{
  "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": 0.14043055474758148,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.4104,
      "step": 100
    },
    {
      "epoch": 0.016813787305590584,
      "grad_norm": 0.2322196364402771,
      "learning_rate": 4.000000000000001e-06,
      "loss": 2.4058,
      "step": 200
    },
    {
      "epoch": 0.025220680958385876,
      "grad_norm": 0.330422580242157,
      "learning_rate": 6e-06,
      "loss": 2.3956,
      "step": 300
    },
    {
      "epoch": 0.03362757461118117,
      "grad_norm": 0.37281373143196106,
      "learning_rate": 8.000000000000001e-06,
      "loss": 2.3817,
      "step": 400
    },
    {
      "epoch": 0.042034468263976464,
      "grad_norm": 0.4371340274810791,
      "learning_rate": 1e-05,
      "loss": 2.2819,
      "step": 500
    },
    {
      "epoch": 0.05044136191677175,
      "grad_norm": 0.5313137173652649,
      "learning_rate": 1.2e-05,
      "loss": 2.2971,
      "step": 600
    },
    {
      "epoch": 0.05884825556956705,
      "grad_norm": 0.5993666648864746,
      "learning_rate": 1.4e-05,
      "loss": 2.2486,
      "step": 700
    },
    {
      "epoch": 0.06725514922236234,
      "grad_norm": 0.885300874710083,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 2.2134,
      "step": 800
    },
    {
      "epoch": 0.07566204287515763,
      "grad_norm": 0.7315155267715454,
      "learning_rate": 1.8e-05,
      "loss": 2.1677,
      "step": 900
    },
    {
      "epoch": 0.08406893652795293,
      "grad_norm": 0.6705694794654846,
      "learning_rate": 2e-05,
      "loss": 2.1677,
      "step": 1000
    },
    {
      "epoch": 0.09247583018074822,
      "grad_norm": 0.6811093688011169,
      "learning_rate": 1.999584295057038e-05,
      "loss": 2.1539,
      "step": 1100
    },
    {
      "epoch": 0.1008827238335435,
      "grad_norm": 0.8088228702545166,
      "learning_rate": 1.9983375258493504e-05,
      "loss": 2.1099,
      "step": 1200
    },
    {
      "epoch": 0.1092896174863388,
      "grad_norm": 0.6950477361679077,
      "learning_rate": 1.996260728953182e-05,
      "loss": 2.0899,
      "step": 1300
    },
    {
      "epoch": 0.1176965111391341,
      "grad_norm": 0.746992290019989,
      "learning_rate": 1.9933556310380036e-05,
      "loss": 2.1047,
      "step": 1400
    },
    {
      "epoch": 0.12610340479192939,
      "grad_norm": 0.9399320483207703,
      "learning_rate": 1.9896246474309414e-05,
      "loss": 2.0344,
      "step": 1500
    },
    {
      "epoch": 0.13451029844472467,
      "grad_norm": 0.8537988066673279,
      "learning_rate": 1.9850708801086507e-05,
      "loss": 2.0576,
      "step": 1600
    },
    {
      "epoch": 0.14291719209751996,
      "grad_norm": 0.8270643949508667,
      "learning_rate": 1.9796981151183013e-05,
      "loss": 2.019,
      "step": 1700
    },
    {
      "epoch": 0.15132408575031525,
      "grad_norm": 1.1590845584869385,
      "learning_rate": 1.9735108194298206e-05,
      "loss": 2.0348,
      "step": 1800
    },
    {
      "epoch": 0.15973097940311054,
      "grad_norm": 0.8914215564727783,
      "learning_rate": 1.9665141372220112e-05,
      "loss": 2.0314,
      "step": 1900
    },
    {
      "epoch": 0.16813787305590586,
      "grad_norm": 0.7579370141029358,
      "learning_rate": 1.9587138856056303e-05,
      "loss": 2.0636,
      "step": 2000
    },
    {
      "epoch": 0.17654476670870115,
      "grad_norm": 0.9895981550216675,
