{
  "best_metric": 0.038124363869428635,
  "best_model_checkpoint": "saves/psy-course/Llama3-OpenBioLLM-8B/train/fold6/checkpoint-1900",
  "epoch": 4.995305164319249,
  "eval_steps": 50,
  "global_step": 3325,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.015023474178403756,
      "grad_norm": 6.414540767669678,
      "learning_rate": 3.003003003003003e-06,
      "loss": 1.3292,
      "step": 10
    },
    {
      "epoch": 0.03004694835680751,
      "grad_norm": 5.740021228790283,
      "learning_rate": 6.006006006006006e-06,
      "loss": 1.3844,
      "step": 20
    },
    {
      "epoch": 0.04507042253521127,
      "grad_norm": 4.104672431945801,
      "learning_rate": 9.00900900900901e-06,
      "loss": 1.1953,
      "step": 30
    },
    {
      "epoch": 0.06009389671361502,
      "grad_norm": 1.9491015672683716,
      "learning_rate": 1.2012012012012012e-05,
      "loss": 0.8062,
      "step": 40
    },
    {
      "epoch": 0.07511737089201878,
      "grad_norm": 1.8273999691009521,
      "learning_rate": 1.5015015015015016e-05,
      "loss": 0.4971,
      "step": 50
    },
    {
      "epoch": 0.07511737089201878,
      "eval_loss": 0.3236617147922516,
      "eval_runtime": 215.6614,
      "eval_samples_per_second": 5.49,
      "eval_steps_per_second": 5.49,
      "step": 50
    },
    {
      "epoch": 0.09014084507042254,
      "grad_norm": 1.789884328842163,
      "learning_rate": 1.801801801801802e-05,
      "loss": 0.3049,
      "step": 60
    },
    {
      "epoch": 0.10516431924882629,
      "grad_norm": 1.026201844215393,
      "learning_rate": 2.102102102102102e-05,
      "loss": 0.1795,
      "step": 70
    },
    {
      "epoch": 0.12018779342723004,
      "grad_norm": 1.4471508264541626,
      "learning_rate": 2.4024024024024024e-05,
      "loss": 0.1482,
      "step": 80
    },
    {
      "epoch": 0.1352112676056338,
      "grad_norm": 1.0525811910629272,
      "learning_rate": 2.702702702702703e-05,
      "loss": 0.0928,
      "step": 90
    },
    {
      "epoch": 0.15023474178403756,
      "grad_norm": 1.223236322402954,
      "learning_rate": 3.0030030030030033e-05,
      "loss": 0.1179,
      "step": 100
    },
    {
      "epoch": 0.15023474178403756,
      "eval_loss": 0.09419180452823639,
      "eval_runtime": 224.392,
      "eval_samples_per_second": 5.276,
      "eval_steps_per_second": 5.276,
      "step": 100
    },
    {
      "epoch": 0.1652582159624413,
      "grad_norm": 1.3412930965423584,
      "learning_rate": 3.3033033033033035e-05,
      "loss": 0.1185,
      "step": 110
    },
    {
      "epoch": 0.18028169014084508,
      "grad_norm": 1.0879250764846802,
      "learning_rate": 3.603603603603604e-05,
      "loss": 0.0985,
      "step": 120
    },
    {
      "epoch": 0.19530516431924883,
      "grad_norm": 0.9736179709434509,
      "learning_rate": 3.903903903903904e-05,
      "loss": 0.0828,
      "step": 130
    },
    {
      "epoch": 0.21032863849765257,
      "grad_norm": 0.6209188103675842,
      "learning_rate": 4.204204204204204e-05,
      "loss": 0.1053,
      "step": 140
    },
    {
      "epoch": 0.22535211267605634,
      "grad_norm": 0.6934787631034851,
      "learning_rate": 4.5045045045045046e-05,
      "loss": 0.0891,
      "step": 150
    },
    {
      "epoch": 0.22535211267605634,
      "eval_loss": 0.07970932126045227,
      "eval_runtime": 225.0802,
      "eval_samples_per_second": 5.26,
      "eval_steps_per_second": 5.26,
      "step": 150
    },
    {
      "epoch": 0.2403755868544601,
      "grad_norm": 1.273955225944519,
      "learning_rate": 4.804804804804805e-05,
      "loss": 0.0755,
      "step": 160
    },
    {
      "epoch": 0.25539906103286386,
      "grad_norm": 1.2673016786575317,
      "learning_rate": 5.105105105105106e-05,
      "loss": 0.0569,
      "step": 170
    },
    {
      "epoch": 0.2704225352112676,
      "grad_norm": 1.060423493385315,
      "learning_rate": 5.405405405405406e-05,
      "loss": 0.0744,
      "step": 180
    },
    {
      "epoch": 0.28544600938967135,
      "grad_norm": 1.2299424409866333,
      "learning_rate": 5.705705705705706e-05,
      "loss": 0.0684,
      "step": 190
    },
    {
      "epoch": 0.3004694835680751,
      "grad_norm": 0.6046801805496216,
      "learning_rate": 6.0060060060060066e-05,
      "loss": 0.0539,
      "step": 200
    },
    {
      "epoch": 0.3004694835680751,
      "eval_loss": 0.05980731174349785,
      "eval_runtime": 226.7678,
      "eval_samples_per_second": 5.221,
      "eval_steps_per_second": 5.221,
      "step": 200
    },
    {
      "epoch": 0.3154929577464789,
      "grad_norm": 0.5380669236183167,
      "learning_rate": 6.306306306306306e-05,
      "loss": 0.054,
      "step": 210
    },
    {
      "epoch": 0.3305164319248826,
      "grad_norm": 0.42053088545799255,
      "learning_rate": 6.606606606606607e-05,
      "loss": 0.0645,
      "step": 220
    },
    {
      "epoch": 0.3455399061032864,
      "grad_norm": 0.6181445717811584,
      "learning_rate": 6.906906906906907e-05,
      "loss": 0.0527,
      "step": 230
    },
    {
      "epoch": 0.36056338028169016,
      "grad_norm": 0.8342611193656921,
      "learning_rate": 7.207207207207208e-05,
      "loss": 0.0622,
      "step": 240
    },
    {
      "epoch": 0.3755868544600939,
      "grad_norm": 0.7963096499443054,
      "learning_rate": 7.507507507507507e-05,
      "loss": 0.065,
      "step": 250
    },
    {
      "epoch": 0.3755868544600939,
      "eval_loss": 0.05752342566847801,
      "eval_runtime": 228.7801,
      "eval_samples_per_second": 5.175,
      "eval_steps_per_second": 5.175,
      "step": 250
    },
    {
      "epoch": 0.39061032863849765,
      "grad_norm": 0.5707411170005798,
      "learning_rate": 7.807807807807808e-05,
      "loss": 0.0526,
      "step": 260
    },
    {
      "epoch": 0.4056338028169014,
      "grad_norm": 0.8613728880882263,
      "learning_rate": 8.108108108108109e-05,
      "loss": 0.0638,
      "step": 270
    },
    {
      "epoch": 0.42065727699530514,
      "grad_norm": 0.20531082153320312,
      "learning_rate": 8.408408408408409e-05,
      "loss": 0.0646,
      "step": 280
    },
    {
      "epoch": 0.4356807511737089,
      "grad_norm": 1.4829570055007935,
      "learning_rate": 8.70870870870871e-05,
      "loss": 0.0644,
      "step": 290
    },
    {
      "epoch": 0.4507042253521127,
      "grad_norm": 0.5894226431846619,
      "learning_rate": 9.009009009009009e-05,
      "loss": 0.052,
      "step": 300
    },
    {
      "epoch": 0.4507042253521127,
      "eval_loss": 0.053021036088466644,
