{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 11895,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008406893652795292,
      "grad_norm": 0.09369386732578278,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.4107,
      "step": 100
    },
    {
      "epoch": 0.016813787305590584,
      "grad_norm": 0.162967249751091,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.4113,
      "step": 200
    },
    {
      "epoch": 0.025220680958385876,
      "grad_norm": 0.22761274874210358,
      "learning_rate": 6e-06,
      "loss": 2.4136,
      "step": 300
    },
    {
      "epoch": 0.03362757461118117,
      "grad_norm": 0.26982206106185913,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.4085,
      "step": 400
    },
    {
      "epoch": 0.042034468263976464,
      "grad_norm": 0.329841673374176,
      "learning_rate": 1e-05,
      "loss": 2.317,
      "step": 500
    },
    {
      "epoch": 0.05044136191677175,
      "grad_norm": 0.42722123861312866,
      "learning_rate": 1.2e-05,
      "loss": 2.334,
      "step": 600
    },
    {
      "epoch": 0.05884825556956705,
      "grad_norm": 0.4785993993282318,
      "learning_rate": 1.4e-05,
      "loss": 2.2867,
      "step": 700
    },
    {
      "epoch": 0.06725514922236234,
      "grad_norm": 0.6988652348518372,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.2499,
      "step": 800
    },
    {
      "epoch": 0.07566204287515763,
      "grad_norm": 0.6594943404197693,
      "learning_rate": 1.8e-05,
      "loss": 2.2158,
      "step": 900
    },
    {
      "epoch": 0.08406893652795293,
      "grad_norm": 0.5762264132499695,
      "learning_rate": 2e-05,
      "loss": 2.2117,
      "step": 1000
    },
    {
      "epoch": 0.09247583018074822,
      "grad_norm": 0.6388180255889893,
      "learning_rate": 1.999584295057038e-05,
      "loss": 2.1964,
      "step": 1100
    },
    {
      "epoch": 0.1008827238335435,
      "grad_norm": 0.7342377305030823,
      "learning_rate": 1.9983375258493504e-05,
      "loss": 2.1529,
      "step": 1200
    },
    {
      "epoch": 0.1092896174863388,
      "grad_norm": 0.637233555316925,
      "learning_rate": 1.996260728953182e-05,
      "loss": 2.14,
      "step": 1300
    },
    {
      "epoch": 0.1176965111391341,
      "grad_norm": 0.6634367108345032,
      "learning_rate": 1.9933556310380036e-05,
      "loss": 2.1433,
      "step": 1400
    },
    {
      "epoch": 0.12610340479192939,
      "grad_norm": 0.8501883149147034,
      "learning_rate": 1.9896246474309414e-05,
      "loss": 2.08,
      "step": 1500
    },
    {
      "epoch": 0.13451029844472467,
      "grad_norm": 0.8380800485610962,
      "learning_rate": 1.9850708801086507e-05,
      "loss": 2.1043,
      "step": 1600
    },
    {
      "epoch": 0.14291719209751996,
      "grad_norm": 0.7901475429534912,
      "learning_rate": 1.9796981151183013e-05,
      "loss": 2.0605,
      "step": 1700
    },
    {
      "epoch": 0.15132408575031525,
      "grad_norm": 1.0150291919708252,
      "learning_rate": 1.9735108194298206e-05,
      "loss": 2.0725,
      "step": 1800
    },
    {
      "epoch": 0.15973097940311054,
      "grad_norm": 0.8598669767379761,
      "learning_rate": 1.9665141372220112e-05,
      "loss": 2.0708,
      "step": 1900
    },
    {
      "epoch": 0.16813787305590586,
      "grad_norm": 0.7413082122802734,
      "learning_rate": 1.9587138856056303e-05,
      "loss": 2.0973,
      "step": 2000
    },
    {
      "epoch": 0.17654476670870115,
