{
  "best_global_step": 5300,
  "best_metric": 0.846949212010463,
  "best_model_checkpoint": "./camelbert-ner-author/checkpoint-5300",
  "epoch": 4.0,
  "eval_steps": 100,
  "global_step": 8236,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04856726566294318,
      "grad_norm": 13.77788257598877,
      "learning_rate": 1.9759592034968435e-05,
      "loss": 2.5655,
      "step": 100
    },
    {
      "epoch": 0.04856726566294318,
      "eval_accuracy": 0.4197738753909069,
      "eval_f1_macro": 0.3205328980120634,
      "eval_f1_micro": 0.4197738753909069,
      "eval_loss": 2.164029121398926,
      "eval_precision_macro": 0.42432855880852344,
      "eval_precision_micro": 0.4197738753909069,
      "eval_recall_macro": 0.39236584110181216,
      "eval_recall_micro": 0.4197738753909069,
      "eval_runtime": 77.4428,
      "eval_samples_per_second": 53.678,
      "eval_steps_per_second": 3.357,
      "step": 100
    },
    {
      "epoch": 0.09713453132588636,
      "grad_norm": 11.284036636352539,
      "learning_rate": 1.9516755706653716e-05,
      "loss": 1.7809,
      "step": 200
    },
    {
      "epoch": 0.09713453132588636,
      "eval_accuracy": 0.5876834255472697,
      "eval_f1_macro": 0.5021998875928434,
      "eval_f1_micro": 0.5876834255472697,
      "eval_loss": 1.547103762626648,
      "eval_precision_macro": 0.5640550767871323,
      "eval_precision_micro": 0.5876834255472697,
      "eval_recall_macro": 0.5447230363092859,
      "eval_recall_micro": 0.5876834255472697,
      "eval_runtime": 77.3875,
      "eval_samples_per_second": 53.717,
      "eval_steps_per_second": 3.36,
      "step": 200
    },
    {
      "epoch": 0.14570179698882954,
      "grad_norm": 15.53082275390625,
      "learning_rate": 1.9273919378339e-05,
      "loss": 1.2903,
      "step": 300
    },
    {
      "epoch": 0.14570179698882954,
      "eval_accuracy": 0.6300216502285302,
      "eval_f1_macro": 0.5639062324413268,
      "eval_f1_micro": 0.6300216502285302,
      "eval_loss": 1.3453640937805176,
      "eval_precision_macro": 0.5883696452747289,
      "eval_precision_micro": 0.6300216502285302,
      "eval_recall_macro": 0.6245065052027649,
      "eval_recall_micro": 0.6300216502285302,
      "eval_runtime": 77.7169,
      "eval_samples_per_second": 53.489,
      "eval_steps_per_second": 3.345,
      "step": 300
    },
    {
      "epoch": 0.1942690626517727,
      "grad_norm": 12.855542182922363,
      "learning_rate": 1.9031083050024287e-05,
      "loss": 1.0303,
      "step": 400
    },
    {
      "epoch": 0.1942690626517727,
      "eval_accuracy": 0.6980995910512389,
      "eval_f1_macro": 0.621757249063289,
      "eval_f1_micro": 0.6980995910512389,
      "eval_loss": 1.0540422201156616,
      "eval_precision_macro": 0.6363865097952114,
      "eval_precision_micro": 0.6980995910512389,
      "eval_recall_macro": 0.6615766427269236,
      "eval_recall_micro": 0.6980995910512389,
      "eval_runtime": 78.209,
      "eval_samples_per_second": 53.152,
      "eval_steps_per_second": 3.324,
      "step": 400
    },
    {
      "epoch": 0.24283632831471588,
      "grad_norm": 13.974908828735352,
      "learning_rate": 1.8788246721709567e-05,
      "loss": 0.7461,
      "step": 500
    },
    {
      "epoch": 0.24283632831471588,
      "eval_accuracy": 0.697618474861679,
      "eval_f1_macro": 0.6189270686060082,
      "eval_f1_micro": 0.697618474861679,
      "eval_loss": 1.0767923593521118,
      "eval_precision_macro": 0.6453274802012455,
      "eval_precision_micro": 0.697618474861679,
      "eval_recall_macro": 0.6631935550298835,
      "eval_recall_micro": 0.697618474861679,
      "eval_runtime": 77.7991,
      "eval_samples_per_second": 53.433,
      "eval_steps_per_second": 3.342,
      "step": 500
    },
    {
      "epoch": 0.2914035939776591,
      "grad_norm": 15.981602668762207,
      "learning_rate": 1.8545410393394855e-05,
      "loss": 0.7715,
      "step": 600
    },
    {
      "epoch": 0.2914035939776591,
      "eval_accuracy": 0.6793360596584075,
      "eval_f1_macro": 0.6412879956515993,
      "eval_f1_micro": 0.6793360596584075,
      "eval_loss": 1.0772430896759033,
      "eval_precision_macro": 0.6604751299557832,
      "eval_precision_micro": 0.6793360596584075,
      "eval_recall_macro": 0.6882200466749144,
      "eval_recall_micro": 0.6793360596584075,
      "eval_runtime": 78.3813,
      "eval_samples_per_second": 53.036,
      "eval_steps_per_second": 3.317,
      "step": 600
    },
    {
      "epoch": 0.33997085964060225,
      "grad_norm": 15.210453033447266,
      "learning_rate": 1.8302574065080135e-05,
      "loss": 0.5764,
      "step": 700
    },
    {
      "epoch": 0.33997085964060225,
      "eval_accuracy": 0.7346644214577821,
      "eval_f1_macro": 0.6644040766448638,
      "eval_f1_micro": 0.7346644214577821,
      "eval_loss": 0.9163544178009033,
      "eval_precision_macro": 0.6814907504685033,
      "eval_precision_micro": 0.7346644214577821,
      "eval_recall_macro": 0.7012321944408646,
      "eval_recall_micro": 0.7346644214577821,
      "eval_runtime": 77.9074,
      "eval_samples_per_second": 53.358,
      "eval_steps_per_second": 3.337,
      "step": 700
    },
    {
      "epoch": 0.3885381253035454,
      "grad_norm": 33.91225051879883,
      "learning_rate": 1.8059737736765423e-05,
      "loss": 0.5638,
      "step": 800
    },
    {
      "epoch": 0.3885381253035454,
      "eval_accuracy": 0.7474140004811162,
      "eval_f1_macro": 0.6913277390868877,
      "eval_f1_micro": 0.7474140004811162,
      "eval_loss": 0.889366090297699,
      "eval_precision_macro": 0.7266313115490612,
      "eval_precision_micro": 0.7474140004811162,
      "eval_recall_macro": 0.715908295360388,
      "eval_recall_micro": 0.7474140004811162,
      "eval_runtime": 78.3481,
      "eval_samples_per_second": 53.058,
      "eval_steps_per_second": 3.319,
      "step": 800
    },
    {
      "epoch": 0.4371053909664886,
      "grad_norm": 8.672567367553711,
      "learning_rate": 1.7816901408450707e-05,
      "loss": 0.5197,
      "step": 900
    },
    {
      "epoch": 0.4371053909664886,
      "eval_accuracy": 0.7926389222997354,
      "eval_f1_macro": 0.7247243935175413,
      "eval_f1_micro": 0.7926389222997354,
      "eval_loss": 0.7225074768066406,
      "eval_precision_macro": 0.7255052815535598,
      "eval_precision_micro": 0.7926389222997354,
      "eval_recall_macro": 0.7576197995601733,
      "eval_recall_micro": 0.7926389222997354,
      "eval_runtime": 78.3587,
      "eval_samples_per_second": 53.051,
      "eval_steps_per_second": 3.318,
      "step": 900
    },
    {
      "epoch": 0.48567265662943176,
      "grad_norm": 5.37787389755249,
      "learning_rate": 1.757406508013599e-05,
      "loss": 0.4454,
      "step": 1000
    },
    {
      "epoch": 0.48567265662943176,
      "eval_accuracy": 0.7611258118835699,
      "eval_f1_macro": 0.6918902246005824,
      "eval_f1_micro": 0.7611258118835699,
      "eval_loss": 0.8200820684432983,
      "eval_precision_macro": 0.7248507826718404,
      "eval_precision_micro": 0.7611258118835699,
      "eval_recall_macro": 0.7240661179940105,
      "eval_recall_micro": 0.7611258118835699,
      "eval_runtime": 77.9125,
      "eval_samples_per_second": 53.355,
      "eval_steps_per_second": 3.337,
      "step": 1000
    },
    {