      "eval_runtime": 224.341,
      "eval_samples_per_second": 5.278,
      "eval_steps_per_second": 5.278,
      "step": 300
    },
    {
      "epoch": 0.46572769953051646,
      "grad_norm": 0.3916723132133484,
      "learning_rate": 9.30930930930931e-05,
      "loss": 0.0449,
      "step": 310
    },
    {
      "epoch": 0.4807511737089202,
      "grad_norm": 0.7790238857269287,
      "learning_rate": 9.60960960960961e-05,
      "loss": 0.0598,
      "step": 320
    },
    {
      "epoch": 0.49577464788732395,
      "grad_norm": 0.6429105997085571,
      "learning_rate": 9.90990990990991e-05,
      "loss": 0.0561,
      "step": 330
    },
    {
      "epoch": 0.5107981220657277,
      "grad_norm": 0.9011141657829285,
      "learning_rate": 9.999864944989638e-05,
      "loss": 0.0683,
      "step": 340
    },
    {
      "epoch": 0.5258215962441315,
      "grad_norm": 0.3162938058376312,
      "learning_rate": 9.999203468625017e-05,
      "loss": 0.0535,
      "step": 350
    },
    {
      "epoch": 0.5258215962441315,
      "eval_loss": 0.054493363946676254,
      "eval_runtime": 227.3472,
      "eval_samples_per_second": 5.208,
      "eval_steps_per_second": 5.208,
      "step": 350
    },
    {
      "epoch": 0.5408450704225352,
      "grad_norm": 0.7987211346626282,
      "learning_rate": 9.997990837719421e-05,
      "loss": 0.0634,
      "step": 360
    },
    {
      "epoch": 0.5558685446009389,
      "grad_norm": 0.30378398299217224,
      "learning_rate": 9.996227185963554e-05,
      "loss": 0.0451,
      "step": 370
    },
    {
      "epoch": 0.5708920187793427,
      "grad_norm": 0.7728825807571411,
      "learning_rate": 9.993912707797329e-05,
      "loss": 0.0467,
      "step": 380
    },
    {
      "epoch": 0.5859154929577465,
      "grad_norm": 0.3161417245864868,
      "learning_rate": 9.99104765838842e-05,
      "loss": 0.0493,
      "step": 390
    },
    {
      "epoch": 0.6009389671361502,
      "grad_norm": 0.26571208238601685,
      "learning_rate": 9.987632353604151e-05,
      "loss": 0.0512,
      "step": 400
    },
    {
      "epoch": 0.6009389671361502,
      "eval_loss": 0.04660526290535927,
      "eval_runtime": 226.5936,
      "eval_samples_per_second": 5.225,
      "eval_steps_per_second": 5.225,
      "step": 400
    },
    {
      "epoch": 0.615962441314554,
      "grad_norm": 0.7127376794815063,
      "learning_rate": 9.98366716997665e-05,
      "loss": 0.0459,
      "step": 410
    },
    {
      "epoch": 0.6309859154929578,
      "grad_norm": 1.0153098106384277,
      "learning_rate": 9.979152544661354e-05,
      "loss": 0.0478,
      "step": 420
    },
    {
      "epoch": 0.6460093896713615,
      "grad_norm": 0.6635463833808899,
      "learning_rate": 9.974088975388802e-05,
      "loss": 0.0498,
      "step": 430
    },
    {
      "epoch": 0.6610328638497652,
      "grad_norm": 0.5931793451309204,
      "learning_rate": 9.968477020409766e-05,
      "loss": 0.0489,
      "step": 440
    },
    {
      "epoch": 0.676056338028169,
      "grad_norm": 0.5385611057281494,
      "learning_rate": 9.962317298433705e-05,
      "loss": 0.0597,
      "step": 450
    },
    {
      "epoch": 0.676056338028169,
      "eval_loss": 0.049538787454366684,
      "eval_runtime": 226.4382,
      "eval_samples_per_second": 5.229,
      "eval_steps_per_second": 5.229,
      "step": 450
    },
    {
      "epoch": 0.6910798122065728,
      "grad_norm": 0.31567174196243286,
      "learning_rate": 9.955610488560551e-05,
      "loss": 0.0595,
      "step": 460
    },
    {
      "epoch": 0.7061032863849765,
      "grad_norm": 0.6918370127677917,
      "learning_rate": 9.948357330205842e-05,
      "loss": 0.061,
      "step": 470
    },
    {
      "epoch": 0.7211267605633803,
      "grad_norm": 0.8577624559402466,
      "learning_rate": 9.940558623019201e-05,
      "loss": 0.032,
      "step": 480
    },
    {
      "epoch": 0.7361502347417841,
      "grad_norm": 0.6278219819068909,
      "learning_rate": 9.932215226796172e-05,
      "loss": 0.0486,
      "step": 490
    },
    {
      "epoch": 0.7511737089201878,
      "grad_norm": 0.3588387072086334,
      "learning_rate": 9.923328061383435e-05,
      "loss": 0.0509,
      "step": 500
    },
    {
      "epoch": 0.7511737089201878,
      "eval_loss": 0.046869516372680664,
      "eval_runtime": 228.8856,
      "eval_samples_per_second": 5.173,
      "eval_steps_per_second": 5.173,
      "step": 500
    },
    {
      "epoch": 0.7661971830985915,
      "grad_norm": 0.24121642112731934,
      "learning_rate": 9.913898106577393e-05,
      "loss": 0.049,
      "step": 510
    },
    {
      "epoch": 0.7812206572769953,
      "grad_norm": 0.3446851670742035,
      "learning_rate": 9.903926402016153e-05,
      "loss": 0.0405,
      "step": 520
    },
    {
      "epoch": 0.7962441314553991,
      "grad_norm": 0.30128219723701477,
      "learning_rate": 9.893414047064897e-05,
      "loss": 0.0452,
      "step": 530
    },
    {
      "epoch": 0.8112676056338028,
      "grad_norm": 0.4867357313632965,
      "learning_rate": 9.88236220069469e-05,
      "loss": 0.0483,
      "step": 540
    },
    {
      "epoch": 0.8262910798122066,
      "grad_norm": 0.3773770034313202,
      "learning_rate": 9.870772081354705e-05,
      "loss": 0.0624,
      "step": 550
    },
    {
      "epoch": 0.8262910798122066,
      "eval_loss": 0.04311800375580788,
      "eval_runtime": 225.9134,
      "eval_samples_per_second": 5.241,
      "eval_steps_per_second": 5.241,
      "step": 550
    },
    {
      "epoch": 0.8413145539906103,
      "grad_norm": 0.3748703896999359,
      "learning_rate": 9.858644966837878e-05,
      "loss": 0.0493,
      "step": 560
    },
    {
      "epoch": 0.856338028169014,
      "grad_norm": 0.19429147243499756,
      "learning_rate": 9.845982194140051e-05,
      "loss": 0.038,
      "step": 570
    },
    {
      "epoch": 0.8713615023474178,
      "grad_norm": 0.24159827828407288,
      "learning_rate": 9.832785159312559e-05,
      "loss": 0.0435,
      "step": 580
    },
    {
      "epoch": 0.8863849765258216,
      "grad_norm": 0.3861151337623596,
      "learning_rate": 9.819055317308317e-05,
      "loss": 0.0389,
      "step": 590
    },
    {
      "epoch": 0.9014084507042254,
      "grad_norm": 0.2947847545146942,
      "learning_rate": 9.804794181821422e-05,
      "loss": 0.035,
      "step": 600
    },
    {
      "epoch": 0.9014084507042254,
      "eval_loss": 0.04581260681152344,
      "eval_runtime": 228.7478,
      "eval_samples_per_second": 5.176,
      "eval_steps_per_second": 5.176,
