{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 10920,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009157509157509158,
      "grad_norm": 0.05460265651345253,
      "learning_rate": 2.289639381797367e-07,
      "loss": 2.2583,
      "step": 100
    },
    {
      "epoch": 0.018315018315018316,
      "grad_norm": 0.07254950702190399,
      "learning_rate": 4.579278763594734e-07,
      "loss": 2.2556,
      "step": 200
    },
    {
      "epoch": 0.027472527472527472,
      "grad_norm": 0.07023929059505463,
      "learning_rate": 6.868918145392102e-07,
      "loss": 2.2658,
      "step": 300
    },
    {
      "epoch": 0.03663003663003663,
      "grad_norm": 0.08579763770103455,
      "learning_rate": 9.158557527189468e-07,
      "loss": 2.2546,
      "step": 400
    },
    {
      "epoch": 0.045787545787545784,
      "grad_norm": 0.13748684525489807,
      "learning_rate": 1.1448196908986836e-06,
      "loss": 2.2453,
      "step": 500
    },
    {
      "epoch": 0.054945054945054944,
      "grad_norm": 0.0766037255525589,
      "learning_rate": 1.3737836290784203e-06,
      "loss": 2.2471,
      "step": 600
    },
    {
      "epoch": 0.0641025641025641,
      "grad_norm": 0.18016816675662994,
      "learning_rate": 1.6027475672581568e-06,
      "loss": 2.2481,
      "step": 700
    },
    {
      "epoch": 0.07326007326007326,
      "grad_norm": 0.10347414016723633,
      "learning_rate": 1.8317115054378937e-06,
      "loss": 2.2388,
      "step": 800
    },
    {
      "epoch": 0.08241758241758242,
      "grad_norm": 0.14010679721832275,
      "learning_rate": 2.0606754436176306e-06,
      "loss": 2.2189,
      "step": 900
    },
    {
      "epoch": 0.09157509157509157,
      "grad_norm": 0.12083210051059723,
      "learning_rate": 2.2896393817973673e-06,
      "loss": 2.2308,
      "step": 1000
    },
    {
      "epoch": 0.10073260073260074,
      "grad_norm": 0.13420909643173218,
      "learning_rate": 2.5186033199771035e-06,
      "loss": 2.215,
      "step": 1100
    },
    {
      "epoch": 0.10989010989010989,
      "grad_norm": 0.15881897509098053,
      "learning_rate": 2.7475672581568406e-06,
      "loss": 2.2177,
      "step": 1200
    },
    {
      "epoch": 0.11904761904761904,
      "grad_norm": 0.1493181735277176,
      "learning_rate": 2.97424155695478e-06,
      "loss": 2.2062,
      "step": 1300
    },
    {
      "epoch": 0.1282051282051282,
      "grad_norm": 0.17832180857658386,
      "learning_rate": 3.2032054951345167e-06,
      "loss": 2.2018,
      "step": 1400
    },
    {
      "epoch": 0.13736263736263737,
      "grad_norm": 0.16811606287956238,
      "learning_rate": 3.4321694333142534e-06,
      "loss": 2.1941,
      "step": 1500
    },
    {
      "epoch": 0.14652014652014653,
      "grad_norm": 0.16803540289402008,
      "learning_rate": 3.66113337149399e-06,
      "loss": 2.1665,
      "step": 1600
    },
    {
      "epoch": 0.15567765567765568,
      "grad_norm": 0.18348592519760132,
      "learning_rate": 3.890097309673726e-06,
      "loss": 2.1702,
      "step": 1700
    },
    {
      "epoch": 0.16483516483516483,
      "grad_norm": 0.19313548505306244,
      "learning_rate": 4.119061247853463e-06,
      "loss": 2.154,
      "step": 1800
    },
    {
      "epoch": 0.17399267399267399,
      "grad_norm": 0.1911778450012207,
      "learning_rate": 4.3480251860332005e-06,
      "loss": 2.1665,
      "step": 1900
    },
    {
      "epoch": 0.18315018315018314,
      "grad_norm": 0.20000025629997253,
      "learning_rate": 4.574699484831139e-06,
      "loss": 2.1645,
      "step": 2000
    },
    {
      "epoch": 0.19230769230769232,
      "grad_norm": 0.23437415063381195,
      "learning_rate": 4.803663423010876e-06,
      "loss": 2.1638,
      "step": 2100
    },
    {
      "epoch": 0.20146520146520147,
      "grad_norm": 0.23130947351455688,