      "epoch": 0.5342399222923749,
      "grad_norm": 50.15217971801758,
      "learning_rate": 1.7331228751821275e-05,
      "loss": 0.3573,
      "step": 1100
    },
    {
      "epoch": 0.5342399222923749,
      "eval_accuracy": 0.7486167909550157,
      "eval_f1_macro": 0.708488926204985,
      "eval_f1_micro": 0.7486167909550158,
      "eval_loss": 0.840596616268158,
      "eval_precision_macro": 0.7458516807418488,
      "eval_precision_micro": 0.7486167909550157,
      "eval_recall_macro": 0.7355897302661839,
      "eval_recall_micro": 0.7486167909550157,
      "eval_runtime": 77.9539,
      "eval_samples_per_second": 53.326,
      "eval_steps_per_second": 3.335,
      "step": 1100
    },
    {
      "epoch": 0.5828071879553182,
      "grad_norm": 4.126067161560059,
      "learning_rate": 1.708839242350656e-05,
      "loss": 0.3992,
      "step": 1200
    },
    {
      "epoch": 0.5828071879553182,
      "eval_accuracy": 0.7969689680057734,
      "eval_f1_macro": 0.7545628662653078,
      "eval_f1_micro": 0.7969689680057734,
      "eval_loss": 0.7257721424102783,
      "eval_precision_macro": 0.7815422636293521,
      "eval_precision_micro": 0.7969689680057734,
      "eval_recall_macro": 0.76812907504405,
      "eval_recall_micro": 0.7969689680057734,
      "eval_runtime": 78.4205,
      "eval_samples_per_second": 53.009,
      "eval_steps_per_second": 3.315,
      "step": 1200
    },
    {
      "epoch": 0.6313744536182613,
      "grad_norm": 9.667169570922852,
      "learning_rate": 1.6845556095191843e-05,
      "loss": 0.3637,
      "step": 1300
    },
    {
      "epoch": 0.6313744536182613,
      "eval_accuracy": 0.8181380803464037,
      "eval_f1_macro": 0.7552426483349393,
      "eval_f1_micro": 0.8181380803464037,
      "eval_loss": 0.6487129330635071,
      "eval_precision_macro": 0.770444338655661,
      "eval_precision_micro": 0.8181380803464037,
      "eval_recall_macro": 0.7814605757463501,
      "eval_recall_micro": 0.8181380803464037,
      "eval_runtime": 77.6598,
      "eval_samples_per_second": 53.528,
      "eval_steps_per_second": 3.348,
      "step": 1300
    },
    {
      "epoch": 0.6799417192812045,
      "grad_norm": 15.598691940307617,
      "learning_rate": 1.6602719766877127e-05,
      "loss": 0.3534,
      "step": 1400
    },
    {
      "epoch": 0.6799417192812045,
      "eval_accuracy": 0.8320904498436372,
      "eval_f1_macro": 0.7756514292421667,
      "eval_f1_micro": 0.8320904498436372,
      "eval_loss": 0.6163076162338257,
      "eval_precision_macro": 0.776765552199579,
      "eval_precision_micro": 0.8320904498436372,
      "eval_recall_macro": 0.807685789299095,
      "eval_recall_micro": 0.8320904498436372,
      "eval_runtime": 78.0973,
      "eval_samples_per_second": 53.228,
      "eval_steps_per_second": 3.329,
      "step": 1400
    },
    {
      "epoch": 0.7285089849441476,
      "grad_norm": 16.200693130493164,
      "learning_rate": 1.635988343856241e-05,
      "loss": 0.2738,
      "step": 1500
    },
    {
      "epoch": 0.7285089849441476,
      "eval_accuracy": 0.8330526822227567,
      "eval_f1_macro": 0.7657709635036466,
      "eval_f1_micro": 0.8330526822227567,
      "eval_loss": 0.5991656184196472,
      "eval_precision_macro": 0.7896026550833485,
      "eval_precision_micro": 0.8330526822227567,
      "eval_recall_macro": 0.7825126366614965,
      "eval_recall_micro": 0.8330526822227567,
      "eval_runtime": 78.1674,
      "eval_samples_per_second": 53.181,
      "eval_steps_per_second": 3.326,
      "step": 1500
    },
    {
      "epoch": 0.7770762506070908,
      "grad_norm": 22.443140029907227,
      "learning_rate": 1.6117047110247694e-05,
      "loss": 0.2976,
      "step": 1600
    },
    {
      "epoch": 0.7770762506070908,
      "eval_accuracy": 0.8277604041375992,
      "eval_f1_macro": 0.7434375371463181,
      "eval_f1_micro": 0.8277604041375993,
      "eval_loss": 0.619115948677063,
      "eval_precision_macro": 0.7596164331235828,
      "eval_precision_micro": 0.8277604041375992,
      "eval_recall_macro": 0.7832912830659067,
      "eval_recall_micro": 0.8277604041375992,
      "eval_runtime": 77.7196,
      "eval_samples_per_second": 53.487,
      "eval_steps_per_second": 3.345,
      "step": 1600
    },
    {
      "epoch": 0.825643516270034,
      "grad_norm": 17.786861419677734,
      "learning_rate": 1.587421078193298e-05,
      "loss": 0.2914,
      "step": 1700
    },
    {
      "epoch": 0.825643516270034,
      "eval_accuracy": 0.7873466442145778,
      "eval_f1_macro": 0.7264208752916804,
      "eval_f1_micro": 0.7873466442145777,
      "eval_loss": 0.7651363015174866,
      "eval_precision_macro": 0.7408201824968061,
      "eval_precision_micro": 0.7873466442145778,
      "eval_recall_macro": 0.7604117672744211,
      "eval_recall_micro": 0.7873466442145778,
      "eval_runtime": 77.8533,
      "eval_samples_per_second": 53.395,
      "eval_steps_per_second": 3.34,
      "step": 1700
    },
    {
      "epoch": 0.8742107819329772,
      "grad_norm": 15.827679634094238,
      "learning_rate": 1.5631374453618262e-05,
      "loss": 0.254,
      "step": 1800
    },
    {
      "epoch": 0.8742107819329772,
      "eval_accuracy": 0.7729131585277844,
      "eval_f1_macro": 0.7078131643612664,
      "eval_f1_micro": 0.7729131585277844,
      "eval_loss": 0.8441910743713379,
      "eval_precision_macro": 0.7184265914939797,
      "eval_precision_micro": 0.7729131585277844,
      "eval_recall_macro": 0.7567655915669744,
      "eval_recall_micro": 0.7729131585277844,
      "eval_runtime": 78.4035,
      "eval_samples_per_second": 53.021,
      "eval_steps_per_second": 3.316,
      "step": 1800
    },
    {
      "epoch": 0.9227780475959203,
      "grad_norm": 13.433658599853516,
      "learning_rate": 1.5388538125303546e-05,
      "loss": 0.2446,
      "step": 1900
    },
    {
      "epoch": 0.9227780475959203,
      "eval_accuracy": 0.8330526822227567,
      "eval_f1_macro": 0.7754170140757469,
      "eval_f1_micro": 0.8330526822227567,
      "eval_loss": 0.6231703162193298,
      "eval_precision_macro": 0.7743561402393088,
      "eval_precision_micro": 0.8330526822227567,
      "eval_recall_macro": 0.8166841562149844,
      "eval_recall_micro": 0.8330526822227567,
      "eval_runtime": 77.9333,
      "eval_samples_per_second": 53.341,
      "eval_steps_per_second": 3.336,
      "step": 1900
    },
    {
      "epoch": 0.9713453132588635,
      "grad_norm": 1.4447236061096191,
      "learning_rate": 1.514570179698883e-05,
      "loss": 0.2468,
      "step": 2000
    },
    {
      "epoch": 0.9713453132588635,
      "eval_accuracy": 0.8448400288669714,
      "eval_f1_macro": 0.785930557684152,
      "eval_f1_micro": 0.8448400288669714,
      "eval_loss": 0.5916985869407654,
      "eval_precision_macro": 0.779867107822518,
      "eval_precision_micro": 0.8448400288669714,
      "eval_recall_macro": 0.8258044114789991,
      "eval_recall_micro": 0.8448400288669714,
      "eval_runtime": 78.2704,
      "eval_samples_per_second": 53.111,
      "eval_steps_per_second": 3.322,
      "step": 2000
    },
    {
      "epoch": 1.0199125789218066,
      "grad_norm": 42.209495544433594,
      "learning_rate": 1.4902865468674114e-05,
      "loss": 0.2197,
      "step": 2100
    },
    {
      "epoch": 1.0199125789218066,