      "learning_rate": 1.9501165497869832e-05,
      "loss": 1.9961,
      "step": 2100
    },
    {
      "epoch": 0.18495166036149643,
      "grad_norm": 0.9374507069587708,
      "learning_rate": 1.940729277676063e-05,
      "loss": 1.9923,
      "step": 2200
    },
    {
      "epoch": 0.19335855401429172,
      "grad_norm": 0.9945927858352661,
      "learning_rate": 1.930559873943704e-05,
      "loss": 1.9944,
      "step": 2300
    },
    {
      "epoch": 0.201765447667087,
      "grad_norm": 0.9538114070892334,
      "learning_rate": 1.9196167935327036e-05,
      "loss": 2.036,
      "step": 2400
    },
    {
      "epoch": 0.2101723413198823,
      "grad_norm": 0.780885636806488,
      "learning_rate": 1.9079091346282977e-05,
      "loss": 1.9894,
      "step": 2500
    },
    {
      "epoch": 0.2185792349726776,
      "grad_norm": 1.2008519172668457,
      "learning_rate": 1.8954466310938405e-05,
      "loss": 1.9837,
      "step": 2600
    },
    {
      "epoch": 0.22698612862547288,
      "grad_norm": 1.089125394821167,
      "learning_rate": 1.8822396443779745e-05,
      "loss": 2.016,
      "step": 2700
    },
    {
      "epoch": 0.2353930222782682,
      "grad_norm": 1.087217092514038,
      "learning_rate": 1.868299154900018e-05,
      "loss": 2.0407,
      "step": 2800
    },
    {
      "epoch": 0.24379991593106348,
      "grad_norm": 0.9000464677810669,
      "learning_rate": 1.8536367529207375e-05,
      "loss": 1.9954,
      "step": 2900
    },
    {
      "epoch": 0.25220680958385877,
      "grad_norm": 0.7694302201271057,
      "learning_rate": 1.83826462890609e-05,
      "loss": 1.9973,
      "step": 3000
    },
    {
      "epoch": 0.26061370323665406,
      "grad_norm": 0.8629293441772461,
      "learning_rate": 1.8221955633919495e-05,
      "loss": 1.9994,
      "step": 3100
    },
    {
      "epoch": 0.26902059688944935,
      "grad_norm": 0.839648962020874,
      "learning_rate": 1.8054429163582415e-05,
      "loss": 2.003,
      "step": 3200
    },
    {
      "epoch": 0.27742749054224464,
      "grad_norm": 0.8285431861877441,
      "learning_rate": 1.7880206161213255e-05,
      "loss": 1.9925,
      "step": 3300
    },
    {
      "epoch": 0.2858343841950399,
      "grad_norm": 0.837740421295166,
      "learning_rate": 1.7699431477538534e-05,
      "loss": 1.9801,
      "step": 3400
    },
    {
      "epoch": 0.2942412778478352,
      "grad_norm": 0.8980201482772827,
      "learning_rate": 1.7512255410417392e-05,
      "loss": 1.9764,
      "step": 3500
    },
    {
      "epoch": 0.3026481715006305,
      "grad_norm": 1.3097134828567505,
      "learning_rate": 1.731883357988244e-05,
      "loss": 1.9771,
      "step": 3600
    },
    {
      "epoch": 0.3110550651534258,
      "grad_norm": 0.7765768766403198,
      "learning_rate": 1.7119326798755734e-05,
      "loss": 1.9589,
      "step": 3700
    },
    {
      "epoch": 0.3194619588062211,
      "grad_norm": 1.0500297546386719,
      "learning_rate": 1.6913900938947417e-05,
      "loss": 1.9478,
      "step": 3800
    },
    {
      "epoch": 0.32786885245901637,
      "grad_norm": 0.9133660793304443,
      "learning_rate": 1.6702726793548153e-05,
      "loss": 1.9782,
      "step": 3900
    },
    {
      "epoch": 0.3362757461118117,
      "grad_norm": 0.9790305495262146,
      "learning_rate": 1.6485979934830084e-05,
      "loss": 1.9712,
      "step": 4000
    },
    {
      "epoch": 0.344682639764607,
      "grad_norm": 1.1819267272949219,