      "step": 600
    },
    {
      "epoch": 0.9164319248826291,
      "grad_norm": 0.9220906496047974,
      "learning_rate": 9.790003325120261e-05,
      "loss": 0.0477,
      "step": 610
    },
    {
      "epoch": 0.9314553990610329,
      "grad_norm": 0.585889995098114,
      "learning_rate": 9.774684377874178e-05,
      "loss": 0.0565,
      "step": 620
    },
    {
      "epoch": 0.9464788732394366,
      "grad_norm": 0.16347403824329376,
      "learning_rate": 9.758839028973692e-05,
      "loss": 0.0356,
      "step": 630
    },
    {
      "epoch": 0.9615023474178404,
      "grad_norm": 0.4304580092430115,
      "learning_rate": 9.742469025344298e-05,
      "loss": 0.0467,
      "step": 640
    },
    {
      "epoch": 0.9765258215962441,
      "grad_norm": 0.2779543697834015,
      "learning_rate": 9.725576171753874e-05,
      "loss": 0.0546,
      "step": 650
    },
    {
      "epoch": 0.9765258215962441,
      "eval_loss": 0.04435986280441284,
      "eval_runtime": 228.5445,
      "eval_samples_per_second": 5.181,
      "eval_steps_per_second": 5.181,
      "step": 650
    },
    {
      "epoch": 0.9915492957746479,
      "grad_norm": 0.8497869372367859,
      "learning_rate": 9.708162330613708e-05,
      "loss": 0.0396,
      "step": 660
    },
    {
      "epoch": 1.0065727699530516,
      "grad_norm": 0.2351168841123581,
      "learning_rate": 9.690229421773167e-05,
      "loss": 0.0434,
      "step": 670
    },
    {
      "epoch": 1.0215962441314554,
      "grad_norm": 0.4020308256149292,
      "learning_rate": 9.67177942230804e-05,
      "loss": 0.0369,
      "step": 680
    },
    {
      "epoch": 1.036619718309859,
      "grad_norm": 0.48213866353034973,
      "learning_rate": 9.652814366302568e-05,
      "loss": 0.0409,
      "step": 690
    },
    {
      "epoch": 1.051643192488263,
      "grad_norm": 0.30210596323013306,
      "learning_rate": 9.633336344625185e-05,
      "loss": 0.0404,
      "step": 700
    },
    {
      "epoch": 1.051643192488263,
      "eval_loss": 0.04425961896777153,
      "eval_runtime": 230.6178,
      "eval_samples_per_second": 5.134,
      "eval_steps_per_second": 5.134,
      "step": 700
    },
    {
      "epoch": 1.0666666666666667,
      "grad_norm": 0.3107031583786011,
      "learning_rate": 9.61334750469801e-05,
      "loss": 0.029,
      "step": 710
    },
    {
      "epoch": 1.0816901408450703,
      "grad_norm": 0.5843893885612488,
      "learning_rate": 9.592850050260089e-05,
      "loss": 0.0373,
      "step": 720
    },
    {
      "epoch": 1.0967136150234742,
      "grad_norm": 0.26578277349472046,
      "learning_rate": 9.571846241124446e-05,
      "loss": 0.0328,
      "step": 730
    },
    {
      "epoch": 1.1117370892018779,
      "grad_norm": 0.22206802666187286,
      "learning_rate": 9.55033839292893e-05,
      "loss": 0.028,
      "step": 740
    },
    {
      "epoch": 1.1267605633802817,
      "grad_norm": 0.5371259450912476,
      "learning_rate": 9.52832887688093e-05,
      "loss": 0.0342,
      "step": 750
    },
    {
      "epoch": 1.1267605633802817,
      "eval_loss": 0.043438900262117386,
      "eval_runtime": 230.5863,
      "eval_samples_per_second": 5.135,
      "eval_steps_per_second": 5.135,
      "step": 750
    },
    {
      "epoch": 1.1417840375586854,
      "grad_norm": 0.43538734316825867,
      "learning_rate": 9.50582011949595e-05,
      "loss": 0.0391,
      "step": 760
    },
    {
      "epoch": 1.1568075117370893,
      "grad_norm": 0.23159150779247284,
      "learning_rate": 9.482814602330084e-05,
      "loss": 0.0354,
      "step": 770
    },
    {
      "epoch": 1.171830985915493,
      "grad_norm": 0.5288726091384888,
      "learning_rate": 9.459314861706435e-05,
      "loss": 0.0386,
      "step": 780
    },
    {
      "epoch": 1.1868544600938966,
      "grad_norm": 0.3396289646625519,
      "learning_rate": 9.435323488435488e-05,
      "loss": 0.0291,
      "step": 790
    },
    {
      "epoch": 1.2018779342723005,
      "grad_norm": 0.5357405543327332,
      "learning_rate": 9.410843127529473e-05,
      "loss": 0.0298,
      "step": 800
    },
    {
      "epoch": 1.2018779342723005,
      "eval_loss": 0.04284869506955147,
      "eval_runtime": 228.5369,
      "eval_samples_per_second": 5.181,
      "eval_steps_per_second": 5.181,
      "step": 800
    },
    {
      "epoch": 1.2169014084507042,
      "grad_norm": 0.6564673781394958,
      "learning_rate": 9.385876477910765e-05,
      "loss": 0.0371,
      "step": 810
    },
    {
      "epoch": 1.231924882629108,
      "grad_norm": 0.15312528610229492,
      "learning_rate": 9.360426292114314e-05,
      "loss": 0.0379,
      "step": 820
    },
    {
      "epoch": 1.2469483568075117,
      "grad_norm": 0.32437649369239807,
      "learning_rate": 9.334495375984212e-05,
      "loss": 0.0334,
      "step": 830
    },
    {
      "epoch": 1.2619718309859156,
      "grad_norm": 0.33815252780914307,
      "learning_rate": 9.30808658836432e-05,
      "loss": 0.0305,
      "step": 840
    },
    {
      "epoch": 1.2769953051643192,
      "grad_norm": 0.43382740020751953,
      "learning_rate": 9.281202840783108e-05,
      "loss": 0.0348,
      "step": 850
    },
    {
      "epoch": 1.2769953051643192,
      "eval_loss": 0.04067714512348175,
      "eval_runtime": 232.2301,
      "eval_samples_per_second": 5.098,
      "eval_steps_per_second": 5.098,
      "step": 850
    },
    {
      "epoch": 1.292018779342723,
      "grad_norm": 0.3475792109966278,
      "learning_rate": 9.253847097132655e-05,
      "loss": 0.0285,
      "step": 860
    },
    {
      "epoch": 1.3070422535211268,
      "grad_norm": 0.24043916165828705,
      "learning_rate": 9.226022373341882e-05,
      "loss": 0.0285,
      "step": 870
    },
    {
      "epoch": 1.3220657276995305,
      "grad_norm": 0.2261325865983963,
      "learning_rate": 9.19773173704406e-05,
      "loss": 0.0263,
      "step": 880
    },
    {
      "epoch": 1.3370892018779343,
      "grad_norm": 0.673026978969574,
      "learning_rate": 9.168978307238594e-05,
      "loss": 0.0371,
      "step": 890
    },
    {
      "epoch": 1.352112676056338,
      "grad_norm": 0.1485815793275833,
      "learning_rate": 9.13976525394717e-05,
      "loss": 0.0291,
      "step": 900
    },
    {
      "epoch": 1.352112676056338,
      "eval_loss": 0.0397120825946331,
      "eval_runtime": 229.6421,
      "eval_samples_per_second": 5.156,
      "eval_steps_per_second": 5.156,
      "step": 900
    },
    {
      "epoch": 1.3671361502347419,
      "grad_norm": 0.20742298662662506,