      "eval_accuracy": 0.8210247774837623,
      "eval_f1_macro": 0.7496958773425116,
      "eval_f1_micro": 0.8210247774837625,
      "eval_loss": 0.6304962038993835,
      "eval_precision_macro": 0.7800275333894653,
      "eval_precision_micro": 0.8210247774837623,
      "eval_recall_macro": 0.7725863519296959,
      "eval_recall_micro": 0.8210247774837623,
      "eval_runtime": 78.4172,
      "eval_samples_per_second": 53.011,
      "eval_steps_per_second": 3.316,
      "step": 2100
    },
    {
      "epoch": 1.0684798445847499,
      "grad_norm": 6.323939800262451,
      "learning_rate": 1.46600291403594e-05,
      "loss": 0.1474,
      "step": 2200
    },
    {
      "epoch": 1.0684798445847499,
      "eval_accuracy": 0.8210247774837623,
      "eval_f1_macro": 0.7654970298380952,
      "eval_f1_micro": 0.8210247774837625,
      "eval_loss": 0.6584117412567139,
      "eval_precision_macro": 0.7776774908498755,
      "eval_precision_micro": 0.8210247774837623,
      "eval_recall_macro": 0.7832612508206293,
      "eval_recall_micro": 0.8210247774837623,
      "eval_runtime": 78.2294,
      "eval_samples_per_second": 53.139,
      "eval_steps_per_second": 3.324,
      "step": 2200
    },
    {
      "epoch": 1.117047110247693,
      "grad_norm": 5.200247287750244,
      "learning_rate": 1.4417192812044682e-05,
      "loss": 0.1385,
      "step": 2300
    },
    {
      "epoch": 1.117047110247693,
      "eval_accuracy": 0.8465239355304306,
      "eval_f1_macro": 0.795171009913394,
      "eval_f1_micro": 0.8465239355304306,
      "eval_loss": 0.5680557489395142,
      "eval_precision_macro": 0.8036743010238906,
      "eval_precision_micro": 0.8465239355304306,
      "eval_recall_macro": 0.8194950943386284,
      "eval_recall_micro": 0.8465239355304306,
      "eval_runtime": 77.9399,
      "eval_samples_per_second": 53.336,
      "eval_steps_per_second": 3.336,
      "step": 2300
    },
    {
      "epoch": 1.1656143759106363,
      "grad_norm": 59.158382415771484,
      "learning_rate": 1.4174356483729968e-05,
      "loss": 0.1135,
      "step": 2400
    },
    {
      "epoch": 1.1656143759106363,
      "eval_accuracy": 0.867452489776281,
      "eval_f1_macro": 0.8221801724207984,
      "eval_f1_micro": 0.867452489776281,
      "eval_loss": 0.5212485194206238,
      "eval_precision_macro": 0.8331447468612765,
      "eval_precision_micro": 0.867452489776281,
      "eval_recall_macro": 0.8320781071525781,
      "eval_recall_micro": 0.867452489776281,
      "eval_runtime": 78.3816,
      "eval_samples_per_second": 53.035,
      "eval_steps_per_second": 3.317,
      "step": 2400
    },
    {
      "epoch": 1.2141816415735793,
      "grad_norm": 1.2092406749725342,
      "learning_rate": 1.393152015541525e-05,
      "loss": 0.153,
      "step": 2500
    },
    {
      "epoch": 1.2141816415735793,
      "eval_accuracy": 0.854462352658167,
      "eval_f1_macro": 0.79162479913702,
      "eval_f1_micro": 0.854462352658167,
      "eval_loss": 0.5905571579933167,
      "eval_precision_macro": 0.8007086193727224,
      "eval_precision_micro": 0.854462352658167,
      "eval_recall_macro": 0.8240486776141968,
      "eval_recall_micro": 0.854462352658167,
      "eval_runtime": 77.8377,
      "eval_samples_per_second": 53.406,
      "eval_steps_per_second": 3.34,
      "step": 2500
    },
    {
      "epoch": 1.2627489072365226,
      "grad_norm": 4.295379638671875,
      "learning_rate": 1.3688683827100536e-05,
      "loss": 0.1221,
      "step": 2600
    },
    {
      "epoch": 1.2627489072365226,
      "eval_accuracy": 0.8780370459465962,
      "eval_f1_macro": 0.8233792046483023,
      "eval_f1_micro": 0.8780370459465962,
      "eval_loss": 0.48117873072624207,
      "eval_precision_macro": 0.8275579818555562,
      "eval_precision_micro": 0.8780370459465962,
      "eval_recall_macro": 0.8399965589517642,
      "eval_recall_micro": 0.8780370459465962,
      "eval_runtime": 78.2827,
      "eval_samples_per_second": 53.102,
      "eval_steps_per_second": 3.321,
      "step": 2600
    },
    {
      "epoch": 1.3113161728994658,
      "grad_norm": 0.7493284940719604,
      "learning_rate": 1.344584749878582e-05,
      "loss": 0.134,
      "step": 2700
    },
    {
      "epoch": 1.3113161728994658,
      "eval_accuracy": 0.8525378878999278,
      "eval_f1_macro": 0.8105294018236757,
      "eval_f1_micro": 0.8525378878999278,
      "eval_loss": 0.6301819682121277,
      "eval_precision_macro": 0.8127798645314742,
      "eval_precision_micro": 0.8525378878999278,
      "eval_recall_macro": 0.8407694300980793,
      "eval_recall_micro": 0.8525378878999278,
      "eval_runtime": 78.1792,
      "eval_samples_per_second": 53.173,
      "eval_steps_per_second": 3.326,
      "step": 2700
    },
    {
      "epoch": 1.359883438562409,
      "grad_norm": 15.269425392150879,
      "learning_rate": 1.3203011170471102e-05,
      "loss": 0.1177,
      "step": 2800
    },
    {
      "epoch": 1.359883438562409,
      "eval_accuracy": 0.8720230935770988,
      "eval_f1_macro": 0.8162750069957628,
      "eval_f1_micro": 0.872023093577099,
      "eval_loss": 0.5413680672645569,
      "eval_precision_macro": 0.8190184865925261,
      "eval_precision_micro": 0.8720230935770988,
      "eval_recall_macro": 0.8430994278929198,
      "eval_recall_micro": 0.8720230935770988,
      "eval_runtime": 77.9146,
      "eval_samples_per_second": 53.353,
      "eval_steps_per_second": 3.337,
      "step": 2800
    },
    {
      "epoch": 1.408450704225352,
      "grad_norm": 2.0257482528686523,
      "learning_rate": 1.2960174842156388e-05,
      "loss": 0.0987,
      "step": 2900
    },
    {
      "epoch": 1.408450704225352,
      "eval_accuracy": 0.8809237430839548,
      "eval_f1_macro": 0.8218294114431427,
      "eval_f1_micro": 0.8809237430839548,
      "eval_loss": 0.5116602182388306,
      "eval_precision_macro": 0.8136201153530077,
      "eval_precision_micro": 0.8809237430839548,
      "eval_recall_macro": 0.8508023667833765,
      "eval_recall_micro": 0.8809237430839548,
      "eval_runtime": 78.002,
      "eval_samples_per_second": 53.294,
      "eval_steps_per_second": 3.333,
      "step": 2900
    },
    {
      "epoch": 1.4570179698882952,
      "grad_norm": 42.31605529785156,
      "learning_rate": 1.2717338513841673e-05,
      "loss": 0.1058,
      "step": 3000
    },
    {
      "epoch": 1.4570179698882952,
      "eval_accuracy": 0.8433966802982921,
      "eval_f1_macro": 0.8068375038460742,
      "eval_f1_micro": 0.8433966802982921,
      "eval_loss": 0.6526976823806763,
      "eval_precision_macro": 0.8158791028657807,
      "eval_precision_micro": 0.8433966802982921,
      "eval_recall_macro": 0.8236984397615126,
      "eval_recall_micro": 0.8433966802982921,
      "eval_runtime": 78.2408,
      "eval_samples_per_second": 53.131,
      "eval_steps_per_second": 3.323,
      "step": 3000
    },
    {
      "epoch": 1.5055852355512385,
      "grad_norm": 0.04906897246837616,
      "learning_rate": 1.2474502185526956e-05,
      "loss": 0.0847,
      "step": 3100
    },
    {
      "epoch": 1.5055852355512385,
      "eval_accuracy": 0.8554245850372865,
      "eval_f1_macro": 0.798783246253302,
      "eval_f1_micro": 0.8554245850372865,
      "eval_loss": 0.5800924301147461,
      "eval_precision_macro": 0.8004222933889796,
      "eval_precision_micro": 0.8554245850372865,