      "learning_rate": 5.032627361190613e-06,
      "loss": 2.1656,
      "step": 2200
    },
    {
      "epoch": 0.21062271062271062,
      "grad_norm": 0.22546882927417755,
      "learning_rate": 5.26159129937035e-06,
      "loss": 2.1426,
      "step": 2300
    },
    {
      "epoch": 0.21978021978021978,
      "grad_norm": 0.23832601308822632,
      "learning_rate": 5.490555237550086e-06,
      "loss": 2.1404,
      "step": 2400
    },
    {
      "epoch": 0.22893772893772893,
      "grad_norm": 0.2475137710571289,
      "learning_rate": 5.719519175729823e-06,
      "loss": 2.132,
      "step": 2500
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 0.27017173171043396,
      "learning_rate": 5.94848311390956e-06,
      "loss": 2.1334,
      "step": 2600
    },
    {
      "epoch": 0.24725274725274726,
      "grad_norm": 0.31527864933013916,
      "learning_rate": 6.177447052089296e-06,
      "loss": 2.1221,
      "step": 2700
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 0.39517447352409363,
      "learning_rate": 6.406410990269033e-06,
      "loss": 2.1134,
      "step": 2800
    },
    {
      "epoch": 0.26556776556776557,
      "grad_norm": 0.2927546501159668,
      "learning_rate": 6.6353749284487705e-06,
      "loss": 2.1196,
      "step": 2900
    },
    {
      "epoch": 0.27472527472527475,
      "grad_norm": 0.28731632232666016,
      "learning_rate": 6.864338866628507e-06,
      "loss": 2.0895,
      "step": 3000
    },
    {
      "epoch": 0.2838827838827839,
      "grad_norm": 0.33531680703163147,
      "learning_rate": 7.093302804808243e-06,
      "loss": 2.0915,
      "step": 3100
    },
    {
      "epoch": 0.29304029304029305,
      "grad_norm": 0.2969292998313904,
      "learning_rate": 7.32226674298798e-06,
      "loss": 2.0853,
      "step": 3200
    },
    {
      "epoch": 0.3021978021978022,
      "grad_norm": 0.32316532731056213,
      "learning_rate": 7.551230681167717e-06,
      "loss": 2.0859,
      "step": 3300
    },
    {
      "epoch": 0.31135531135531136,
      "grad_norm": 0.3245663344860077,
      "learning_rate": 7.780194619347453e-06,
      "loss": 2.0655,
      "step": 3400
    },
    {
      "epoch": 0.32051282051282054,
      "grad_norm": 0.6936108469963074,
      "learning_rate": 8.00915855752719e-06,
      "loss": 2.06,
      "step": 3500
    },
    {
      "epoch": 0.32967032967032966,
      "grad_norm": 0.36271369457244873,
      "learning_rate": 8.238122495706927e-06,
      "loss": 2.0726,
      "step": 3600
    },
    {
      "epoch": 0.33882783882783885,
      "grad_norm": 0.3658630847930908,
      "learning_rate": 8.467086433886663e-06,
      "loss": 2.0698,
      "step": 3700
    },
    {
      "epoch": 0.34798534798534797,
      "grad_norm": 0.4526204764842987,
      "learning_rate": 8.696050372066401e-06,
      "loss": 2.0497,
      "step": 3800
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.3695688843727112,
      "learning_rate": 8.925014310246137e-06,
      "loss": 2.0638,
      "step": 3900
    },
    {
      "epoch": 0.3663003663003663,
      "grad_norm": 0.3850938379764557,
      "learning_rate": 9.153978248425874e-06,
      "loss": 2.0525,
      "step": 4000
    },
    {
      "epoch": 0.37545787545787546,
      "grad_norm": 0.42660748958587646,
      "learning_rate": 9.38294218660561e-06,
      "loss": 2.06,
      "step": 4100
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.41539886593818665,
      "learning_rate": 9.611906124785348e-06,
      "loss": 2.0474,
      "step": 4200
    },
    {
      "epoch": 0.39377289377289376,
      "grad_norm": 0.40648505091667175,
      "learning_rate": 9.840870062965084e-06,
      "loss": 2.0379,
      "step": 4300
    },
    {
      "epoch": 0.40293040293040294,
      "grad_norm": 0.5012279152870178,
      "learning_rate": 1.006983400114482e-05,
      "loss": 2.043,
      "step": 4400
    },
    {