      "learning_rate": 9.110095797864263e-05,
      "loss": 0.0295,
      "step": 910
    },
    {
      "epoch": 1.3821596244131455,
      "grad_norm": 0.1670617312192917,
      "learning_rate": 9.079973210002051e-05,
      "loss": 0.0317,
      "step": 920
    },
    {
      "epoch": 1.3971830985915492,
      "grad_norm": 0.4337206482887268,
      "learning_rate": 9.049400811329807e-05,
      "loss": 0.044,
      "step": 930
    },
    {
      "epoch": 1.412206572769953,
      "grad_norm": 0.36409109830856323,
      "learning_rate": 9.01838197240775e-05,
      "loss": 0.0267,
      "step": 940
    },
    {
      "epoch": 1.4272300469483568,
      "grad_norm": 0.45880192518234253,
      "learning_rate": 8.986920113015461e-05,
      "loss": 0.0373,
      "step": 950
    },
    {
      "epoch": 1.4272300469483568,
      "eval_loss": 0.03974553570151329,
      "eval_runtime": 232.2398,
      "eval_samples_per_second": 5.098,
      "eval_steps_per_second": 5.098,
      "step": 950
    },
    {
      "epoch": 1.4422535211267606,
      "grad_norm": 0.2591331899166107,
      "learning_rate": 8.955018701774846e-05,
      "loss": 0.0277,
      "step": 960
    },
    {
      "epoch": 1.4572769953051643,
      "grad_norm": 0.20547696948051453,
      "learning_rate": 8.922681255767731e-05,
      "loss": 0.0418,
      "step": 970
    },
    {
      "epoch": 1.4723004694835682,
      "grad_norm": 0.2559107542037964,
      "learning_rate": 8.889911340148112e-05,
      "loss": 0.035,
      "step": 980
    },
    {
      "epoch": 1.4873239436619718,
      "grad_norm": 0.20421642065048218,
      "learning_rate": 8.856712567749095e-05,
      "loss": 0.0315,
      "step": 990
    },
    {
      "epoch": 1.5023474178403755,
      "grad_norm": 0.35287970304489136,
      "learning_rate": 8.82308859868459e-05,
      "loss": 0.0289,
      "step": 1000
    },
    {
      "epoch": 1.5023474178403755,
      "eval_loss": 0.04059433564543724,
      "eval_runtime": 232.2328,
      "eval_samples_per_second": 5.098,
      "eval_steps_per_second": 5.098,
      "step": 1000
    },
    {
      "epoch": 1.5173708920187794,
      "grad_norm": 0.3293749690055847,
      "learning_rate": 8.789043139945795e-05,
      "loss": 0.0332,
      "step": 1010
    },
    {
      "epoch": 1.532394366197183,
      "grad_norm": 0.26048192381858826,
      "learning_rate": 8.754579944992491e-05,
      "loss": 0.0419,
      "step": 1020
    },
    {
      "epoch": 1.5474178403755867,
      "grad_norm": 0.2248084545135498,
      "learning_rate": 8.719702813339248e-05,
      "loss": 0.0327,
      "step": 1030
    },
    {
      "epoch": 1.5624413145539906,
      "grad_norm": 0.29436999559402466,
      "learning_rate": 8.684415590136518e-05,
      "loss": 0.0305,
      "step": 1040
    },
    {
      "epoch": 1.5774647887323945,
      "grad_norm": 0.2553965151309967,
      "learning_rate": 8.648722165746722e-05,
      "loss": 0.0339,
      "step": 1050
    },
    {
      "epoch": 1.5774647887323945,
      "eval_loss": 0.0449674129486084,
      "eval_runtime": 229.1439,
      "eval_samples_per_second": 5.167,
      "eval_steps_per_second": 5.167,
      "step": 1050
    },
    {
      "epoch": 1.5924882629107981,
      "grad_norm": 0.2621479034423828,
      "learning_rate": 8.61262647531534e-05,
      "loss": 0.0273,
      "step": 1060
    },
    {
      "epoch": 1.6075117370892018,
      "grad_norm": 0.29867053031921387,
      "learning_rate": 8.576132498337068e-05,
      "loss": 0.0349,
      "step": 1070
    },
    {
      "epoch": 1.6225352112676057,
      "grad_norm": 0.5712867379188538,
      "learning_rate": 8.539244258217088e-05,
      "loss": 0.0332,
      "step": 1080
    },
    {
      "epoch": 1.6375586854460094,
      "grad_norm": 0.2706688940525055,
      "learning_rate": 8.501965821827485e-05,
      "loss": 0.0408,
      "step": 1090
    },
    {
      "epoch": 1.652582159624413,
      "grad_norm": 0.2529209554195404,
      "learning_rate": 8.464301299058892e-05,
      "loss": 0.0244,
      "step": 1100
    },
    {
      "epoch": 1.652582159624413,
      "eval_loss": 0.04117948189377785,
      "eval_runtime": 232.3365,
      "eval_samples_per_second": 5.096,
      "eval_steps_per_second": 5.096,
      "step": 1100
    },
    {
      "epoch": 1.667605633802817,
      "grad_norm": 0.2566877007484436,
      "learning_rate": 8.426254842367374e-05,
      "loss": 0.0296,
      "step": 1110
    },
    {
      "epoch": 1.6826291079812208,
      "grad_norm": 0.28313955664634705,
      "learning_rate": 8.387830646316623e-05,
      "loss": 0.0382,
      "step": 1120
    },
    {
      "epoch": 1.6976525821596244,
      "grad_norm": 0.5802918672561646,
      "learning_rate": 8.349032947115525e-05,
      "loss": 0.0285,
      "step": 1130
    },
    {
      "epoch": 1.712676056338028,
      "grad_norm": 0.31035321950912476,
      "learning_rate": 8.309866022151107e-05,
      "loss": 0.0422,
      "step": 1140
    },
    {
      "epoch": 1.727699530516432,
      "grad_norm": 0.37052640318870544,
      "learning_rate": 8.270334189516983e-05,
      "loss": 0.0288,
      "step": 1150
    },
    {
      "epoch": 1.727699530516432,
      "eval_loss": 0.04018314927816391,
      "eval_runtime": 229.361,
      "eval_samples_per_second": 5.162,
      "eval_steps_per_second": 5.162,
      "step": 1150
    },
    {
      "epoch": 1.7427230046948357,
      "grad_norm": 0.13857990503311157,
      "learning_rate": 8.230441807537277e-05,
      "loss": 0.0418,
      "step": 1160
    },
    {
      "epoch": 1.7577464788732393,
      "grad_norm": 0.20266057550907135,
      "learning_rate": 8.190193274286122e-05,
      "loss": 0.0272,
      "step": 1170
    },
    {
      "epoch": 1.7727699530516432,
      "grad_norm": 0.26995351910591125,
      "learning_rate": 8.149593027102789e-05,
      "loss": 0.0336,
      "step": 1180
    },
    {
      "epoch": 1.787793427230047,
      "grad_norm": 0.5027073621749878,
      "learning_rate": 8.108645542102469e-05,
      "loss": 0.0282,
      "step": 1190
    },
    {
      "epoch": 1.8028169014084507,
      "grad_norm": 0.3197694420814514,
      "learning_rate": 8.067355333682798e-05,
      "loss": 0.036,
      "step": 1200
    },
    {
      "epoch": 1.8028169014084507,
      "eval_loss": 0.03951646015048027,
      "eval_runtime": 233.3104,
      "eval_samples_per_second": 5.075,
      "eval_steps_per_second": 5.075,
      "step": 1200
    },
    {
      "epoch": 1.8178403755868544,
      "grad_norm": 0.16591861844062805,
      "learning_rate": 8.025726954026138e-05,