      "eval_recall_macro": 0.8243767820708686,
      "eval_recall_micro": 0.8554245850372865,
      "eval_runtime": 77.9481,
      "eval_samples_per_second": 53.33,
      "eval_steps_per_second": 3.336,
      "step": 3100
    },
    {
      "epoch": 1.5541525012141817,
      "grad_norm": 0.13342243432998657,
      "learning_rate": 1.223166585721224e-05,
      "loss": 0.0934,
      "step": 3200
    },
    {
      "epoch": 1.5541525012141817,
      "eval_accuracy": 0.8433966802982921,
      "eval_f1_macro": 0.7975033136779321,
      "eval_f1_micro": 0.8433966802982921,
      "eval_loss": 0.6552587151527405,
      "eval_precision_macro": 0.7929638472956879,
      "eval_precision_micro": 0.8433966802982921,
      "eval_recall_macro": 0.8272082545582277,
      "eval_recall_micro": 0.8433966802982921,
      "eval_runtime": 78.5124,
      "eval_samples_per_second": 52.947,
      "eval_steps_per_second": 3.312,
      "step": 3200
    },
    {
      "epoch": 1.602719766877125,
      "grad_norm": 0.7083119750022888,
      "learning_rate": 1.1988829528897524e-05,
      "loss": 0.0969,
      "step": 3300
    },
    {
      "epoch": 1.602719766877125,
      "eval_accuracy": 0.82607649747414,
      "eval_f1_macro": 0.7832101352133586,
      "eval_f1_micro": 0.82607649747414,
      "eval_loss": 0.7666494250297546,
      "eval_precision_macro": 0.7876268324182853,
      "eval_precision_micro": 0.82607649747414,
      "eval_recall_macro": 0.8241865106118899,
      "eval_recall_micro": 0.82607649747414,
      "eval_runtime": 78.311,
      "eval_samples_per_second": 53.083,
      "eval_steps_per_second": 3.32,
      "step": 3300
    },
    {
      "epoch": 1.651287032540068,
      "grad_norm": 16.786699295043945,
      "learning_rate": 1.1745993200582808e-05,
      "loss": 0.1428,
      "step": 3400
    },
    {
      "epoch": 1.651287032540068,
      "eval_accuracy": 0.8462833774356507,
      "eval_f1_macro": 0.7894688554105801,
      "eval_f1_micro": 0.8462833774356507,
      "eval_loss": 0.685663104057312,
      "eval_precision_macro": 0.7989143890731024,
      "eval_precision_micro": 0.8462833774356507,
      "eval_recall_macro": 0.8135123764257445,
      "eval_recall_micro": 0.8462833774356507,
      "eval_runtime": 78.0482,
      "eval_samples_per_second": 53.262,
      "eval_steps_per_second": 3.331,
      "step": 3400
    },
    {
      "epoch": 1.6998542982030111,
      "grad_norm": 27.461700439453125,
      "learning_rate": 1.1503156872268093e-05,
      "loss": 0.0681,
      "step": 3500
    },
    {
      "epoch": 1.6998542982030111,
      "eval_accuracy": 0.8898243925908107,
      "eval_f1_macro": 0.8280214132166845,
      "eval_f1_micro": 0.8898243925908107,
      "eval_loss": 0.5203650593757629,
      "eval_precision_macro": 0.8341272420720794,
      "eval_precision_micro": 0.8898243925908107,
      "eval_recall_macro": 0.8595091501105108,
      "eval_recall_micro": 0.8898243925908107,
      "eval_runtime": 78.1461,
      "eval_samples_per_second": 53.195,
      "eval_steps_per_second": 3.327,
      "step": 3500
    },
    {
      "epoch": 1.7484215638659544,
      "grad_norm": 69.4913101196289,
      "learning_rate": 1.1260320543953375e-05,
      "loss": 0.0931,
      "step": 3600
    },
    {
      "epoch": 1.7484215638659544,
      "eval_accuracy": 0.8684147221554005,
      "eval_f1_macro": 0.8151692727718802,
      "eval_f1_micro": 0.8684147221554005,
      "eval_loss": 0.6120151281356812,
      "eval_precision_macro": 0.8240793067449873,
      "eval_precision_micro": 0.8684147221554005,
      "eval_recall_macro": 0.8314024627761833,
      "eval_recall_micro": 0.8684147221554005,
      "eval_runtime": 78.1922,
      "eval_samples_per_second": 53.164,
      "eval_steps_per_second": 3.325,
      "step": 3600
    },
    {
      "epoch": 1.7969888295288974,
      "grad_norm": 0.8026056885719299,
      "learning_rate": 1.1017484215638661e-05,
      "loss": 0.0935,
      "step": 3700
    },
    {
      "epoch": 1.7969888295288974,
      "eval_accuracy": 0.8715419773875391,
      "eval_f1_macro": 0.8220476579310915,
      "eval_f1_micro": 0.8715419773875391,
      "eval_loss": 0.5801733732223511,
      "eval_precision_macro": 0.8170628217436031,
      "eval_precision_micro": 0.8715419773875391,
      "eval_recall_macro": 0.8414628001544893,
      "eval_recall_micro": 0.8715419773875391,
      "eval_runtime": 77.92,
      "eval_samples_per_second": 53.35,
      "eval_steps_per_second": 3.337,
      "step": 3700
    },
    {
      "epoch": 1.8455560951918408,
      "grad_norm": 20.50851058959961,
      "learning_rate": 1.0774647887323943e-05,
      "loss": 0.1051,
      "step": 3800
    },
    {
      "epoch": 1.8455560951918408,
      "eval_accuracy": 0.8638441183545826,
      "eval_f1_macro": 0.8138781743364819,
      "eval_f1_micro": 0.8638441183545826,
      "eval_loss": 0.6387995481491089,
      "eval_precision_macro": 0.8120807002790287,
      "eval_precision_micro": 0.8638441183545826,
      "eval_recall_macro": 0.8384780204857788,
      "eval_recall_micro": 0.8638441183545826,
      "eval_runtime": 78.3578,
      "eval_samples_per_second": 53.052,
      "eval_steps_per_second": 3.318,
      "step": 3800
    },
    {
      "epoch": 1.8941233608547838,
      "grad_norm": 0.4359864890575409,
      "learning_rate": 1.0531811559009229e-05,
      "loss": 0.1141,
      "step": 3900
    },
    {
      "epoch": 1.8941233608547838,
      "eval_accuracy": 0.8734664421457782,
      "eval_f1_macro": 0.8181325108037866,
      "eval_f1_micro": 0.8734664421457783,
      "eval_loss": 0.6033037304878235,
      "eval_precision_macro": 0.8257092361144394,
      "eval_precision_micro": 0.8734664421457782,
      "eval_recall_macro": 0.8280682968840635,
      "eval_recall_micro": 0.8734664421457782,
      "eval_runtime": 78.3807,
      "eval_samples_per_second": 53.036,
      "eval_steps_per_second": 3.317,
      "step": 3900
    },
    {
      "epoch": 1.942690626517727,
      "grad_norm": 0.026569893583655357,
      "learning_rate": 1.0288975230694513e-05,
      "loss": 0.0782,
      "step": 4000
    },
    {
      "epoch": 1.942690626517727,
      "eval_accuracy": 0.8713014192927592,
      "eval_f1_macro": 0.8158890342266224,
      "eval_f1_micro": 0.8713014192927592,
      "eval_loss": 0.6082781553268433,
      "eval_precision_macro": 0.8100721903094688,
      "eval_precision_micro": 0.8713014192927592,
      "eval_recall_macro": 0.8378298843929596,
      "eval_recall_micro": 0.8713014192927592,
      "eval_runtime": 77.7827,
      "eval_samples_per_second": 53.444,
      "eval_steps_per_second": 3.343,
      "step": 4000
    },
    {
      "epoch": 1.9912578921806703,
      "grad_norm": 0.27337446808815,
      "learning_rate": 1.0046138902379797e-05,
      "loss": 0.0935,
      "step": 4100
    },
    {
      "epoch": 1.9912578921806703,
      "eval_accuracy": 0.8385855184026942,
      "eval_f1_macro": 0.7755030296607576,
      "eval_f1_micro": 0.8385855184026941,
      "eval_loss": 0.7882886528968811,
      "eval_precision_macro": 0.772528897611925,
      "eval_precision_micro": 0.8385855184026942,
      "eval_recall_macro": 0.821396700276969,
      "eval_recall_micro": 0.8385855184026942,
      "eval_runtime": 78.0832,
      "eval_samples_per_second": 53.238,
      "eval_steps_per_second": 3.33,
      "step": 4100