      "learning_rate": 1.626384056827429e-05,
      "loss": 1.9597,
      "step": 4100
    },
    {
      "epoch": 0.3530895334174023,
      "grad_norm": 1.196151852607727,
      "learning_rate": 1.6036493382746178e-05,
      "loss": 1.9571,
      "step": 4200
    },
    {
      "epoch": 0.3614964270701976,
      "grad_norm": 0.8918384909629822,
      "learning_rate": 1.580412739694333e-05,
      "loss": 1.9769,
      "step": 4300
    },
    {
      "epoch": 0.36990332072299287,
      "grad_norm": 2.0516951084136963,
      "learning_rate": 1.5566935802243496e-05,
      "loss": 1.9757,
      "step": 4400
    },
    {
      "epoch": 0.37831021437578816,
      "grad_norm": 0.9879031777381897,
      "learning_rate": 1.5325115802083373e-05,
      "loss": 1.9893,
      "step": 4500
    },
    {
      "epoch": 0.38671710802858345,
      "grad_norm": 0.6235149502754211,
      "learning_rate": 1.5078868448001704e-05,
      "loss": 2.008,
      "step": 4600
    },
    {
      "epoch": 0.39512400168137873,
      "grad_norm": 0.9613019824028015,
      "learning_rate": 1.4828398472483057e-05,
      "loss": 1.9669,
      "step": 4700
    },
    {
      "epoch": 0.403530895334174,
      "grad_norm": 1.0296316146850586,
      "learning_rate": 1.4573914118741201e-05,
      "loss": 1.8964,
      "step": 4800
    },
    {
      "epoch": 0.4119377889869693,
      "grad_norm": 0.8810610771179199,
      "learning_rate": 1.4315626967583657e-05,
      "loss": 1.9226,
      "step": 4900
    },
    {
      "epoch": 0.4203446826397646,
      "grad_norm": 0.9355632066726685,
      "learning_rate": 1.40537517615013e-05,
      "loss": 1.9678,
      "step": 5000
    },
    {
      "epoch": 0.4287515762925599,
      "grad_norm": 0.7143288850784302,
      "learning_rate": 1.3788506226129348e-05,
      "loss": 1.9328,
      "step": 5100
    },
    {
      "epoch": 0.4371584699453552,
      "grad_norm": 0.987920880317688,
      "learning_rate": 1.3520110889228104e-05,
      "loss": 1.8968,
      "step": 5200
    },
    {
      "epoch": 0.44556536359815047,
      "grad_norm": 0.8306659460067749,
      "learning_rate": 1.3248788897334005e-05,
      "loss": 1.9294,
      "step": 5300
    },
    {
      "epoch": 0.45397225725094575,
      "grad_norm": 1.0377694368362427,
      "learning_rate": 1.2974765830233383e-05,
      "loss": 1.9835,
      "step": 5400
    },
    {
      "epoch": 0.46237915090374104,
      "grad_norm": 0.9360339641571045,
      "learning_rate": 1.269826951341319e-05,
      "loss": 1.9121,
      "step": 5500
    },
    {
      "epoch": 0.4707860445565364,
      "grad_norm": 1.0214323997497559,
      "learning_rate": 1.2419529828644661e-05,
      "loss": 1.9133,
      "step": 5600
    },
    {
      "epoch": 0.4791929382093317,
      "grad_norm": 1.1800683736801147,
      "learning_rate": 1.2138778522857307e-05,
      "loss": 1.934,
      "step": 5700
    },
    {
      "epoch": 0.48759983186212696,
      "grad_norm": 1.0064023733139038,
      "learning_rate": 1.1856249015462242e-05,
      "loss": 1.9665,
      "step": 5800
    },
    {
      "epoch": 0.49600672551492225,
      "grad_norm": 1.3952449560165405,
      "learning_rate": 1.1572176204284986e-05,
      "loss": 1.9426,
      "step": 5900
    },
    {
      "epoch": 0.5044136191677175,
      "grad_norm": 1.0473861694335938,
      "learning_rate": 1.1286796270269076e-05,
      "loss": 1.9559,
      "step": 6000
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 1.101931095123291,