      "loss": 0.0257,
      "step": 1210
    },
    {
      "epoch": 1.8328638497652583,
      "grad_norm": 0.5136527419090271,
      "learning_rate": 7.983764992597716e-05,
      "loss": 0.0305,
      "step": 1220
    },
    {
      "epoch": 1.847887323943662,
      "grad_norm": 0.4316839575767517,
      "learning_rate": 7.94147407563964e-05,
      "loss": 0.0321,
      "step": 1230
    },
    {
      "epoch": 1.8629107981220656,
      "grad_norm": 0.38824647665023804,
      "learning_rate": 7.89885886566086e-05,
      "loss": 0.0352,
      "step": 1240
    },
    {
      "epoch": 1.8779342723004695,
      "grad_norm": 0.4996744990348816,
      "learning_rate": 7.855924060923141e-05,
      "loss": 0.0334,
      "step": 1250
    },
    {
      "epoch": 1.8779342723004695,
      "eval_loss": 0.039211343973875046,
      "eval_runtime": 231.7629,
      "eval_samples_per_second": 5.109,
      "eval_steps_per_second": 5.109,
      "step": 1250
    },
    {
      "epoch": 1.8929577464788734,
      "grad_norm": 0.23325949907302856,
      "learning_rate": 7.812674394923077e-05,
      "loss": 0.0346,
      "step": 1260
    },
    {
      "epoch": 1.907981220657277,
      "grad_norm": 0.27697038650512695,
      "learning_rate": 7.769114635870231e-05,
      "loss": 0.0317,
      "step": 1270
    },
    {
      "epoch": 1.9230046948356807,
      "grad_norm": 0.4152398407459259,
      "learning_rate": 7.725249586161463e-05,
      "loss": 0.0341,
      "step": 1280
    },
    {
      "epoch": 1.9380281690140846,
      "grad_norm": 0.20453670620918274,
      "learning_rate": 7.68108408185145e-05,
      "loss": 0.0299,
      "step": 1290
    },
    {
      "epoch": 1.9530516431924883,
      "grad_norm": 0.38925862312316895,
      "learning_rate": 7.636622992119536e-05,
      "loss": 0.0568,
      "step": 1300
    },
    {
      "epoch": 1.9530516431924883,
      "eval_loss": 0.04275708645582199,
      "eval_runtime": 226.0867,
      "eval_samples_per_second": 5.237,
      "eval_steps_per_second": 5.237,
      "step": 1300
    },
    {
      "epoch": 1.968075117370892,
      "grad_norm": 0.254408597946167,
      "learning_rate": 7.591871218732902e-05,
      "loss": 0.0371,
      "step": 1310
    },
    {
      "epoch": 1.9830985915492958,
      "grad_norm": 0.279060959815979,
      "learning_rate": 7.54683369550616e-05,
      "loss": 0.0443,
      "step": 1320
    },
    {
      "epoch": 1.9981220657276997,
      "grad_norm": 0.1406373232603073,
      "learning_rate": 7.501515387757404e-05,
      "loss": 0.034,
      "step": 1330
    },
    {
      "epoch": 2.013145539906103,
      "grad_norm": 0.2928116023540497,
      "learning_rate": 7.455921291760796e-05,
      "loss": 0.0376,
      "step": 1340
    },
    {
      "epoch": 2.028169014084507,
      "grad_norm": 0.10005494207143784,
      "learning_rate": 7.410056434195725e-05,
      "loss": 0.0206,
      "step": 1350
    },
    {
      "epoch": 2.028169014084507,
      "eval_loss": 0.04088796302676201,
      "eval_runtime": 225.6026,
      "eval_samples_per_second": 5.248,
      "eval_steps_per_second": 5.248,
      "step": 1350
    },
    {
      "epoch": 2.043192488262911,
      "grad_norm": 0.28136786818504333,
      "learning_rate": 7.363925871592629e-05,
      "loss": 0.0218,
      "step": 1360
    },
    {
      "epoch": 2.0582159624413148,
      "grad_norm": 0.15816110372543335,
      "learning_rate": 7.317534689775528e-05,
      "loss": 0.019,
      "step": 1370
    },
    {
      "epoch": 2.073239436619718,
      "grad_norm": 0.15044282376766205,
      "learning_rate": 7.270888003301304e-05,
      "loss": 0.016,
      "step": 1380
    },
    {
      "epoch": 2.088262910798122,
      "grad_norm": 0.6464121341705322,
      "learning_rate": 7.22399095489584e-05,
      "loss": 0.0205,
      "step": 1390
    },
    {
      "epoch": 2.103286384976526,
      "grad_norm": 0.21453744173049927,
      "learning_rate": 7.176848714887042e-05,
      "loss": 0.0226,
      "step": 1400
    },
    {
      "epoch": 2.103286384976526,
      "eval_loss": 0.04083973169326782,
      "eval_runtime": 220.6924,
      "eval_samples_per_second": 5.365,
      "eval_steps_per_second": 5.365,
      "step": 1400
    },
    {
      "epoch": 2.1183098591549294,
      "grad_norm": 0.32658934593200684,
      "learning_rate": 7.129466480634806e-05,
      "loss": 0.021,
      "step": 1410
    },
    {
      "epoch": 2.1333333333333333,
      "grad_norm": 0.5312864184379578,
      "learning_rate": 7.081849475958042e-05,
      "loss": 0.0184,
      "step": 1420
    },
    {
      "epoch": 2.148356807511737,
      "grad_norm": 0.30401676893234253,
      "learning_rate": 7.034002950558723e-05,
      "loss": 0.0181,
      "step": 1430
    },
    {
      "epoch": 2.1633802816901406,
      "grad_norm": 0.2417522370815277,
      "learning_rate": 6.985932179443144e-05,
      "loss": 0.0225,
      "step": 1440
    },
    {
      "epoch": 2.1784037558685445,
      "grad_norm": 0.31011879444122314,
      "learning_rate": 6.937642462340342e-05,
      "loss": 0.0257,
      "step": 1450
    },
    {
      "epoch": 2.1784037558685445,
      "eval_loss": 0.04065319150686264,
      "eval_runtime": 222.471,
      "eval_samples_per_second": 5.322,
      "eval_steps_per_second": 5.322,
      "step": 1450
    },
    {
      "epoch": 2.1934272300469484,
      "grad_norm": 0.2239389568567276,
      "learning_rate": 6.889139123117817e-05,
      "loss": 0.0183,
      "step": 1460
    },
    {
      "epoch": 2.2084507042253523,
      "grad_norm": 0.22514520585536957,
      "learning_rate": 6.840427509194575e-05,
      "loss": 0.0242,
      "step": 1470
    },
    {
      "epoch": 2.2234741784037557,
      "grad_norm": 0.2953309416770935,
      "learning_rate": 6.791512990951597e-05,
      "loss": 0.015,
      "step": 1480
    },
    {
      "epoch": 2.2384976525821596,
      "grad_norm": 0.35825031995773315,
      "learning_rate": 6.74240096113975e-05,
      "loss": 0.0254,
      "step": 1490
    },
    {
      "epoch": 2.2535211267605635,
      "grad_norm": 0.15820668637752533,
      "learning_rate": 6.693096834285256e-05,
      "loss": 0.0138,
      "step": 1500
    },
    {
      "epoch": 2.2535211267605635,
      "eval_loss": 0.04137204959988594,
      "eval_runtime": 221.3626,
      "eval_samples_per_second": 5.349,
      "eval_steps_per_second": 5.349,
      "step": 1500
    },
    {
      "epoch": 2.2685446009389674,
      "grad_norm": 0.8325169086456299,
      "learning_rate": 6.643606046092732e-05,
      "loss": 0.0239,
      "step": 1510
    },
    {
      "epoch": 2.283568075117371,