    },
    {
      "epoch": 2.0398251578436133,
      "grad_norm": 0.06082015484571457,
      "learning_rate": 9.803302574065081e-06,
      "loss": 0.0446,
      "step": 4200
    },
    {
      "epoch": 2.0398251578436133,
      "eval_accuracy": 0.8462833774356507,
      "eval_f1_macro": 0.793800305134788,
      "eval_f1_micro": 0.8462833774356507,
      "eval_loss": 0.7242459654808044,
      "eval_precision_macro": 0.8042449609141427,
      "eval_precision_micro": 0.8462833774356507,
      "eval_recall_macro": 0.8070235910631676,
      "eval_recall_micro": 0.8462833774356507,
      "eval_runtime": 78.366,
      "eval_samples_per_second": 53.046,
      "eval_steps_per_second": 3.318,
      "step": 4200
    },
    {
      "epoch": 2.0883924235065567,
      "grad_norm": 1.1284111738204956,
      "learning_rate": 9.560466245750365e-06,
      "loss": 0.0394,
      "step": 4300
    },
    {
      "epoch": 2.0883924235065567,
      "eval_accuracy": 0.8573490497955256,
      "eval_f1_macro": 0.8097667721853947,
      "eval_f1_micro": 0.8573490497955256,
      "eval_loss": 0.6780639886856079,
      "eval_precision_macro": 0.8206783431890832,
      "eval_precision_micro": 0.8573490497955256,
      "eval_recall_macro": 0.8218729335160754,
      "eval_recall_micro": 0.8573490497955256,
      "eval_runtime": 78.4567,
      "eval_samples_per_second": 52.985,
      "eval_steps_per_second": 3.314,
      "step": 4300
    },
    {
      "epoch": 2.1369596891694997,
      "grad_norm": 1.0478459596633911,
      "learning_rate": 9.317629917435649e-06,
      "loss": 0.023,
      "step": 4400
    },
    {
      "epoch": 2.1369596891694997,
      "eval_accuracy": 0.8381044022131344,
      "eval_f1_macro": 0.7761721486362632,
      "eval_f1_micro": 0.8381044022131344,
      "eval_loss": 0.8235397934913635,
      "eval_precision_macro": 0.8022053549256318,
      "eval_precision_micro": 0.8381044022131344,
      "eval_recall_macro": 0.8109538575923662,
      "eval_recall_micro": 0.8381044022131344,
      "eval_runtime": 77.944,
      "eval_samples_per_second": 53.333,
      "eval_steps_per_second": 3.336,
      "step": 4400
    },
    {
      "epoch": 2.1855269548324427,
      "grad_norm": 0.0286689605563879,
      "learning_rate": 9.074793589120935e-06,
      "loss": 0.0338,
      "step": 4500
    },
    {
      "epoch": 2.1855269548324427,
      "eval_accuracy": 0.8703391869136397,
      "eval_f1_macro": 0.8094122567680977,
      "eval_f1_micro": 0.8703391869136397,
      "eval_loss": 0.6315681338310242,
      "eval_precision_macro": 0.8074299681427781,
      "eval_precision_micro": 0.8703391869136397,
      "eval_recall_macro": 0.8337866144492019,
      "eval_recall_micro": 0.8703391869136397,
      "eval_runtime": 78.3498,
      "eval_samples_per_second": 53.057,
      "eval_steps_per_second": 3.318,
      "step": 4500
    },
    {
      "epoch": 2.234094220495386,
      "grad_norm": 0.06468763947486877,
      "learning_rate": 8.831957260806217e-06,
      "loss": 0.0414,
      "step": 4600
    },
    {
      "epoch": 2.234094220495386,
      "eval_accuracy": 0.8840509983160934,
      "eval_f1_macro": 0.824537485440517,
      "eval_f1_micro": 0.8840509983160935,
      "eval_loss": 0.5948061347007751,
      "eval_precision_macro": 0.8233324585853661,
      "eval_precision_micro": 0.8840509983160934,
      "eval_recall_macro": 0.8505518969806567,
      "eval_recall_micro": 0.8840509983160934,
      "eval_runtime": 77.9733,
      "eval_samples_per_second": 53.313,
      "eval_steps_per_second": 3.334,
      "step": 4600
    },
    {
      "epoch": 2.282661486158329,
      "grad_norm": 4.641617298126221,
      "learning_rate": 8.5891209324915e-06,
      "loss": 0.0379,
      "step": 4700
    },
    {
      "epoch": 2.282661486158329,
      "eval_accuracy": 0.8765936973779168,
      "eval_f1_macro": 0.825230740105461,
      "eval_f1_micro": 0.8765936973779168,
      "eval_loss": 0.5780928134918213,
      "eval_precision_macro": 0.8206894619069279,
      "eval_precision_micro": 0.8765936973779168,
      "eval_recall_macro": 0.841894179116351,
      "eval_recall_micro": 0.8765936973779168,
      "eval_runtime": 78.4226,
      "eval_samples_per_second": 53.008,
      "eval_steps_per_second": 3.315,
      "step": 4700
    },
    {
      "epoch": 2.3312287518212726,
      "grad_norm": 0.23332637548446655,
      "learning_rate": 8.346284604176785e-06,
      "loss": 0.0473,
      "step": 4800
    },
    {
      "epoch": 2.3312287518212726,
      "eval_accuracy": 0.8729853259562185,
      "eval_f1_macro": 0.8052063570572925,
      "eval_f1_micro": 0.8729853259562185,
      "eval_loss": 0.6347105503082275,
      "eval_precision_macro": 0.8294626941554183,
      "eval_precision_micro": 0.8729853259562185,
      "eval_recall_macro": 0.8174296733333687,
      "eval_recall_micro": 0.8729853259562185,
      "eval_runtime": 77.9725,
      "eval_samples_per_second": 53.314,
      "eval_steps_per_second": 3.335,
      "step": 4800
    },
    {
      "epoch": 2.3797960174842157,
      "grad_norm": 0.014487605541944504,
      "learning_rate": 8.103448275862069e-06,
      "loss": 0.0159,
      "step": 4900
    },
    {
      "epoch": 2.3797960174842157,
      "eval_accuracy": 0.8681741640606206,
      "eval_f1_macro": 0.8094189525603713,
      "eval_f1_micro": 0.8681741640606206,
      "eval_loss": 0.6882294416427612,
      "eval_precision_macro": 0.8262511053264366,
      "eval_precision_micro": 0.8681741640606206,
      "eval_recall_macro": 0.824772348858174,
      "eval_recall_micro": 0.8681741640606206,
      "eval_runtime": 78.4876,
      "eval_samples_per_second": 52.964,
      "eval_steps_per_second": 3.313,
      "step": 4900
    },
    {
      "epoch": 2.4283632831471587,
      "grad_norm": 0.011654403060674667,
      "learning_rate": 7.860611947547354e-06,
      "loss": 0.0115,
      "step": 5000
    },
    {
      "epoch": 2.4283632831471587,
      "eval_accuracy": 0.8830887659369738,
      "eval_f1_macro": 0.8288533499086042,
      "eval_f1_micro": 0.8830887659369738,
      "eval_loss": 0.5934479832649231,
      "eval_precision_macro": 0.8396603529798649,
      "eval_precision_micro": 0.8830887659369738,
      "eval_recall_macro": 0.8453069551945437,
      "eval_recall_micro": 0.8830887659369738,
      "eval_runtime": 78.5682,
      "eval_samples_per_second": 52.909,
      "eval_steps_per_second": 3.309,
      "step": 5000
    },
    {
      "epoch": 2.476930548810102,
      "grad_norm": 21.12113380432129,
      "learning_rate": 7.617775619232638e-06,
      "loss": 0.0494,
      "step": 5100
    },
    {
      "epoch": 2.476930548810102,
      "eval_accuracy": 0.8506134231416888,
      "eval_f1_macro": 0.7896669170968705,
      "eval_f1_micro": 0.8506134231416888,
      "eval_loss": 0.732039749622345,
      "eval_precision_macro": 0.7898243286153485,
      "eval_precision_micro": 0.8506134231416888,
      "eval_recall_macro": 0.8132701383447537,
      "eval_recall_micro": 0.8506134231416888,
      "eval_runtime": 78.0419,
      "eval_samples_per_second": 53.266,
      "eval_steps_per_second": 3.332,
      "step": 5100
    },
    {
      "epoch": 2.525497814473045,
      "grad_norm": 0.14766599237918854,
      "learning_rate": 7.3749392909179215e-06,
      "loss": 0.0318,
      "step": 5200
    },
    {
      "epoch": 2.525497814473045,