      "learning_rate": 1.1000346481112887e-05,
      "loss": 1.9103,
      "step": 6100
    },
    {
      "epoch": 0.5212274064733081,
      "grad_norm": 1.0466363430023193,
      "learning_rate": 1.0713064994002956e-05,
      "loss": 1.9395,
      "step": 6200
    },
    {
      "epoch": 0.5296343001261034,
      "grad_norm": 0.9257377982139587,
      "learning_rate": 1.0425190657607702e-05,
      "loss": 1.9141,
      "step": 6300
    },
    {
      "epoch": 0.5380411937788987,
      "grad_norm": 0.779354453086853,
      "learning_rate": 1.0136962813496306e-05,
      "loss": 1.9647,
      "step": 6400
    },
    {
      "epoch": 0.546448087431694,
      "grad_norm": 0.8354730606079102,
      "learning_rate": 9.848621097147772e-06,
      "loss": 2.0283,
      "step": 6500
    },
    {
      "epoch": 0.5548549810844893,
      "grad_norm": 1.0222142934799194,
      "learning_rate": 9.560405238715588e-06,
      "loss": 1.9444,
      "step": 6600
    },
    {
      "epoch": 0.5632618747372846,
      "grad_norm": 1.3046057224273682,
      "learning_rate": 9.272554863713739e-06,
      "loss": 1.9555,
      "step": 6700
    },
    {
      "epoch": 0.5716687683900799,
      "grad_norm": 0.8095905184745789,
      "learning_rate": 8.985309293789662e-06,
      "loss": 1.9761,
      "step": 6800
    },
    {
      "epoch": 0.5800756620428752,
      "grad_norm": 1.0805299282073975,
      "learning_rate": 8.698907347749885e-06,
      "loss": 1.9043,
      "step": 6900
    },
    {
      "epoch": 0.5884825556956704,
      "grad_norm": 1.3454480171203613,
      "learning_rate": 8.413587143003691e-06,
      "loss": 1.902,
      "step": 7000
    },
    {
      "epoch": 0.5968894493484658,
      "grad_norm": 1.1316404342651367,
      "learning_rate": 8.129585897589964e-06,
      "loss": 1.942,
      "step": 7100
    },
    {
      "epoch": 0.605296343001261,
      "grad_norm": 0.9173849821090698,
      "learning_rate": 7.847139732951751e-06,
      "loss": 1.9336,
      "step": 7200
    },
    {
      "epoch": 0.6137032366540563,
      "grad_norm": 0.8359085321426392,
      "learning_rate": 7.566483477622577e-06,
      "loss": 1.9584,
      "step": 7300
    },
    {
      "epoch": 0.6221101303068516,
      "grad_norm": 1.6915035247802734,
      "learning_rate": 7.28785047198767e-06,
      "loss": 1.9113,
      "step": 7400
    },
    {
      "epoch": 0.6305170239596469,
      "grad_norm": 0.8478715419769287,
      "learning_rate": 7.011472374282457e-06,
      "loss": 1.9362,
      "step": 7500
    },
    {
      "epoch": 0.6389239176124422,
      "grad_norm": 1.5024291276931763,
      "learning_rate": 6.737578967989626e-06,
      "loss": 1.9636,
      "step": 7600
    },
    {
      "epoch": 0.6473308112652375,
      "grad_norm": 0.7754480242729187,
      "learning_rate": 6.466397970794855e-06,
      "loss": 1.9107,
      "step": 7700
    },
    {
      "epoch": 0.6557377049180327,
      "grad_norm": 0.7962978482246399,
      "learning_rate": 6.198154845260089e-06,
      "loss": 1.8819,
      "step": 7800
    },
    {
      "epoch": 0.6641445985708281,
      "grad_norm": 0.9278832077980042,
      "learning_rate": 5.933072611371725e-06,
      "loss": 1.9741,
      "step": 7900
    },
    {
      "epoch": 0.6725514922236234,
      "grad_norm": 1.244887351989746,
      "learning_rate": 5.671371661119609e-06,
      "loss": 2.0055,
      "step": 8000
    },
    {
      "epoch": 0.6809583858764187,