      "grad_norm": 0.994321346282959,
      "learning_rate": 1.9501165497869832e-05,
      "loss": 2.0324,
      "step": 2100
    },
    {
      "epoch": 0.18495166036149643,
      "grad_norm": 0.9418250918388367,
      "learning_rate": 1.940729277676063e-05,
      "loss": 2.0269,
      "step": 2200
    },
    {
      "epoch": 0.19335855401429172,
      "grad_norm": 0.9771312475204468,
      "learning_rate": 1.930559873943704e-05,
      "loss": 2.0302,
      "step": 2300
    },
    {
      "epoch": 0.201765447667087,
      "grad_norm": 0.9368925094604492,
      "learning_rate": 1.9196167935327036e-05,
      "loss": 2.068,
      "step": 2400
    },
    {
      "epoch": 0.2101723413198823,
      "grad_norm": 0.797228991985321,
      "learning_rate": 1.9079091346282977e-05,
      "loss": 2.0227,
      "step": 2500
    },
    {
      "epoch": 0.2185792349726776,
      "grad_norm": 1.143477439880371,
      "learning_rate": 1.8954466310938405e-05,
      "loss": 2.0179,
      "step": 2600
    },
    {
      "epoch": 0.22698612862547288,
      "grad_norm": 1.1722630262374878,
      "learning_rate": 1.8822396443779745e-05,
      "loss": 2.0476,
      "step": 2700
    },
    {
      "epoch": 0.2353930222782682,
      "grad_norm": 1.155578374862671,
      "learning_rate": 1.868299154900018e-05,
      "loss": 2.0693,
      "step": 2800
    },
    {
      "epoch": 0.24379991593106348,
      "grad_norm": 0.8934180736541748,
      "learning_rate": 1.8536367529207375e-05,
      "loss": 2.0253,
      "step": 2900
    },
    {
      "epoch": 0.25220680958385877,
      "grad_norm": 0.8130605816841125,
      "learning_rate": 1.83826462890609e-05,
      "loss": 2.028,
      "step": 3000
    },
    {
      "epoch": 0.26061370323665406,
      "grad_norm": 0.89918452501297,
      "learning_rate": 1.8221955633919495e-05,
      "loss": 2.0296,
      "step": 3100
    },
    {
      "epoch": 0.26902059688944935,
      "grad_norm": 0.8746682405471802,
      "learning_rate": 1.8054429163582415e-05,
      "loss": 2.0331,
      "step": 3200
    },
    {
      "epoch": 0.27742749054224464,
      "grad_norm": 0.8571292161941528,
      "learning_rate": 1.7880206161213255e-05,
      "loss": 2.024,
      "step": 3300
    },
    {
      "epoch": 0.2858343841950399,
      "grad_norm": 0.870735764503479,
      "learning_rate": 1.7699431477538534e-05,
      "loss": 2.0122,
      "step": 3400
    },
    {
      "epoch": 0.2942412778478352,
      "grad_norm": 0.9568698406219482,
      "learning_rate": 1.7512255410417392e-05,
      "loss": 2.0081,
      "step": 3500
    },
    {
      "epoch": 0.3026481715006305,
      "grad_norm": 1.3627972602844238,
      "learning_rate": 1.731883357988244e-05,
      "loss": 2.0088,
      "step": 3600
    },
    {
      "epoch": 0.3110550651534258,
      "grad_norm": 0.857333779335022,
      "learning_rate": 1.7119326798755734e-05,
      "loss": 1.9904,
      "step": 3700
    },
    {
      "epoch": 0.3194619588062211,
      "grad_norm": 1.0471360683441162,
      "learning_rate": 1.6913900938947417e-05,
      "loss": 1.98,
      "step": 3800
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 0.9307909607887268,
      "learning_rate": 1.6702726793548153e-05,
      "loss": 2.0083,
      "step": 3900
    },
    {
      "epoch": 0.3362757461118117,
      "grad_norm": 1.0177804231643677,
      "learning_rate": 1.6485979934830084e-05,
      "loss": 2.0035,
      "step": 4000
    },
    {
      "epoch": 0.344682639764607,