      "grad_norm": 0.2664135694503784,
      "learning_rate": 6.593934052845929e-05,
      "loss": 0.0217,
      "step": 1520
    },
    {
      "epoch": 2.2985915492957747,
      "grad_norm": 0.3122292160987854,
      "learning_rate": 6.544086330806181e-05,
      "loss": 0.022,
      "step": 1530
    },
    {
      "epoch": 2.3136150234741786,
      "grad_norm": 0.24781066179275513,
      "learning_rate": 6.494068375608646e-05,
      "loss": 0.0218,
      "step": 1540
    },
    {
      "epoch": 2.328638497652582,
      "grad_norm": 0.32878419756889343,
      "learning_rate": 6.443885701656432e-05,
      "loss": 0.0167,
      "step": 1550
    },
    {
      "epoch": 2.328638497652582,
      "eval_loss": 0.040929365903139114,
      "eval_runtime": 219.4831,
      "eval_samples_per_second": 5.394,
      "eval_steps_per_second": 5.394,
      "step": 1550
    },
    {
      "epoch": 2.343661971830986,
      "grad_norm": 0.49727535247802734,
      "learning_rate": 6.393543841512632e-05,
      "loss": 0.0223,
      "step": 1560
    },
    {
      "epoch": 2.35868544600939,
      "grad_norm": 0.617466926574707,
      "learning_rate": 6.343048345290386e-05,
      "loss": 0.0159,
      "step": 1570
    },
    {
      "epoch": 2.3737089201877932,
      "grad_norm": 0.3165408968925476,
      "learning_rate": 6.292404780040961e-05,
      "loss": 0.0278,
      "step": 1580
    },
    {
      "epoch": 2.388732394366197,
      "grad_norm": 0.2089942991733551,
      "learning_rate": 6.241618729140018e-05,
      "loss": 0.0187,
      "step": 1590
    },
    {
      "epoch": 2.403755868544601,
      "grad_norm": 0.3965398669242859,
      "learning_rate": 6.190695791672042e-05,
      "loss": 0.0217,
      "step": 1600
    },
    {
      "epoch": 2.403755868544601,
      "eval_loss": 0.03883142024278641,
      "eval_runtime": 221.119,
      "eval_samples_per_second": 5.355,
      "eval_steps_per_second": 5.355,
      "step": 1600
    },
    {
      "epoch": 2.418779342723005,
      "grad_norm": 0.16717387735843658,
      "learning_rate": 6.139641581813052e-05,
      "loss": 0.0173,
      "step": 1610
    },
    {
      "epoch": 2.4338028169014083,
      "grad_norm": 0.1234317496418953,
      "learning_rate": 6.088461728211642e-05,
      "loss": 0.016,
      "step": 1620
    },
    {
      "epoch": 2.448826291079812,
      "grad_norm": 0.3614145815372467,
      "learning_rate": 6.0371618733684474e-05,
      "loss": 0.0122,
      "step": 1630
    },
    {
      "epoch": 2.463849765258216,
      "grad_norm": 0.14805041253566742,
      "learning_rate": 5.9857476730140485e-05,
      "loss": 0.0216,
      "step": 1640
    },
    {
      "epoch": 2.4788732394366195,
      "grad_norm": 0.24672311544418335,
      "learning_rate": 5.9342247954854466e-05,
      "loss": 0.0195,
      "step": 1650
    },
    {
      "epoch": 2.4788732394366195,
      "eval_loss": 0.042663708329200745,
      "eval_runtime": 218.9509,
      "eval_samples_per_second": 5.408,
      "eval_steps_per_second": 5.408,
      "step": 1650
    },
    {
      "epoch": 2.4938967136150234,
      "grad_norm": 0.2416851967573166,
      "learning_rate": 5.8825989211011335e-05,
      "loss": 0.0239,
      "step": 1660
    },
    {
      "epoch": 2.5089201877934273,
      "grad_norm": 0.2621941864490509,
      "learning_rate": 5.830875741534852e-05,
      "loss": 0.0166,
      "step": 1670
    },
    {
      "epoch": 2.523943661971831,
      "grad_norm": 0.38329723477363586,
      "learning_rate": 5.7790609591880826e-05,
      "loss": 0.0252,
      "step": 1680
    },
    {
      "epoch": 2.5389671361502346,
      "grad_norm": 0.25781792402267456,
      "learning_rate": 5.727160286561386e-05,
      "loss": 0.0203,
      "step": 1690
    },
    {
      "epoch": 2.5539906103286385,
      "grad_norm": 0.2038751244544983,
      "learning_rate": 5.675179445624581e-05,
      "loss": 0.0223,
      "step": 1700
    },
    {
      "epoch": 2.5539906103286385,
      "eval_loss": 0.04374394193291664,
      "eval_runtime": 219.3848,
      "eval_samples_per_second": 5.397,
      "eval_steps_per_second": 5.397,
      "step": 1700
    },
    {
      "epoch": 2.5690140845070424,
      "grad_norm": 0.1650547981262207,
      "learning_rate": 5.62312416718593e-05,
      "loss": 0.0183,
      "step": 1710
    },
    {
      "epoch": 2.584037558685446,
      "grad_norm": 0.06820981949567795,
      "learning_rate": 5.5710001902603116e-05,
      "loss": 0.0217,
      "step": 1720
    },
    {
      "epoch": 2.5990610328638497,
      "grad_norm": 0.1459759920835495,
      "learning_rate": 5.5188132614365094e-05,
      "loss": 0.0164,
      "step": 1730
    },
    {
      "epoch": 2.6140845070422536,
      "grad_norm": 0.18846748769283295,
      "learning_rate": 5.4665691342436565e-05,
      "loss": 0.0145,
      "step": 1740
    },
    {
      "epoch": 2.629107981220657,
      "grad_norm": 0.44601181149482727,
      "learning_rate": 5.414273568516919e-05,
      "loss": 0.0195,
      "step": 1750
    },
    {
      "epoch": 2.629107981220657,
      "eval_loss": 0.0427960604429245,
      "eval_runtime": 219.0798,
      "eval_samples_per_second": 5.404,
      "eval_steps_per_second": 5.404,
      "step": 1750
    },
    {
      "epoch": 2.644131455399061,
      "grad_norm": 0.5126632452011108,
      "learning_rate": 5.361932329762481e-05,
      "loss": 0.0215,
      "step": 1760
    },
    {
      "epoch": 2.659154929577465,
      "grad_norm": 0.2724759876728058,
      "learning_rate": 5.309551188521914e-05,
      "loss": 0.0232,
      "step": 1770
    },
    {
      "epoch": 2.6741784037558687,
      "grad_norm": 0.4603884816169739,
      "learning_rate": 5.2571359197359704e-05,
      "loss": 0.0168,
      "step": 1780
    },
    {
      "epoch": 2.6892018779342726,
      "grad_norm": 0.3353332579135895,
      "learning_rate": 5.2046923021079175e-05,
      "loss": 0.0144,
      "step": 1790
    },
    {
      "epoch": 2.704225352112676,
      "grad_norm": 0.2024739384651184,
      "learning_rate": 5.1522261174664346e-05,
      "loss": 0.0218,
      "step": 1800
    },
    {
      "epoch": 2.704225352112676,
      "eval_loss": 0.04094401001930237,
      "eval_runtime": 211.573,
      "eval_samples_per_second": 5.596,
      "eval_steps_per_second": 5.596,
      "step": 1800
    },
    {
      "epoch": 2.71924882629108,
      "grad_norm": 1.624040126800537,
      "learning_rate": 5.0997431501281835e-05,
      "loss": 0.0262,
      "step": 1810
    },
    {
      "epoch": 2.7342723004694838,
      "grad_norm": 0.3581642806529999,