      "grad_norm": 1.1086173057556152,
      "learning_rate": 5.415832074569403e-06,
      "loss": 1.8962,
      "step": 8100
    },
    {
      "epoch": 0.689365279529214,
      "grad_norm": 1.0748802423477173,
      "learning_rate": 5.161504255249876e-06,
      "loss": 1.9434,
      "step": 8200
    },
    {
      "epoch": 0.6977721731820092,
      "grad_norm": 0.7687519788742065,
      "learning_rate": 4.911199209125536e-06,
      "loss": 1.95,
      "step": 8300
    },
    {
      "epoch": 0.7061790668348046,
      "grad_norm": 0.9328936338424683,
      "learning_rate": 4.665125042286226e-06,
      "loss": 1.9271,
      "step": 8400
    },
    {
      "epoch": 0.7145859604875998,
      "grad_norm": 1.0116279125213623,
      "learning_rate": 4.423486343226934e-06,
      "loss": 1.8916,
      "step": 8500
    },
    {
      "epoch": 0.7229928541403952,
      "grad_norm": 0.9574350714683533,
      "learning_rate": 4.186484012750876e-06,
      "loss": 2.0135,
      "step": 8600
    },
    {
      "epoch": 0.7313997477931904,
      "grad_norm": 1.5315290689468384,
      "learning_rate": 3.954315096938598e-06,
      "loss": 1.9159,
      "step": 8700
    },
    {
      "epoch": 0.7398066414459857,
      "grad_norm": 1.2323825359344482,
      "learning_rate": 3.7271726233219098e-06,
      "loss": 1.9175,
      "step": 8800
    },
    {
      "epoch": 0.748213535098781,
      "grad_norm": 0.7751179933547974,
      "learning_rate": 3.50524544039889e-06,
      "loss": 1.9278,
      "step": 8900
    },
    {
      "epoch": 0.7566204287515763,
      "grad_norm": 1.096082329750061,
      "learning_rate": 3.288718060623376e-06,
      "loss": 1.9253,
      "step": 9000
    },
    {
      "epoch": 0.7650273224043715,
      "grad_norm": 0.9929354786872864,
      "learning_rate": 3.077770506999487e-06,
      "loss": 1.9252,
      "step": 9100
    },
    {
      "epoch": 0.7734342160571669,
      "grad_norm": 0.8709311485290527,
      "learning_rate": 2.872578163408717e-06,
      "loss": 1.9219,
      "step": 9200
    },
    {
      "epoch": 0.7818411097099621,
      "grad_norm": 1.3237998485565186,
      "learning_rate": 2.673311628794043e-06,
      "loss": 1.9318,
      "step": 9300
    },
    {
      "epoch": 0.7902480033627575,
      "grad_norm": 1.3490018844604492,
      "learning_rate": 2.4801365753222795e-06,
      "loss": 1.9017,
      "step": 9400
    },
    {
      "epoch": 0.7986548970155527,
      "grad_norm": 1.5231995582580566,
      "learning_rate": 2.293213610642594e-06,
      "loss": 1.8539,
      "step": 9500
    },
    {
      "epoch": 0.807061790668348,
      "grad_norm": 0.9112081527709961,
      "learning_rate": 2.1126981443557303e-06,
      "loss": 1.9095,
      "step": 9600
    },
    {
      "epoch": 0.8154686843211434,
      "grad_norm": 1.3989226818084717,
      "learning_rate": 1.9387402588049167e-06,
      "loss": 1.9411,
      "step": 9700
    },
    {
      "epoch": 0.8238755779739386,
      "grad_norm": 0.9119231104850769,
      "learning_rate": 1.7714845842959415e-06,
      "loss": 1.8862,
      "step": 9800
    },
    {
      "epoch": 0.832282471626734,
      "grad_norm": 1.1409211158752441,
      "learning_rate": 1.6110701788500682e-06,
      "loss": 1.897,
      "step": 9900
    },
    {
      "epoch": 0.8406893652795292,
      "grad_norm": 1.3277921676635742,
      "learning_rate": 1.4576304125898234e-06,
      "loss": 1.9049,
      "step": 10000
    },
    {
      "epoch": 0.8490962589323245,