      "grad_norm": 1.2335155010223389,
      "learning_rate": 1.626384056827429e-05,
      "loss": 1.991,
      "step": 4100
    },
    {
      "epoch": 0.3530895334174023,
      "grad_norm": 1.20236074924469,
      "learning_rate": 1.6036493382746178e-05,
      "loss": 1.9894,
      "step": 4200
    },
    {
      "epoch": 0.3614964270701976,
      "grad_norm": 0.9622140526771545,
      "learning_rate": 1.580412739694333e-05,
      "loss": 2.0082,
      "step": 4300
    },
    {
      "epoch": 0.36990332072299287,
      "grad_norm": 2.1568732261657715,
      "learning_rate": 1.5566935802243496e-05,
      "loss": 2.0051,
      "step": 4400
    },
    {
      "epoch": 0.37831021437578816,
      "grad_norm": 1.0069299936294556,
      "learning_rate": 1.5325115802083373e-05,
      "loss": 2.0201,
      "step": 4500
    },
    {
      "epoch": 0.38671710802858345,
      "grad_norm": 0.6649059057235718,
      "learning_rate": 1.5078868448001704e-05,
      "loss": 2.0366,
      "step": 4600
    },
    {
      "epoch": 0.39512400168137873,
      "grad_norm": 1.0054945945739746,
      "learning_rate": 1.4828398472483057e-05,
      "loss": 1.9987,
      "step": 4700
    },
    {
      "epoch": 0.403530895334174,
      "grad_norm": 1.0743929147720337,
      "learning_rate": 1.4573914118741201e-05,
      "loss": 1.9302,
      "step": 4800
    },
    {
      "epoch": 0.4119377889869693,
      "grad_norm": 0.9288580417633057,
      "learning_rate": 1.4315626967583657e-05,
      "loss": 1.9546,
      "step": 4900
    },
    {
      "epoch": 0.4203446826397646,
      "grad_norm": 0.9618598818778992,
      "learning_rate": 1.40537517615013e-05,
      "loss": 2.0003,
      "step": 5000
    },
    {
      "epoch": 0.4287515762925599,
      "grad_norm": 0.7777699828147888,
      "learning_rate": 1.3788506226129348e-05,
      "loss": 1.9651,
      "step": 5100
    },
    {
      "epoch": 0.4371584699453552,
      "grad_norm": 1.044360876083374,
      "learning_rate": 1.3520110889228104e-05,
      "loss": 1.9318,
      "step": 5200
    },
    {
      "epoch": 0.44556536359815047,
      "grad_norm": 0.8843343257904053,
      "learning_rate": 1.3248788897334005e-05,
      "loss": 1.9613,
      "step": 5300
    },
    {
      "epoch": 0.45397225725094575,
      "grad_norm": 1.093237042427063,
      "learning_rate": 1.2974765830233383e-05,
      "loss": 2.0139,
      "step": 5400
    },
    {
      "epoch": 0.46237915090374104,
      "grad_norm": 1.0078516006469727,
      "learning_rate": 1.269826951341319e-05,
      "loss": 1.9448,
      "step": 5500
    },
    {
      "epoch": 0.4707860445565364,
      "grad_norm": 1.0868773460388184,
      "learning_rate": 1.2419529828644661e-05,
      "loss": 1.9463,
      "step": 5600
    },
    {
      "epoch": 0.4791929382093317,
      "grad_norm": 1.2549597024917603,
      "learning_rate": 1.2138778522857307e-05,
      "loss": 1.966,
      "step": 5700
    },
    {
      "epoch": 0.48759983186212696,
      "grad_norm": 1.0770331621170044,
      "learning_rate": 1.1856249015462242e-05,
      "loss": 1.9976,
      "step": 5800
    },
    {
      "epoch": 0.49600672551492225,
      "grad_norm": 1.5205333232879639,
      "learning_rate": 1.1572176204284986e-05,
      "loss": 1.9743,
      "step": 5900
    },
    {
      "epoch": 0.5044136191677175,
      "grad_norm": 1.101060390472412,
      "learning_rate": 1.1286796270269076e-05,
      "loss": 1.9878,
      "step": 6000