      "learning_rate": 5.0472491862600915e-05,
      "loss": 0.019,
      "step": 1820
    },
    {
      "epoch": 2.749295774647887,
      "grad_norm": 0.44195398688316345,
      "learning_rate": 4.994750013241435e-05,
      "loss": 0.0248,
      "step": 1830
    },
    {
      "epoch": 2.764319248826291,
      "grad_norm": 0.22482897341251373,
      "learning_rate": 4.9422514190257974e-05,
      "loss": 0.0221,
      "step": 1840
    },
    {
      "epoch": 2.779342723004695,
      "grad_norm": 0.2989714741706848,
      "learning_rate": 4.88975919150294e-05,
      "loss": 0.0189,
      "step": 1850
    },
    {
      "epoch": 2.779342723004695,
      "eval_loss": 0.040074534714221954,
      "eval_runtime": 207.949,
      "eval_samples_per_second": 5.694,
      "eval_steps_per_second": 5.694,
      "step": 1850
    },
    {
      "epoch": 2.7943661971830984,
      "grad_norm": 0.3000681400299072,
      "learning_rate": 4.83727911786071e-05,
      "loss": 0.0254,
      "step": 1860
    },
    {
      "epoch": 2.8093896713615023,
      "grad_norm": 0.3129828870296478,
      "learning_rate": 4.7848169839470145e-05,
      "loss": 0.0221,
      "step": 1870
    },
    {
      "epoch": 2.824413145539906,
      "grad_norm": 0.5919480323791504,
      "learning_rate": 4.7323785736319244e-05,
      "loss": 0.0239,
      "step": 1880
    },
    {
      "epoch": 2.8394366197183096,
      "grad_norm": 0.23092535138130188,
      "learning_rate": 4.679969668170024e-05,
      "loss": 0.0227,
      "step": 1890
    },
    {
      "epoch": 2.8544600938967135,
      "grad_norm": 0.10704797506332397,
      "learning_rate": 4.627596045563031e-05,
      "loss": 0.0195,
      "step": 1900
    },
    {
      "epoch": 2.8544600938967135,
      "eval_loss": 0.038124363869428635,
      "eval_runtime": 204.7966,
      "eval_samples_per_second": 5.781,
      "eval_steps_per_second": 5.781,
      "step": 1900
    },
    {
      "epoch": 2.8694835680751174,
      "grad_norm": 0.3655933439731598,
      "learning_rate": 4.575263479922783e-05,
      "loss": 0.014,
      "step": 1910
    },
    {
      "epoch": 2.8845070422535213,
      "grad_norm": 0.13271917402744293,
      "learning_rate": 4.522977740834651e-05,
      "loss": 0.0112,
      "step": 1920
    },
    {
      "epoch": 2.8995305164319247,
      "grad_norm": 0.12882518768310547,
      "learning_rate": 4.4707445927214456e-05,
      "loss": 0.0124,
      "step": 1930
    },
    {
      "epoch": 2.9145539906103286,
      "grad_norm": 0.3468819558620453,
      "learning_rate": 4.4185697942079115e-05,
      "loss": 0.0219,
      "step": 1940
    },
    {
      "epoch": 2.9295774647887325,
      "grad_norm": 0.16386888921260834,
      "learning_rate": 4.366459097485832e-05,
      "loss": 0.0188,
      "step": 1950
    },
    {
      "epoch": 2.9295774647887325,
      "eval_loss": 0.039931606501340866,
      "eval_runtime": 202.6841,
      "eval_samples_per_second": 5.842,
      "eval_steps_per_second": 5.842,
      "step": 1950
    },
    {
      "epoch": 2.9446009389671364,
      "grad_norm": 0.29388657212257385,
      "learning_rate": 4.314418247679866e-05,
      "loss": 0.0205,
      "step": 1960
    },
    {
      "epoch": 2.95962441314554,
      "grad_norm": 0.19597427546977997,
      "learning_rate": 4.26245298221416e-05,
      "loss": 0.0137,
      "step": 1970
    },
    {
      "epoch": 2.9746478873239437,
      "grad_norm": 0.16075938940048218,
      "learning_rate": 4.2105690301798014e-05,
      "loss": 0.0162,
      "step": 1980
    },
    {
      "epoch": 2.9896713615023476,
      "grad_norm": 0.39698508381843567,
      "learning_rate": 4.158772111703194e-05,
      "loss": 0.0163,
      "step": 1990
    },
    {
      "epoch": 3.004694835680751,
      "grad_norm": 0.044948216527700424,
      "learning_rate": 4.107067937315429e-05,
      "loss": 0.0105,
      "step": 2000
    },
    {
      "epoch": 3.004694835680751,
      "eval_loss": 0.041156113147735596,
      "eval_runtime": 200.8348,
      "eval_samples_per_second": 5.895,
      "eval_steps_per_second": 5.895,
      "step": 2000
    },
    {
      "epoch": 3.019718309859155,
      "grad_norm": 0.14478544890880585,
      "learning_rate": 4.055462207322698e-05,
      "loss": 0.0076,
      "step": 2010
    },
    {
      "epoch": 3.034741784037559,
      "grad_norm": 0.05940256267786026,
      "learning_rate": 4.003960611177855e-05,
      "loss": 0.0115,
      "step": 2020
    },
    {
      "epoch": 3.0497652582159622,
      "grad_norm": 0.13706958293914795,
      "learning_rate": 3.952568826853152e-05,
      "loss": 0.0102,
      "step": 2030
    },
    {
      "epoch": 3.064788732394366,
      "grad_norm": 0.22277607023715973,
      "learning_rate": 3.901292520214256e-05,
      "loss": 0.0064,
      "step": 2040
    },
    {
      "epoch": 3.07981220657277,
      "grad_norm": 0.30167967081069946,
      "learning_rate": 3.850137344395598e-05,
      "loss": 0.0073,
      "step": 2050
    },
    {
      "epoch": 3.07981220657277,
      "eval_loss": 0.04430518299341202,
      "eval_runtime": 200.5966,
      "eval_samples_per_second": 5.902,
      "eval_steps_per_second": 5.902,
      "step": 2050
    },
    {
      "epoch": 3.094835680751174,
      "grad_norm": 0.6749436855316162,
      "learning_rate": 3.799108939177118e-05,
      "loss": 0.0064,
      "step": 2060
    },
    {
      "epoch": 3.1098591549295773,
      "grad_norm": 0.16386157274246216,
      "learning_rate": 3.7482129303624934e-05,
      "loss": 0.0077,
      "step": 2070
    },
    {
      "epoch": 3.124882629107981,
      "grad_norm": 0.35549381375312805,
      "learning_rate": 3.697454929158901e-05,
      "loss": 0.0089,
      "step": 2080
    },
    {
      "epoch": 3.139906103286385,
      "grad_norm": 0.2858954071998596,
      "learning_rate": 3.6468405315583854e-05,
      "loss": 0.0102,
      "step": 2090
    },
    {
      "epoch": 3.1549295774647885,
      "grad_norm": 0.43447062373161316,
      "learning_rate": 3.59637531772092e-05,
      "loss": 0.0082,
      "step": 2100
    },
    {
      "epoch": 3.1549295774647885,
      "eval_loss": 0.0465645007789135,
      "eval_runtime": 200.3422,
      "eval_samples_per_second": 5.91,
      "eval_steps_per_second": 5.91,
      "step": 2100
    },
    {
      "epoch": 3.1699530516431924,
      "grad_norm": 0.16099733114242554,
      "learning_rate": 3.546064851359192e-05,
      "loss": 0.0037,
      "step": 2110
    },
    {
      "epoch": 3.1849765258215963,
      "grad_norm": 0.33014243841171265,
      "learning_rate": 3.495914679125212e-05,
      "loss": 0.0068,