      "epoch": 0.41208791208791207,
      "grad_norm": 0.38856762647628784,
      "learning_rate": 1.0298797939324557e-05,
      "loss": 2.0299,
      "step": 4500
    },
    {
      "epoch": 0.42124542124542125,
      "grad_norm": 0.4283836781978607,
      "learning_rate": 1.0527761877504295e-05,
      "loss": 2.0429,
      "step": 4600
    },
    {
      "epoch": 0.43040293040293043,
      "grad_norm": 0.5659844279289246,
      "learning_rate": 1.075672581568403e-05,
      "loss": 2.0415,
      "step": 4700
    },
    {
      "epoch": 0.43956043956043955,
      "grad_norm": 0.47702351212501526,
      "learning_rate": 1.0985689753863767e-05,
      "loss": 2.0181,
      "step": 4800
    },
    {
      "epoch": 0.44871794871794873,
      "grad_norm": 0.4613106846809387,
      "learning_rate": 1.1212364052661706e-05,
      "loss": 2.0224,
      "step": 4900
    },
    {
      "epoch": 0.45787545787545786,
      "grad_norm": 0.4302317500114441,
      "learning_rate": 1.1441327990841444e-05,
      "loss": 2.031,
      "step": 5000
    },
    {
      "epoch": 0.46703296703296704,
      "grad_norm": 0.41913115978240967,
      "learning_rate": 1.167029192902118e-05,
      "loss": 2.0182,
      "step": 5100
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 0.5346239805221558,
      "learning_rate": 1.1899255867200917e-05,
      "loss": 2.0158,
      "step": 5200
    },
    {
      "epoch": 0.48534798534798534,
      "grad_norm": 0.4939456284046173,
      "learning_rate": 1.2128219805380655e-05,
      "loss": 2.0215,
      "step": 5300
    },
    {
      "epoch": 0.4945054945054945,
      "grad_norm": 0.42702722549438477,
      "learning_rate": 1.2357183743560391e-05,
      "loss": 2.0081,
      "step": 5400
    },
    {
      "epoch": 0.5036630036630036,
      "grad_norm": 0.5567723512649536,
      "learning_rate": 1.2586147681740126e-05,
      "loss": 2.0116,
      "step": 5500
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.4391794204711914,
      "learning_rate": 1.2815111619919864e-05,
      "loss": 2.0117,
      "step": 5600
    },
    {
      "epoch": 0.521978021978022,
      "grad_norm": 0.6212931275367737,
      "learning_rate": 1.30440755580996e-05,
      "loss": 2.0111,
      "step": 5700
    },
    {
      "epoch": 0.5311355311355311,
      "grad_norm": 0.4400519132614136,
      "learning_rate": 1.3273039496279336e-05,
      "loss": 2.0099,
      "step": 5800
    },
    {
      "epoch": 0.5402930402930403,
      "grad_norm": 0.5068506598472595,
      "learning_rate": 1.3502003434459074e-05,
      "loss": 2.014,
      "step": 5900
    },
    {
      "epoch": 0.5494505494505495,
      "grad_norm": 0.5231282711029053,
      "learning_rate": 1.373096737263881e-05,
      "loss": 2.0055,
      "step": 6000
    },
    {
      "epoch": 0.5586080586080586,
      "grad_norm": 0.5154737234115601,
      "learning_rate": 1.3959931310818547e-05,
      "loss": 2.0046,
      "step": 6100
    },
    {
      "epoch": 0.5677655677655677,
      "grad_norm": 0.4546205699443817,
      "learning_rate": 1.4188895248998285e-05,
      "loss": 1.9846,
      "step": 6200
    },
    {
      "epoch": 0.5769230769230769,
      "grad_norm": 0.5952470302581787,
      "learning_rate": 1.441785918717802e-05,
      "loss": 1.9931,
      "step": 6300
    },
    {
      "epoch": 0.5860805860805861,
      "grad_norm": 0.4483753740787506,
      "learning_rate": 1.4646823125357757e-05,
      "loss": 1.9976,
      "step": 6400
    },
    {
      "epoch": 0.5952380952380952,
      "grad_norm": 0.5263549089431763,
      "learning_rate": 1.4875787063537495e-05,
      "loss": 1.9938,
      "step": 6500
    },
    {
      "epoch": 0.6043956043956044,
      "grad_norm": 0.5647267699241638,
      "learning_rate": 1.5104751001717231e-05,
      "loss": 2.0073,
      "step": 6600
    },
    {
      "epoch": 0.6135531135531136,
      "grad_norm": 0.4960547089576721,