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 1.1352288722991943,
      "learning_rate": 1.1000346481112887e-05,
      "loss": 1.9468,
      "step": 6100
    },
    {
      "epoch": 0.5212274064733081,
      "grad_norm": 1.0924268960952759,
      "learning_rate": 1.0713064994002956e-05,
      "loss": 1.9728,
      "step": 6200
    },
    {
      "epoch": 0.5296343001261034,
      "grad_norm": 1.0215404033660889,
      "learning_rate": 1.0425190657607702e-05,
      "loss": 1.9489,
      "step": 6300
    },
    {
      "epoch": 0.5380411937788987,
      "grad_norm": 0.8387424349784851,
      "learning_rate": 1.0136962813496306e-05,
      "loss": 1.9977,
      "step": 6400
    },
    {
      "epoch": 0.546448087431694,
      "grad_norm": 0.8960559368133545,
      "learning_rate": 9.848621097147772e-06,
      "loss": 2.0569,
      "step": 6500
    },
    {
      "epoch": 0.5548549810844893,
      "grad_norm": 1.0311048030853271,
      "learning_rate": 9.560405238715588e-06,
      "loss": 1.9769,
      "step": 6600
    },
    {
      "epoch": 0.5632618747372846,
      "grad_norm": 1.3905280828475952,
      "learning_rate": 9.272554863713739e-06,
      "loss": 1.9868,
      "step": 6700
    },
    {
      "epoch": 0.5716687683900799,
      "grad_norm": 0.8861886858940125,
      "learning_rate": 8.985309293789662e-06,
      "loss": 2.0083,
      "step": 6800
    },
    {
      "epoch": 0.5800756620428752,
      "grad_norm": 1.1910375356674194,
      "learning_rate": 8.698907347749885e-06,
      "loss": 1.9396,
      "step": 6900
    },
    {
      "epoch": 0.5884825556956704,
      "grad_norm": 1.4200043678283691,
      "learning_rate": 8.413587143003691e-06,
      "loss": 1.938,
      "step": 7000
    },
    {
      "epoch": 0.5968894493484658,
      "grad_norm": 1.2304203510284424,
      "learning_rate": 8.129585897589964e-06,
      "loss": 1.9756,
      "step": 7100
    },
    {
      "epoch": 0.605296343001261,
      "grad_norm": 0.9935517907142639,
      "learning_rate": 7.847139732951751e-06,
      "loss": 1.9672,
      "step": 7200
    },
    {
      "epoch": 0.6137032366540563,
      "grad_norm": 0.8960947394371033,
      "learning_rate": 7.566483477622577e-06,
      "loss": 1.9919,
      "step": 7300
    },
    {
      "epoch": 0.6221101303068516,
      "grad_norm": 1.761006236076355,
      "learning_rate": 7.28785047198767e-06,
      "loss": 1.9473,
      "step": 7400
    },
    {
      "epoch": 0.6305170239596469,
      "grad_norm": 0.9094328880310059,
      "learning_rate": 7.011472374282457e-06,
      "loss": 1.9713,
      "step": 7500
    },
    {
      "epoch": 0.6389239176124422,
      "grad_norm": 1.645095944404602,
      "learning_rate": 6.737578967989626e-06,
      "loss": 1.9944,
      "step": 7600
    },
    {
      "epoch": 0.6473308112652375,
      "grad_norm": 0.8385418653488159,
      "learning_rate": 6.466397970794855e-06,
      "loss": 1.9474,
      "step": 7700
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 0.8688055276870728,
      "learning_rate": 6.198154845260089e-06,
      "loss": 1.9178,
      "step": 7800
    },
    {
      "epoch": 0.6641445985708281,
      "grad_norm": 1.0304285287857056,
      "learning_rate": 5.933072611371725e-06,
      "loss": 2.0074,
      "step": 7900
    },
    {
      "epoch": 0.6725514922236234,
      "grad_norm": 1.312940001487732,
      "learning_rate": 5.671371661119609e-06,
      "loss": 2.0371,
      "step": 8000
    },
    {
      "epoch": 0.6809583858764187,
      "grad_norm": 1.2208335399627686,
