,
      "loss": 0.9666,
      "step": 1680
    },
    {
      "epoch": 0.03214822423861972,
      "grad_norm": 4.810030937194824,
      "learning_rate": 3.2147612706867036e-06,
      "loss": 1.0979,
      "step": 1690
    },
    {
      "epoch": 0.03233845041754646,
      "grad_norm": 9.718046188354492,
      "learning_rate": 3.23378352672627e-06,
      "loss": 1.2383,
      "step": 1700
    },
    {
      "epoch": 0.032528676596473204,
      "grad_norm": 2.0624125003814697,
      "learning_rate": 3.252805782765836e-06,
      "loss": 1.2589,
      "step": 1710
    },
    {
      "epoch": 0.032718902775399954,
      "grad_norm": 2.8751795291900635,
      "learning_rate": 3.2718280388054026e-06,
      "loss": 1.1319,
      "step": 1720
    },
    {
      "epoch": 0.0329091289543267,
      "grad_norm": 2.805138349533081,
      "learning_rate": 3.2908502948449687e-06,
      "loss": 0.9922,
      "step": 1730
    },
    {
      "epoch": 0.03309935513325344,
      "grad_norm": 4.693927764892578,
      "learning_rate": 3.3098725508845353e-06,
      "loss": 1.0518,
      "step": 1740
    },
    {
      "epoch": 0.03328958131218018,
      "grad_norm": 9.922603607177734,
      "learning_rate": 3.3288948069241015e-06,
      "loss": 1.2417,
      "step": 1750
    },
    {
      "epoch": 0.033479807491106926,
      "grad_norm": 1.9441560506820679,
      "learning_rate": 3.3479170629636677e-06,
      "loss": 1.2156,
      "step": 1760
    },
    {
      "epoch": 0.03367003367003367,
      "grad_norm": 2.1576952934265137,
      "learning_rate": 3.366939319003234e-06,
      "loss": 1.1009,
      "step": 1770
    },
    {
      "epoch": 0.03386025984896041,
      "grad_norm": 2.93638014793396,
      "learning_rate": 3.3859615750428005e-06,
      "loss": 0.9806,
      "step": 1780
    },
    {
      "epoch": 0.034050486027887156,
      "grad_norm": 4.772079944610596,
      "learning_rate": 3.4049838310823667e-06,
      "loss": 1.0505,
      "step": 1790
    },
    {
      "epoch": 0.0342407122068139,
      "grad_norm": 13.75841236114502,
      "learning_rate": 3.424006087121933e-06,
      "loss": 1.2653,
      "step": 1800
    },
    {
      "epoch": 0.03443093838574065,
      "grad_norm": 2.011861801147461,
      "learning_rate": 3.443028343161499e-06,
      "loss": 1.1962,
      "step": 1810
    },
    {
      "epoch": 0.03462116456466739,
      "grad_norm": 2.2049922943115234,
      "learning_rate": 3.4620505992010656e-06,
      "loss": 1.1636,
      "step": 1820
    },
    {
      "epoch": 0.034811390743594135,
      "grad_norm": 3.468641757965088,
      "learning_rate": 3.481072855240632e-06,
      "loss": 1.006,
      "step": 1830
    },
    {
      "epoch": 0.03500161692252088,
      "grad_norm": 5.194962024688721,
      "learning_rate": 3.500095111280198e-06,
      "loss": 1.035,
      "step": 1840
    },
    {
      "epoch": 0.03519184310144762,
      "grad_norm": 9.455286979675293,
      "learning_rate": 3.519117367319764e-06,
      "loss": 1.1639,
      "step": 1850
    },
    {
      "epoch": 0.035382069280374365,
      "grad_norm": 2.2892448902130127,
      "learning_rate": 3.5381396233593308e-06,
      "loss": 1.2648,
      "step": 1860
    },
    {
      "epoch": 0.03557229545930111,
      "grad_norm": 2.668325185775757,
      "learning_rate": 3.557161879398897e-06,
      "loss": 1.1385,
      "step": 1870
    },
    {
      "epoch": 0.03576252163822785,
      "grad_norm": 2.897562265396118,
      "learning_rate": 3.5761841354384635e-06,
      "loss": 0.9518,
      "step": 1880
    },
    {
      "epoch": 0.035952747817154594,
      "grad_norm": 5.355794906616211,
      "learning_rate": 3.5952063914780293e-06,
      "loss": 1.059,
      "step": 1890
    },
    {
      "epoch": 0.03614297399608134,
      "grad_norm": 11.731189727783203,
      "learning_rate": 3.614228647517596e-06,
      "loss": 1.2406,
      "step": 1900
    },
    {
      "epoch": 0.03633320017500809,
      "grad_norm": 2.4913065433502197,
      "learning_rate": 3.633250903557162e-06,
      "loss": 1.1554,
      "step": 1910
    },
    {
      "epoch": 0.03652342635393483,
      "grad_norm": 2.4660778045654297,
      "learning_rate": 3.6522731595967287e-06,
      "loss": 1.1604,
      "step": 1920
    },
    {
      "epoch": 0.036713652532861574,
      "grad_norm": 2.8727564811706543,
      "learning_rate": 3.6712954156362944e-06,
      "loss": 0.986,
      "step": 1930
    },
    {
      "epoch": 0.03690387871178832,
      "grad_norm": 5.380633354187012,
      "learning_rate": 3.690317671675861e-06,
      "loss": 0.9519,
      "step": 1940
    },
    {
      "epoch": 0.03709410489071506,
      "grad_norm": 9.196549415588379,
      "learning_rate": 3.709339927715427e-06,
      "loss": 1.253,
      "step": 1950
    },
    {
      "epoch": 0.0372843310696418,
      "grad_norm": 1.7812504768371582,
      "learning_rate": 3.728362183754994e-06,
      "loss": 1.2104,
      "step": 1960
    },
    {
      "epoch": 0.037474557248568546,
      "grad_norm": 2.1316113471984863,
      "learning_rate": 3.7473844397945596e-06,
      "loss": 1.1207,
      "step": 1970
    },
    {
      "epoch": 0.03766478342749529,
      "grad_norm": 2.813387155532837,
      "learning_rate": 3.766406695834126e-06,
      "loss": 0.9173,
      "step": 1980
    },
    {
      "epoch": 0.03785500960642203,
      "grad_norm": 4.0899128913879395,
      "learning_rate": 3.7854289518736923e-06,
      "loss": 0.9493,
      "step": 1990
    },
    {
      "epoch": 0.03804523578534878,
      "grad_norm": 10.877523422241211,
      "learning_rate": 3.804451207913259e-06,
      "loss": 1.2015,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 105138,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "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": 2.7069734052126044e+18,
  "train_batch_size": 7,
  "trial_name": null,
  "trial_params": null
}