      "eval_accuracy": 0.8491700745730094,
      "eval_f1_macro": 0.8013080246715508,
      "eval_f1_micro": 0.8491700745730094,
      "eval_loss": 0.775058925151825,
      "eval_precision_macro": 0.8191511470633912,
      "eval_precision_micro": 0.8491700745730094,
      "eval_recall_macro": 0.8104006475463327,
      "eval_recall_micro": 0.8491700745730094,
      "eval_runtime": 78.0698,
      "eval_samples_per_second": 53.247,
      "eval_steps_per_second": 3.33,
      "step": 5200
    },
    {
      "epoch": 2.5740650801359886,
      "grad_norm": 0.01593843102455139,
      "learning_rate": 7.1321029626032054e-06,
      "loss": 0.0493,
      "step": 5300
    },
    {
      "epoch": 2.5740650801359886,
      "eval_accuracy": 0.8907866249699302,
      "eval_f1_macro": 0.846949212010463,
      "eval_f1_micro": 0.8907866249699302,
      "eval_loss": 0.5424800515174866,
      "eval_precision_macro": 0.8565661804143393,
      "eval_precision_micro": 0.8907866249699302,
      "eval_recall_macro": 0.852065989778345,
      "eval_recall_micro": 0.8907866249699302,
      "eval_runtime": 78.4856,
      "eval_samples_per_second": 52.965,
      "eval_steps_per_second": 3.313,
      "step": 5300
    },
    {
      "epoch": 2.6226323457989316,
      "grad_norm": 39.92009353637695,
      "learning_rate": 6.88926663428849e-06,
      "loss": 0.0265,
      "step": 5400
    },
    {
      "epoch": 2.6226323457989316,
      "eval_accuracy": 0.8811643011787347,
      "eval_f1_macro": 0.8343599063711409,
      "eval_f1_micro": 0.8811643011787347,
      "eval_loss": 0.5994859933853149,
      "eval_precision_macro": 0.8496569155300632,
      "eval_precision_micro": 0.8811643011787347,
      "eval_recall_macro": 0.8421658331646679,
      "eval_recall_micro": 0.8811643011787347,
      "eval_runtime": 78.0929,
      "eval_samples_per_second": 53.231,
      "eval_steps_per_second": 3.329,
      "step": 5400
    },
    {
      "epoch": 2.6711996114618746,
      "grad_norm": 0.05156758427619934,
      "learning_rate": 6.646430305973774e-06,
      "loss": 0.0644,
      "step": 5500
    },
    {
      "epoch": 2.6711996114618746,
      "eval_accuracy": 0.8898243925908107,
      "eval_f1_macro": 0.8456373260763799,
      "eval_f1_micro": 0.8898243925908107,
      "eval_loss": 0.5272171497344971,
      "eval_precision_macro": 0.8489836763758699,
      "eval_precision_micro": 0.8898243925908107,
      "eval_recall_macro": 0.8557074629075649,
      "eval_recall_micro": 0.8898243925908107,
      "eval_runtime": 78.4213,
      "eval_samples_per_second": 53.009,
      "eval_steps_per_second": 3.315,
      "step": 5500
    },
    {
      "epoch": 2.719766877124818,
      "grad_norm": 0.8775114417076111,
      "learning_rate": 6.403593977659058e-06,
      "loss": 0.0452,
      "step": 5600
    },
    {
      "epoch": 2.719766877124818,
      "eval_accuracy": 0.8650469088284821,
      "eval_f1_macro": 0.8179080927250358,
      "eval_f1_micro": 0.8650469088284821,
      "eval_loss": 0.6951431035995483,
      "eval_precision_macro": 0.8176656256318844,
      "eval_precision_micro": 0.8650469088284821,
      "eval_recall_macro": 0.8412836305297059,
      "eval_recall_micro": 0.8650469088284821,
      "eval_runtime": 78.6127,
      "eval_samples_per_second": 52.879,
      "eval_steps_per_second": 3.307,
      "step": 5600
    },
    {
      "epoch": 2.768334142787761,
      "grad_norm": 0.023337820544838905,
      "learning_rate": 6.160757649344342e-06,
      "loss": 0.0238,
      "step": 5700
    },
    {
      "epoch": 2.768334142787761,
      "eval_accuracy": 0.8691363964397402,
      "eval_f1_macro": 0.8141029552401969,
      "eval_f1_micro": 0.8691363964397402,
      "eval_loss": 0.6843555569648743,
      "eval_precision_macro": 0.828107626781137,
      "eval_precision_micro": 0.8691363964397402,
      "eval_recall_macro": 0.8223290635441496,
      "eval_recall_micro": 0.8691363964397402,
      "eval_runtime": 77.8775,
      "eval_samples_per_second": 53.379,
      "eval_steps_per_second": 3.339,
      "step": 5700
    },
    {
      "epoch": 2.816901408450704,
      "grad_norm": 2.3684067726135254,
      "learning_rate": 5.917921321029626e-06,
      "loss": 0.0344,
      "step": 5800
    },
    {
      "epoch": 2.816901408450704,
      "eval_accuracy": 0.8842915564108732,
      "eval_f1_macro": 0.8254107617822057,
      "eval_f1_micro": 0.8842915564108732,
      "eval_loss": 0.5869230628013611,
      "eval_precision_macro": 0.8172320894762466,
      "eval_precision_micro": 0.8842915564108732,
      "eval_recall_macro": 0.8495662880335549,
      "eval_recall_micro": 0.8842915564108732,
      "eval_runtime": 78.0202,
      "eval_samples_per_second": 53.281,
      "eval_steps_per_second": 3.332,
      "step": 5800
    },
    {
      "epoch": 2.8654686741136475,
      "grad_norm": 0.07361145317554474,
      "learning_rate": 5.675084992714911e-06,
      "loss": 0.0212,
      "step": 5900
    },
    {
      "epoch": 2.8654686741136475,
      "eval_accuracy": 0.8768342554726967,
      "eval_f1_macro": 0.8234793626505037,
      "eval_f1_micro": 0.8768342554726967,
      "eval_loss": 0.6304789185523987,
      "eval_precision_macro": 0.8184524998357775,
      "eval_precision_micro": 0.8768342554726967,
      "eval_recall_macro": 0.8476992326610299,
      "eval_recall_micro": 0.8768342554726967,
      "eval_runtime": 78.6516,
      "eval_samples_per_second": 52.853,
      "eval_steps_per_second": 3.306,
      "step": 5900
    },
    {
      "epoch": 2.9140359397765905,
      "grad_norm": 0.009860685095191002,
      "learning_rate": 5.432248664400195e-06,
      "loss": 0.0239,
      "step": 6000
    },
    {
      "epoch": 2.9140359397765905,
      "eval_accuracy": 0.8518162136155881,
      "eval_f1_macro": 0.8017569225322448,
      "eval_f1_micro": 0.8518162136155881,
      "eval_loss": 0.7799412608146667,
      "eval_precision_macro": 0.7982453929078707,
      "eval_precision_micro": 0.8518162136155881,
      "eval_recall_macro": 0.8304382158522342,
      "eval_recall_micro": 0.8518162136155881,
      "eval_runtime": 78.0376,
      "eval_samples_per_second": 53.269,
      "eval_steps_per_second": 3.332,
      "step": 6000
    },
    {
      "epoch": 2.9626032054395335,
      "grad_norm": 0.0258096382021904,
      "learning_rate": 5.189412336085479e-06,
      "loss": 0.0293,
      "step": 6100
    },
    {
      "epoch": 2.9626032054395335,
      "eval_accuracy": 0.8741881164301178,
      "eval_f1_macro": 0.8230892222950853,
      "eval_f1_micro": 0.8741881164301178,
      "eval_loss": 0.6950201392173767,
      "eval_precision_macro": 0.8197692931333949,
      "eval_precision_micro": 0.8741881164301178,
      "eval_recall_macro": 0.8477652020234965,
      "eval_recall_micro": 0.8741881164301178,
      "eval_runtime": 78.3208,
      "eval_samples_per_second": 53.077,
      "eval_steps_per_second": 3.32,
      "step": 6100
    },
    {
      "epoch": 3.011170471102477,
      "grad_norm": 0.1370718628168106,
      "learning_rate": 4.946576007770763e-06,
      "loss": 0.0288,
      "step": 6200
    },
    {
      "epoch": 3.011170471102477,
      "eval_accuracy": 0.8789992783257157,
      "eval_f1_macro": 0.829792442447409,
      "eval_f1_micro": 0.8789992783257157,
      "eval_loss": 0.6330971717834473,
      "eval_precision_macro": 0.8231458659168076,