      "learning_rate": 5.413269575260924e-06,
      "loss": 1.9319,
      "step": 8100
    },
    {
      "epoch": 0.689365279529214,
      "grad_norm": 1.1482669115066528,
      "learning_rate": 5.161504255249876e-06,
      "loss": 1.9766,
      "step": 8200
    },
    {
      "epoch": 0.6977721731820092,
      "grad_norm": 0.839103102684021,
      "learning_rate": 4.911199209125536e-06,
      "loss": 1.9854,
      "step": 8300
    },
    {
      "epoch": 0.7061790668348046,
      "grad_norm": 0.9994194507598877,
      "learning_rate": 4.665125042286226e-06,
      "loss": 1.961,
      "step": 8400
    },
    {
      "epoch": 0.7145859604875998,
      "grad_norm": 1.0879062414169312,
      "learning_rate": 4.423486343226934e-06,
      "loss": 1.9287,
      "step": 8500
    },
    {
      "epoch": 0.7229928541403952,
      "grad_norm": 1.0109790563583374,
      "learning_rate": 4.186484012750876e-06,
      "loss": 2.0442,
      "step": 8600
    },
    {
      "epoch": 0.7313997477931904,
      "grad_norm": 1.5959651470184326,
      "learning_rate": 3.954315096938598e-06,
      "loss": 1.9502,
      "step": 8700
    },
    {
      "epoch": 0.7398066414459857,
      "grad_norm": 1.3326095342636108,
      "learning_rate": 3.7271726233219098e-06,
      "loss": 1.9527,
      "step": 8800
    },
    {
      "epoch": 0.748213535098781,
      "grad_norm": 0.8534725308418274,
      "learning_rate": 3.50524544039889e-06,
      "loss": 1.9636,
      "step": 8900
    },
    {
      "epoch": 0.7566204287515763,
      "grad_norm": 1.1570713520050049,
      "learning_rate": 3.288718060623376e-06,
      "loss": 1.9624,
      "step": 9000
    },
    {
      "epoch": 0.7650273224043715,
      "grad_norm": 1.052301287651062,
      "learning_rate": 3.079851782604133e-06,
      "loss": 1.9601,
      "step": 9100
    },
    {
      "epoch": 0.7734342160571669,
      "grad_norm": 0.9507755637168884,
      "learning_rate": 2.8746010343552233e-06,
      "loss": 1.958,
      "step": 9200
    },
    {
      "epoch": 0.7818411097099621,
      "grad_norm": 1.406473159790039,
      "learning_rate": 2.6752744132475063e-06,
      "loss": 1.9672,
      "step": 9300
    },
    {
      "epoch": 0.7902480033627575,
      "grad_norm": 1.400251030921936,
      "learning_rate": 2.482037641404298e-06,
      "loss": 1.9376,
      "step": 9400
    },
    {
      "epoch": 0.7986548970155527,
      "grad_norm": 1.6398078203201294,
      "learning_rate": 2.295051377788038e-06,
      "loss": 1.8933,
      "step": 9500
    },
    {
      "epoch": 0.807061790668348,
      "grad_norm": 0.9706113338470459,
      "learning_rate": 2.1144710846268247e-06,
      "loss": 1.9434,
      "step": 9600
    },
    {
      "epoch": 0.8154686843211434,
      "grad_norm": 1.5242693424224854,
      "learning_rate": 1.9404468981615997e-06,
      "loss": 1.9754,
      "step": 9700
    },
    {
      "epoch": 0.8238755779739386,
      "grad_norm": 0.9875818490982056,
      "learning_rate": 1.7731235038213745e-06,
      "loss": 1.9225,
      "step": 9800
    },
    {
      "epoch": 0.832282471626734,
      "grad_norm": 1.1812236309051514,
      "learning_rate": 1.6126400159303524e-06,
      "loss": 1.933,
      "step": 9900
    },
    {
      "epoch": 0.8406893652795292,
      "grad_norm": 1.3592262268066406,
      "learning_rate": 1.4591298620468918e-06,
      "loss": 1.9412,
      "step": 10000
    },
    {
      "epoch": 0.8490962589323245,
      "grad_norm": 1.4087212085723877,