      "learning_rate": 1.5333714939896968e-05,
      "loss": 1.9973,
      "step": 6700
    },
    {
      "epoch": 0.6227106227106227,
      "grad_norm": 0.6195764541625977,
      "learning_rate": 1.5562678878076704e-05,
      "loss": 1.9929,
      "step": 6800
    },
    {
      "epoch": 0.6318681318681318,
      "grad_norm": 0.48598435521125793,
      "learning_rate": 1.579164281625644e-05,
      "loss": 1.9864,
      "step": 6900
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 0.6534644365310669,
      "learning_rate": 1.6020606754436176e-05,
      "loss": 1.9883,
      "step": 7000
    },
    {
      "epoch": 0.6501831501831502,
      "grad_norm": 0.5080538392066956,
      "learning_rate": 1.6249570692615916e-05,
      "loss": 1.9651,
      "step": 7100
    },
    {
      "epoch": 0.6593406593406593,
      "grad_norm": 0.502566397190094,
      "learning_rate": 1.6478534630795652e-05,
      "loss": 1.9877,
      "step": 7200
    },
    {
      "epoch": 0.6684981684981685,
      "grad_norm": 0.5017273426055908,
      "learning_rate": 1.670749856897539e-05,
      "loss": 1.9841,
      "step": 7300
    },
    {
      "epoch": 0.6776556776556777,
      "grad_norm": 0.6656813621520996,
      "learning_rate": 1.6936462507155125e-05,
      "loss": 1.9798,
      "step": 7400
    },
    {
      "epoch": 0.6868131868131868,
      "grad_norm": 0.5949518084526062,
      "learning_rate": 1.716542644533486e-05,
      "loss": 1.9929,
      "step": 7500
    },
    {
      "epoch": 0.6959706959706959,
      "grad_norm": 0.738613486289978,
      "learning_rate": 1.7394390383514597e-05,
      "loss": 1.996,
      "step": 7600
    },
    {
      "epoch": 0.7051282051282052,
      "grad_norm": 0.5886286497116089,
      "learning_rate": 1.7623354321694334e-05,
      "loss": 1.9669,
      "step": 7700
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.552151083946228,
      "learning_rate": 1.785231825987407e-05,
      "loss": 1.9565,
      "step": 7800
    },
    {
      "epoch": 0.7234432234432234,
      "grad_norm": 0.5507295727729797,
      "learning_rate": 1.808128219805381e-05,
      "loss": 1.9857,
      "step": 7900
    },
    {
      "epoch": 0.7326007326007326,
      "grad_norm": 0.7184144258499146,
      "learning_rate": 1.8310246136233546e-05,
      "loss": 1.969,
      "step": 8000
    },
    {
      "epoch": 0.7417582417582418,
      "grad_norm": 0.5713202953338623,
      "learning_rate": 1.8539210074413282e-05,
      "loss": 1.9797,
      "step": 8100
    },
    {
      "epoch": 0.7509157509157509,
      "grad_norm": 0.563890278339386,
      "learning_rate": 1.8768174012593018e-05,
      "loss": 1.9643,
      "step": 8200
    },
    {
      "epoch": 0.76007326007326,
      "grad_norm": 0.6094496846199036,
      "learning_rate": 1.8997137950772755e-05,
      "loss": 1.9758,
      "step": 8300
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.8818446397781372,
      "learning_rate": 1.922610188895249e-05,
      "loss": 1.966,
      "step": 8400
    },
    {
      "epoch": 0.7783882783882784,
      "grad_norm": 0.5603904128074646,
      "learning_rate": 1.945506582713223e-05,
      "loss": 1.97,
      "step": 8500
    },
    {
      "epoch": 0.7875457875457875,
      "grad_norm": 0.6132967472076416,
      "learning_rate": 1.9684029765311963e-05,
      "loss": 1.9606,
      "step": 8600
    },
    {
      "epoch": 0.7967032967032966,
      "grad_norm": 0.5720909237861633,
      "learning_rate": 1.99129937034917e-05,
      "loss": 1.9789,
      "step": 8700
    },
    {
      "epoch": 0.8058608058608059,
      "grad_norm": 0.5442408919334412,
      "learning_rate": 1.9960293406409122e-05,
      "loss": 1.9667,
      "step": 8800
    },
    {
      "epoch": 0.815018315018315,
      "grad_norm": 0.6531534194946289,
      "learning_rate": 1.9729957256223114e-05,
      "loss": 1.9748,
      "step": 8900
    },