      "eval_precision_micro": 0.8789992783257157,
      "eval_recall_macro": 0.8504144231450652,
      "eval_recall_micro": 0.8789992783257157,
      "eval_runtime": 78.7114,
      "eval_samples_per_second": 52.813,
      "eval_steps_per_second": 3.303,
      "step": 6200
    },
    {
      "epoch": 3.05973773676542,
      "grad_norm": 0.055757131427526474,
      "learning_rate": 4.703739679456047e-06,
      "loss": 0.0254,
      "step": 6300
    },
    {
      "epoch": 3.05973773676542,
      "eval_accuracy": 0.8799615107048352,
      "eval_f1_macro": 0.8318246689081815,
      "eval_f1_micro": 0.8799615107048352,
      "eval_loss": 0.63609379529953,
      "eval_precision_macro": 0.8269628089569904,
      "eval_precision_micro": 0.8799615107048352,
      "eval_recall_macro": 0.8565965365714439,
      "eval_recall_micro": 0.8799615107048352,
      "eval_runtime": 78.354,
      "eval_samples_per_second": 53.054,
      "eval_steps_per_second": 3.318,
      "step": 6300
    },
    {
      "epoch": 3.1083050024283634,
      "grad_norm": 0.006881752982735634,
      "learning_rate": 4.460903351141332e-06,
      "loss": 0.0252,
      "step": 6400
    },
    {
      "epoch": 3.1083050024283634,
      "eval_accuracy": 0.8862160211691124,
      "eval_f1_macro": 0.8441970931986851,
      "eval_f1_micro": 0.8862160211691124,
      "eval_loss": 0.6103322505950928,
      "eval_precision_macro": 0.8494002256031983,
      "eval_precision_micro": 0.8862160211691124,
      "eval_recall_macro": 0.8594469740432842,
      "eval_recall_micro": 0.8862160211691124,
      "eval_runtime": 78.3491,
      "eval_samples_per_second": 53.057,
      "eval_steps_per_second": 3.318,
      "step": 6400
    },
    {
      "epoch": 3.1568722680913064,
      "grad_norm": 0.02003752812743187,
      "learning_rate": 4.2180670228266156e-06,
      "loss": 0.0062,
      "step": 6500
    },
    {
      "epoch": 3.1568722680913064,
      "eval_accuracy": 0.8833293240317537,
      "eval_f1_macro": 0.8395368491034129,
      "eval_f1_micro": 0.8833293240317537,
      "eval_loss": 0.6034556031227112,
      "eval_precision_macro": 0.8398338415533984,
      "eval_precision_micro": 0.8833293240317537,
      "eval_recall_macro": 0.8524929004333107,
      "eval_recall_micro": 0.8833293240317537,
      "eval_runtime": 78.6267,
      "eval_samples_per_second": 52.87,
      "eval_steps_per_second": 3.307,
      "step": 6500
    },
    {
      "epoch": 3.20543953375425,
      "grad_norm": 0.09360187500715256,
      "learning_rate": 3.9752306945118995e-06,
      "loss": 0.0176,
      "step": 6600
    },
    {
      "epoch": 3.20543953375425,
      "eval_accuracy": 0.8850132306952129,
      "eval_f1_macro": 0.8304379003006633,
      "eval_f1_micro": 0.8850132306952128,
      "eval_loss": 0.6116442084312439,
      "eval_precision_macro": 0.8326738669282544,
      "eval_precision_micro": 0.8850132306952129,
      "eval_recall_macro": 0.85264914984025,
      "eval_recall_micro": 0.8850132306952129,
      "eval_runtime": 78.4069,
      "eval_samples_per_second": 53.018,
      "eval_steps_per_second": 3.316,
      "step": 6600
    },
    {
      "epoch": 3.254006799417193,
      "grad_norm": 0.0037081497721374035,
      "learning_rate": 3.7323943661971835e-06,
      "loss": 0.0077,
      "step": 6700
    },
    {
      "epoch": 3.254006799417193,
      "eval_accuracy": 0.8845321145056532,
      "eval_f1_macro": 0.8347831431371906,
      "eval_f1_micro": 0.8845321145056532,
      "eval_loss": 0.6256404519081116,
      "eval_precision_macro": 0.8436031947040645,
      "eval_precision_micro": 0.8845321145056532,
      "eval_recall_macro": 0.8486489503127234,
      "eval_recall_micro": 0.8845321145056532,
      "eval_runtime": 78.5317,
      "eval_samples_per_second": 52.934,
      "eval_steps_per_second": 3.311,
      "step": 6700
    },
    {
      "epoch": 3.302574065080136,
      "grad_norm": 0.06124697998166084,
      "learning_rate": 3.4895580378824675e-06,
      "loss": 0.0127,
      "step": 6800
    },
    {
      "epoch": 3.302574065080136,
      "eval_accuracy": 0.8794803945152755,
      "eval_f1_macro": 0.8278216288594707,
      "eval_f1_micro": 0.8794803945152755,
      "eval_loss": 0.6568040251731873,
      "eval_precision_macro": 0.8467736273399344,
      "eval_precision_micro": 0.8794803945152755,
      "eval_recall_macro": 0.8387083670261228,
      "eval_recall_micro": 0.8794803945152755,
      "eval_runtime": 78.3608,
      "eval_samples_per_second": 53.05,
      "eval_steps_per_second": 3.318,
      "step": 6800
    },
    {
      "epoch": 3.3511413307430793,
      "grad_norm": 0.25665995478630066,
      "learning_rate": 3.246721709567752e-06,
      "loss": 0.0178,
      "step": 6900
    },
    {
      "epoch": 3.3511413307430793,
      "eval_accuracy": 0.8713014192927592,
      "eval_f1_macro": 0.8179505523753748,
      "eval_f1_micro": 0.8713014192927592,
      "eval_loss": 0.7154167890548706,
      "eval_precision_macro": 0.8310341284957633,
      "eval_precision_micro": 0.8713014192927592,
      "eval_recall_macro": 0.8341873689934125,
      "eval_recall_micro": 0.8713014192927592,
      "eval_runtime": 78.685,
      "eval_samples_per_second": 52.831,
      "eval_steps_per_second": 3.304,
      "step": 6900
    },
    {
      "epoch": 3.3997085964060223,
      "grad_norm": 0.037899695336818695,
      "learning_rate": 3.0038853812530354e-06,
      "loss": 0.0019,
      "step": 7000
    },
    {
      "epoch": 3.3997085964060223,
      "eval_accuracy": 0.8816454173682945,
      "eval_f1_macro": 0.833404594848917,
      "eval_f1_micro": 0.8816454173682945,
      "eval_loss": 0.6380771994590759,
      "eval_precision_macro": 0.8392890539432627,
      "eval_precision_micro": 0.8816454173682945,
      "eval_recall_macro": 0.8437175960246225,
      "eval_recall_micro": 0.8816454173682945,
      "eval_runtime": 78.6949,
      "eval_samples_per_second": 52.824,
      "eval_steps_per_second": 3.304,
      "step": 7000
    },
    {
      "epoch": 3.4482758620689653,
      "grad_norm": 31.638689041137695,
      "learning_rate": 2.7610490529383198e-06,
      "loss": 0.0142,
      "step": 7100
    },
    {
      "epoch": 3.4482758620689653,
      "eval_accuracy": 0.8828482078421939,
      "eval_f1_macro": 0.833083270199735,
      "eval_f1_micro": 0.8828482078421939,
      "eval_loss": 0.6579678058624268,
      "eval_precision_macro": 0.8455418017703596,
      "eval_precision_micro": 0.8828482078421939,
      "eval_recall_macro": 0.8454918923836,
      "eval_recall_micro": 0.8828482078421939,
      "eval_runtime": 78.211,
      "eval_samples_per_second": 53.151,
      "eval_steps_per_second": 3.324,
      "step": 7100
    },
    {
      "epoch": 3.4968431277319088,
      "grad_norm": 28.2841796875,
      "learning_rate": 2.5182127246236037e-06,
      "loss": 0.0018,
      "step": 7200
    },
    {
      "epoch": 3.4968431277319088,
      "eval_accuracy": 0.8746692326196777,
      "eval_f1_macro": 0.8263362849592439,
      "eval_f1_micro": 0.8746692326196777,
      "eval_loss": 0.6722654104232788,
      "eval_precision_macro": 0.8233328721924625,
      "eval_precision_micro": 0.8746692326196777,
      "eval_recall_macro": 0.8458647636918942,
      "eval_recall_micro": 0.8746692326196777,
      "eval_runtime": 78.2301,
      "eval_samples_per_second": 53.138,