      "learning_rate": 1.3127206720305264e-06,
      "loss": 1.9143,
      "step": 10100
    },
    {
      "epoch": 0.8575031525851198,
      "grad_norm": 1.1176539659500122,
      "learning_rate": 1.1735341719292214e-06,
      "loss": 1.9434,
      "step": 10200
    },
    {
      "epoch": 0.8659100462379151,
      "grad_norm": 1.1962157487869263,
      "learning_rate": 1.0416860827751508e-06,
      "loss": 1.9573,
      "step": 10300
    },
    {
      "epoch": 0.8743169398907104,
      "grad_norm": 1.2867956161499023,
      "learning_rate": 9.172860243730752e-07,
      "loss": 1.95,
      "step": 10400
    },
    {
      "epoch": 0.8827238335435057,
      "grad_norm": 1.2648465633392334,
      "learning_rate": 8.004374241613621e-07,
      "loss": 2.0074,
      "step": 10500
    },
    {
      "epoch": 0.8911307271963009,
      "grad_norm": 0.9942221641540527,
      "learning_rate": 6.912374312213821e-07,
      "loss": 2.0051,
      "step": 10600
    },
    {
      "epoch": 0.8995376208490963,
      "grad_norm": 1.1690579652786255,
      "learning_rate": 5.897768355068101e-07,
      "loss": 1.9404,
      "step": 10700
    },
    {
      "epoch": 0.9079445145018915,
      "grad_norm": 1.0716674327850342,
      "learning_rate": 4.96139992359953e-07,
      "loss": 1.8911,
      "step": 10800
    },
    {
      "epoch": 0.9163514081546869,
      "grad_norm": 0.888635516166687,
      "learning_rate": 4.10404752377892e-07,
      "loss": 1.9684,
      "step": 10900
    },
    {
      "epoch": 0.9247583018074821,
      "grad_norm": 1.379387617111206,
      "learning_rate": 3.3264239668671873e-07,
      "loss": 1.9436,
      "step": 11000
    },
    {
      "epoch": 0.9331651954602774,
      "grad_norm": 1.0186644792556763,
      "learning_rate": 2.629175776777071e-07,
      "loss": 1.9699,
      "step": 11100
    },
    {
      "epoch": 0.9415720891130728,
      "grad_norm": 0.8766846060752869,
      "learning_rate": 2.0128826525467772e-07,
      "loss": 2.0177,
      "step": 11200
    },
    {
      "epoch": 0.949978982765868,
      "grad_norm": 1.1881805658340454,
      "learning_rate": 1.4780569863723894e-07,
      "loss": 1.9004,
      "step": 11300
    },
    {
      "epoch": 0.9583858764186634,
      "grad_norm": 1.2002122402191162,
      "learning_rate": 1.0251434376000402e-07,
      "loss": 1.9331,
      "step": 11400
    },
    {
      "epoch": 0.9667927700714586,
      "grad_norm": 1.5120991468429565,
      "learning_rate": 6.545185630316364e-08,
      "loss": 2.0019,
      "step": 11500
    },
    {
      "epoch": 0.9751996637242539,
      "grad_norm": 1.2913029193878174,
      "learning_rate": 3.6649050385185734e-08,
      "loss": 1.9446,
      "step": 11600
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 1.2726212739944458,
      "learning_rate": 1.612987294365431e-08,
      "loss": 1.933,
      "step": 11700
    },
    {
      "epoch": 0.9920134510298445,
      "grad_norm": 1.3567242622375488,
      "learning_rate": 3.9113838255433466e-09,
      "loss": 2.0148,
      "step": 11800
    },
    {
      "epoch": 1.0,
      "step": 11895,
      "total_flos": 2.1665532198912e+17,
      "train_loss": 2.01907276283447,
      "train_runtime": 3147.1141,
      "train_samples_per_second": 7.559,
      "train_steps_per_second": 3.78
    }
  ],
  "logging_steps": 100,
  "max_steps": 11895,
  "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": 2.1665532198912e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}