    {
      "epoch": 0.8241758241758241,
      "grad_norm": 0.5392922163009644,
      "learning_rate": 1.9298823053295396e-05,
      "loss": 1.9635,
      "step": 9000
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.6137576699256897,
      "learning_rate": 1.8675788155151654e-05,
      "loss": 1.9586,
      "step": 9100
    },
    {
      "epoch": 0.8424908424908425,
      "grad_norm": 0.5812509655952454,
      "learning_rate": 1.787371019220515e-05,
      "loss": 1.9504,
      "step": 9200
    },
    {
      "epoch": 0.8516483516483516,
      "grad_norm": 0.6085131764411926,
      "learning_rate": 1.6909141723615757e-05,
      "loss": 1.9583,
      "step": 9300
    },
    {
      "epoch": 0.8608058608058609,
      "grad_norm": 0.5231878757476807,
      "learning_rate": 1.5801988640553246e-05,
      "loss": 1.959,
      "step": 9400
    },
    {
      "epoch": 0.86996336996337,
      "grad_norm": 0.5782052278518677,
      "learning_rate": 1.4575099366429102e-05,
      "loss": 1.9578,
      "step": 9500
    },
    {
      "epoch": 0.8791208791208791,
      "grad_norm": 0.7281983494758606,
      "learning_rate": 1.3253793331810845e-05,
      "loss": 1.9626,
      "step": 9600
    },
    {
      "epoch": 0.8882783882783882,
      "grad_norm": 0.5513895750045776,
      "learning_rate": 1.1865338454926736e-05,
      "loss": 1.954,
      "step": 9700
    },
    {
      "epoch": 0.8974358974358975,
      "grad_norm": 0.6550291180610657,
      "learning_rate": 1.0438388411045472e-05,
      "loss": 1.9417,
      "step": 9800
    },
    {
      "epoch": 0.9065934065934066,
      "grad_norm": 0.6144053936004639,
      "learning_rate": 9.002391303856074e-06,
      "loss": 1.938,
      "step": 9900
    },
    {
      "epoch": 0.9157509157509157,
      "grad_norm": 0.6759700775146484,
      "learning_rate": 7.586981942153e-06,
      "loss": 1.964,
      "step": 10000
    },
    {
      "epoch": 0.924908424908425,
      "grad_norm": 0.6374625563621521,
      "learning_rate": 6.221370263470199e-06,
      "loss": 1.9537,
      "step": 10100
    },
    {
      "epoch": 0.9340659340659341,
      "grad_norm": 0.6961682438850403,
      "learning_rate": 4.933738525823817e-06,
      "loss": 1.9329,
      "step": 10200
    },
    {
      "epoch": 0.9432234432234432,
      "grad_norm": 0.6878940463066101,
      "learning_rate": 3.750659707774723e-06,
      "loss": 1.9359,
      "step": 10300
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 0.5925911068916321,
      "learning_rate": 2.696549119344161e-06,
      "loss": 1.9381,
      "step": 10400
    },
    {
      "epoch": 0.9615384615384616,
      "grad_norm": 0.5766844749450684,
      "learning_rate": 1.7931605409401575e-06,
      "loss": 1.9414,
      "step": 10500
    },
    {
      "epoch": 0.9706959706959707,
      "grad_norm": 0.5555196404457092,
      "learning_rate": 1.0591372885234885e-06,
      "loss": 1.9472,
      "step": 10600
    },
    {
      "epoch": 0.9798534798534798,
      "grad_norm": 0.7189186811447144,
      "learning_rate": 5.096274697241354e-07,
      "loss": 1.9425,
      "step": 10700
    },
    {
      "epoch": 0.989010989010989,
      "grad_norm": 0.5945115685462952,
      "learning_rate": 1.5597137090428095e-07,
      "loss": 1.9325,
      "step": 10800
    },
    {
      "epoch": 0.9981684981684982,
      "grad_norm": 0.5117030739784241,
      "learning_rate": 5.467426590739511e-09,
      "loss": 1.9445,
      "step": 10900
    },
    {
      "epoch": 1.0,
      "step": 10920,
      "total_flos": 1.59124040841796e+18,
      "train_loss": 2.048668092511076,
      "train_runtime": 3400.905,
      "train_samples_per_second": 51.372,
      "train_steps_per_second": 3.211
    }
  ],
  "logging_steps": 100,
  "max_steps": 10920,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.59124040841796e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}