      "eval_steps_per_second": 3.324,
      "step": 7200
    },
    {
      "epoch": 3.5454103933948518,
      "grad_norm": 0.006923174951225519,
      "learning_rate": 2.2753763963088877e-06,
      "loss": 0.0159,
      "step": 7300
    },
    {
      "epoch": 3.5454103933948518,
      "eval_accuracy": 0.8700986288188598,
      "eval_f1_macro": 0.8196568446600343,
      "eval_f1_micro": 0.8700986288188598,
      "eval_loss": 0.7045225501060486,
      "eval_precision_macro": 0.8167725390706483,
      "eval_precision_micro": 0.8700986288188598,
      "eval_recall_macro": 0.8395234839263153,
      "eval_recall_micro": 0.8700986288188598,
      "eval_runtime": 78.7461,
      "eval_samples_per_second": 52.79,
      "eval_steps_per_second": 3.302,
      "step": 7300
    },
    {
      "epoch": 3.593977659057795,
      "grad_norm": 0.011453253217041492,
      "learning_rate": 2.032540067994172e-06,
      "loss": 0.0037,
      "step": 7400
    },
    {
      "epoch": 3.593977659057795,
      "eval_accuracy": 0.882607649747414,
      "eval_f1_macro": 0.8337455736704167,
      "eval_f1_micro": 0.882607649747414,
      "eval_loss": 0.6420170068740845,
      "eval_precision_macro": 0.8351021322092593,
      "eval_precision_micro": 0.882607649747414,
      "eval_recall_macro": 0.8453310977676587,
      "eval_recall_micro": 0.882607649747414,
      "eval_runtime": 78.3451,
      "eval_samples_per_second": 53.06,
      "eval_steps_per_second": 3.319,
      "step": 7400
    },
    {
      "epoch": 3.642544924720738,
      "grad_norm": 0.004663607105612755,
      "learning_rate": 1.789703739679456e-06,
      "loss": 0.0084,
      "step": 7500
    },
    {
      "epoch": 3.642544924720738,
      "eval_accuracy": 0.8802020687996152,
      "eval_f1_macro": 0.8320621801285876,
      "eval_f1_micro": 0.8802020687996152,
      "eval_loss": 0.652312695980072,
      "eval_precision_macro": 0.8306346210019144,
      "eval_precision_micro": 0.8802020687996152,
      "eval_recall_macro": 0.8493781023440272,
      "eval_recall_micro": 0.8802020687996152,
      "eval_runtime": 78.7968,
      "eval_samples_per_second": 52.756,
      "eval_steps_per_second": 3.3,
      "step": 7500
    },
    {
      "epoch": 3.6911121903836817,
      "grad_norm": 0.05554908514022827,
      "learning_rate": 1.5468674113647402e-06,
      "loss": 0.0017,
      "step": 7600
    },
    {
      "epoch": 3.6911121903836817,
      "eval_accuracy": 0.8874188116430118,
      "eval_f1_macro": 0.8401905388217382,
      "eval_f1_micro": 0.8874188116430118,
      "eval_loss": 0.6221657395362854,
      "eval_precision_macro": 0.8387107750272869,
      "eval_precision_micro": 0.8874188116430118,
      "eval_recall_macro": 0.8610361407962509,
      "eval_recall_micro": 0.8874188116430118,
      "eval_runtime": 78.6101,
      "eval_samples_per_second": 52.881,
      "eval_steps_per_second": 3.307,
      "step": 7600
    },
    {
      "epoch": 3.7396794560466247,
      "grad_norm": 2.2692418098449707,
      "learning_rate": 1.3040310830500244e-06,
      "loss": 0.0071,
      "step": 7700
    },
    {
      "epoch": 3.7396794560466247,
      "eval_accuracy": 0.8785181621361559,
      "eval_f1_macro": 0.8326744354650085,
      "eval_f1_micro": 0.878518162136156,
      "eval_loss": 0.664691686630249,
      "eval_precision_macro": 0.8341535564737329,
      "eval_precision_micro": 0.8785181621361559,
      "eval_recall_macro": 0.8463902152998535,
      "eval_recall_micro": 0.8785181621361559,
      "eval_runtime": 78.4331,
      "eval_samples_per_second": 53.001,
      "eval_steps_per_second": 3.315,
      "step": 7700
    },
    {
      "epoch": 3.7882467217095677,
      "grad_norm": 0.009112412109971046,
      "learning_rate": 1.0611947547353086e-06,
      "loss": 0.0018,
      "step": 7800
    },
    {
      "epoch": 3.7882467217095677,
      "eval_accuracy": 0.8787587202309358,
      "eval_f1_macro": 0.8287278332969615,
      "eval_f1_micro": 0.8787587202309358,
      "eval_loss": 0.6747535467147827,
      "eval_precision_macro": 0.8265516787617085,
      "eval_precision_micro": 0.8787587202309358,
      "eval_recall_macro": 0.8490595593522744,
      "eval_recall_micro": 0.8787587202309358,
      "eval_runtime": 78.5872,
      "eval_samples_per_second": 52.897,
      "eval_steps_per_second": 3.308,
      "step": 7800
    },
    {
      "epoch": 3.836813987372511,
      "grad_norm": 0.08465871959924698,
      "learning_rate": 8.183584264205927e-07,
      "loss": 0.0063,
      "step": 7900
    },
    {
      "epoch": 3.836813987372511,
      "eval_accuracy": 0.8802020687996152,
      "eval_f1_macro": 0.8275737635437894,
      "eval_f1_micro": 0.8802020687996152,
      "eval_loss": 0.6518697142601013,
      "eval_precision_macro": 0.8241036586009164,
      "eval_precision_micro": 0.8802020687996152,
      "eval_recall_macro": 0.8505984084112971,
      "eval_recall_micro": 0.8802020687996152,
      "eval_runtime": 78.3552,
      "eval_samples_per_second": 53.053,
      "eval_steps_per_second": 3.318,
      "step": 7900
    },
    {
      "epoch": 3.885381253035454,
      "grad_norm": 0.01252959668636322,
      "learning_rate": 5.755220981058767e-07,
      "loss": 0.0012,
      "step": 8000
    },
    {
      "epoch": 3.885381253035454,
      "eval_accuracy": 0.8816454173682945,
      "eval_f1_macro": 0.8325617260526238,
      "eval_f1_micro": 0.8816454173682945,
      "eval_loss": 0.6509329080581665,
      "eval_precision_macro": 0.8261296227157098,
      "eval_precision_micro": 0.8816454173682945,
      "eval_recall_macro": 0.8554119401439346,
      "eval_recall_micro": 0.8816454173682945,
      "eval_runtime": 78.6734,
      "eval_samples_per_second": 52.839,
      "eval_steps_per_second": 3.305,
      "step": 8000
    },
    {
      "epoch": 3.933948518698397,
      "grad_norm": 0.00854838639497757,
      "learning_rate": 3.3268576979116075e-07,
      "loss": 0.0062,
      "step": 8100
    },
    {
      "epoch": 3.933948518698397,
      "eval_accuracy": 0.8811643011787347,
      "eval_f1_macro": 0.8331789648254515,
      "eval_f1_micro": 0.8811643011787347,
      "eval_loss": 0.6525153517723083,
      "eval_precision_macro": 0.8280847186790202,
      "eval_precision_micro": 0.8811643011787347,
      "eval_recall_macro": 0.8554521062472861,
      "eval_recall_micro": 0.8811643011787347,
      "eval_runtime": 78.2235,
      "eval_samples_per_second": 53.143,
      "eval_steps_per_second": 3.324,
      "step": 8100
    },
    {
      "epoch": 3.9825157843613406,
      "grad_norm": 0.004602006636559963,
      "learning_rate": 8.984944147644487e-08,
      "loss": 0.0097,
      "step": 8200
    },
    {
      "epoch": 3.9825157843613406,
      "eval_accuracy": 0.8802020687996152,
      "eval_f1_macro": 0.832469524971897,
      "eval_f1_micro": 0.8802020687996152,
      "eval_loss": 0.6588266491889954,
      "eval_precision_macro": 0.8274534781731434,
      "eval_precision_micro": 0.8802020687996152,
      "eval_recall_macro": 0.8546630804582604,
      "eval_recall_micro": 0.8802020687996152,
      "eval_runtime": 78.8748,
      "eval_samples_per_second": 52.704,
      "eval_steps_per_second": 3.296,
      "step": 8200
    }
  ],
  "logging_steps": 100,
  "max_steps": 8236,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.467447932454093e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}