      "grad_norm": 1.3236161470413208,
      "learning_rate": 1.311292856853772e-06,
      "loss": 1.8771,
      "step": 10100
    },
    {
      "epoch": 0.8575031525851198,
      "grad_norm": 1.0454002618789673,
      "learning_rate": 1.1721791781324343e-06,
      "loss": 1.9074,
      "step": 10200
    },
    {
      "epoch": 0.8659100462379151,
      "grad_norm": 1.196260929107666,
      "learning_rate": 1.0404050369135698e-06,
      "loss": 1.9241,
      "step": 10300
    },
    {
      "epoch": 0.8743169398907104,
      "grad_norm": 1.2088180780410767,
      "learning_rate": 9.160799915208962e-07,
      "loss": 1.9161,
      "step": 10400
    },
    {
      "epoch": 0.8827238335435057,
      "grad_norm": 1.2079304456710815,
      "learning_rate": 7.993074070262185e-07,
      "loss": 1.9722,
      "step": 10500
    },
    {
      "epoch": 0.8911307271963009,
      "grad_norm": 0.9400559067726135,
      "learning_rate": 6.901843693106935e-07,
      "loss": 1.972,
      "step": 10600
    },
    {
      "epoch": 0.8995376208490963,
      "grad_norm": 1.1368870735168457,
      "learning_rate": 5.888016043466583e-07,
      "loss": 1.9051,
      "step": 10700
    },
    {
      "epoch": 0.9079445145018915,
      "grad_norm": 1.0495091676712036,
      "learning_rate": 4.952434027671659e-07,
      "loss": 1.8549,
      "step": 10800
    },
    {
      "epoch": 0.9163514081546869,
      "grad_norm": 0.832502007484436,
      "learning_rate": 4.095875497859192e-07,
      "loss": 1.9349,
      "step": 10900
    },
    {
      "epoch": 0.9247583018074821,
      "grad_norm": 1.295791745185852,
      "learning_rate": 3.3190526052587545e-07,
      "loss": 1.9073,
      "step": 11000
    },
    {
      "epoch": 0.9331651954602774,
      "grad_norm": 0.9404931664466858,
      "learning_rate": 2.6226112081028544e-07,
      "loss": 1.9349,
      "step": 11100
    },
    {
      "epoch": 0.9415720891130728,
      "grad_norm": 0.8282467722892761,
      "learning_rate": 2.0071303346540505e-07,
      "loss": 1.9868,
      "step": 11200
    },
    {
      "epoch": 0.949978982765868,
      "grad_norm": 1.1069015264511108,
      "learning_rate": 1.473121701795144e-07,
      "loss": 1.8612,
      "step": 11300
    },
    {
      "epoch": 0.9583858764186634,
      "grad_norm": 1.072883129119873,
      "learning_rate": 1.0210292895826423e-07,
      "loss": 1.8956,
      "step": 11400
    },
    {
      "epoch": 0.9667927700714586,
      "grad_norm": 1.38330078125,
      "learning_rate": 6.512289721174126e-08,
      "loss": 1.9686,
      "step": 11500
    },
    {
      "epoch": 0.9751996637242539,
      "grad_norm": 1.1821762323379517,
      "learning_rate": 3.640282050392241e-08,
      "loss": 1.9082,
      "step": 11600
    },
    {
      "epoch": 0.9836065573770492,
      "grad_norm": 1.2466627359390259,
      "learning_rate": 1.5966576990507078e-08,
      "loss": 1.896,
      "step": 11700
    },
    {
      "epoch": 0.9920134510298445,
      "grad_norm": 1.2905254364013672,
      "learning_rate": 3.83115756638297e-09,
      "loss": 1.9837,
      "step": 11800
    },
    {
      "epoch": 1.0,
      "step": 11895,
      "total_flos": 2.17890078425088e+17,
      "train_loss": 1.9851969016204616,
      "train_runtime": 3196.2319,
      "train_samples_per_second": 7.443,
      "train_steps_per_second": 3.722
    }
  ],
  "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": 2.17890078425088e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}