      "step": 2120
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.04894431680440903,
      "learning_rate": 3.445930329998819e-05,
      "loss": 0.007,
      "step": 2130
    },
    {
      "epoch": 3.2150234741784036,
      "grad_norm": 0.33057722449302673,
      "learning_rate": 3.396117314678097e-05,
      "loss": 0.0137,
      "step": 2140
    },
    {
      "epoch": 3.2300469483568075,
      "grad_norm": 0.4633444845676422,
      "learning_rate": 3.3464811249718474e-05,
      "loss": 0.0107,
      "step": 2150
    },
    {
      "epoch": 3.2300469483568075,
      "eval_loss": 0.04904927685856819,
      "eval_runtime": 199.8955,
      "eval_samples_per_second": 5.923,
      "eval_steps_per_second": 5.923,
      "step": 2150
    },
    {
      "epoch": 3.2450704225352114,
      "grad_norm": 0.27093496918678284,
      "learning_rate": 3.297027233194114e-05,
      "loss": 0.0087,
      "step": 2160
    },
    {
      "epoch": 3.260093896713615,
      "grad_norm": 0.047191720455884933,
      "learning_rate": 3.2477610915608704e-05,
      "loss": 0.0066,
      "step": 2170
    },
    {
      "epoch": 3.2751173708920187,
      "grad_norm": 0.3168286383152008,
      "learning_rate": 3.1986881315889315e-05,
      "loss": 0.0125,
      "step": 2180
    },
    {
      "epoch": 3.2901408450704226,
      "grad_norm": 0.5179242491722107,
      "learning_rate": 3.149813763497124e-05,
      "loss": 0.0167,
      "step": 2190
    },
    {
      "epoch": 3.3051643192488265,
      "grad_norm": 0.2823299169540405,
      "learning_rate": 3.101143375609818e-05,
      "loss": 0.0143,
      "step": 2200
    },
    {
      "epoch": 3.3051643192488265,
      "eval_loss": 0.044749025255441666,
      "eval_runtime": 199.0042,
      "eval_samples_per_second": 5.95,
      "eval_steps_per_second": 5.95,
      "step": 2200
    },
    {
      "epoch": 3.32018779342723,
      "grad_norm": 0.27977901697158813,
      "learning_rate": 3.0526823337628915e-05,
      "loss": 0.012,
      "step": 2210
    },
    {
      "epoch": 3.335211267605634,
      "grad_norm": 0.1816035658121109,
      "learning_rate": 3.004435980712129e-05,
      "loss": 0.0071,
      "step": 2220
    },
    {
      "epoch": 3.3502347417840377,
      "grad_norm": 0.20612962543964386,
      "learning_rate": 2.9564096355442116e-05,
      "loss": 0.0109,
      "step": 2230
    },
    {
      "epoch": 3.365258215962441,
      "grad_norm": 0.0782838836312294,
      "learning_rate": 2.9086085930902824e-05,
      "loss": 0.0109,
      "step": 2240
    },
    {
      "epoch": 3.380281690140845,
      "grad_norm": 0.1939852386713028,
      "learning_rate": 2.8610381233422058e-05,
      "loss": 0.005,
      "step": 2250
    },
    {
      "epoch": 3.380281690140845,
      "eval_loss": 0.04711684584617615,
      "eval_runtime": 199.3048,
      "eval_samples_per_second": 5.941,
      "eval_steps_per_second": 5.941,
      "step": 2250
    },
    {
      "epoch": 3.395305164319249,
      "grad_norm": 0.15407223999500275,
      "learning_rate": 2.8137034708715592e-05,
      "loss": 0.0051,
      "step": 2260
    },
    {
      "epoch": 3.4103286384976528,
      "grad_norm": 0.0431525744497776,
      "learning_rate": 2.7666098542514273e-05,
      "loss": 0.0061,
      "step": 2270
    },
    {
      "epoch": 3.425352112676056,
      "grad_norm": 0.4643717408180237,
      "learning_rate": 2.719762465481055e-05,
      "loss": 0.0095,
      "step": 2280
    },
    {
      "epoch": 3.44037558685446,
      "grad_norm": 0.13977164030075073,
      "learning_rate": 2.6731664694134473e-05,
      "loss": 0.0114,
      "step": 2290
    },
    {
      "epoch": 3.455399061032864,
      "grad_norm": 0.34736713767051697,
      "learning_rate": 2.6268270031859476e-05,
      "loss": 0.0079,
      "step": 2300
    },
    {
      "epoch": 3.455399061032864,
      "eval_loss": 0.0480869896709919,
      "eval_runtime": 198.5254,
      "eval_samples_per_second": 5.964,
      "eval_steps_per_second": 5.964,
      "step": 2300
    },
    {
      "epoch": 3.4704225352112674,
      "grad_norm": 0.5037516355514526,
      "learning_rate": 2.580749175653877e-05,
      "loss": 0.0097,
      "step": 2310
    },
    {
      "epoch": 3.4854460093896713,
      "grad_norm": 0.5058691501617432,
      "learning_rate": 2.5349380668272905e-05,
      "loss": 0.011,
      "step": 2320
    },
    {
      "epoch": 3.500469483568075,
      "grad_norm": 0.3391149938106537,
      "learning_rate": 2.489398727310908e-05,
      "loss": 0.0105,
      "step": 2330
    },
    {
      "epoch": 3.5154929577464786,
      "grad_norm": 0.5301526188850403,
      "learning_rate": 2.4441361777473066e-05,
      "loss": 0.0097,
      "step": 2340
    },
    {
      "epoch": 3.5305164319248825,
      "grad_norm": 0.5552868247032166,
      "learning_rate": 2.3991554082633912e-05,
      "loss": 0.0085,
      "step": 2350
    },
    {
      "epoch": 3.5305164319248825,
      "eval_loss": 0.05010828375816345,
      "eval_runtime": 198.3007,
      "eval_samples_per_second": 5.971,
      "eval_steps_per_second": 5.971,
      "step": 2350
    },
    {
      "epoch": 3.5455399061032864,
      "grad_norm": 0.15667784214019775,
      "learning_rate": 2.354461377920239e-05,
      "loss": 0.0078,
      "step": 2360
    },
    {
      "epoch": 3.5605633802816903,
      "grad_norm": 0.485943466424942,
      "learning_rate": 2.3100590141663807e-05,
      "loss": 0.0108,
      "step": 2370
    },
    {
      "epoch": 3.575586854460094,
      "grad_norm": 0.49499568343162537,
      "learning_rate": 2.265953212294551e-05,
      "loss": 0.0115,
      "step": 2380
    },
    {
      "epoch": 3.5906103286384976,
      "grad_norm": 0.559502899646759,
      "learning_rate": 2.2221488349019903e-05,
      "loss": 0.017,
      "step": 2390
    },
    {
      "epoch": 3.6056338028169015,
      "grad_norm": 0.260917067527771,
      "learning_rate": 2.1786507113543457e-05,
      "loss": 0.0073,
      "step": 2400
    },
    {
      "epoch": 3.6056338028169015,
      "eval_loss": 0.046827565878629684,
      "eval_runtime": 192.641,
      "eval_samples_per_second": 6.146,
      "eval_steps_per_second": 6.146,
      "step": 2400
    },
    {
      "epoch": 3.6206572769953054,
      "grad_norm": 0.07891547679901123,
      "learning_rate": 2.1354636372532523e-05,
      "loss": 0.0125,
      "step": 2410
    },
    {
      "epoch": 3.635680751173709,
      "grad_norm": 0.23800870776176453,
      "learning_rate": 2.092592373907617e-05,
      "loss": 0.0082,
      "step": 2420
    },
    {
      "epoch": 3.6507042253521127,
      "grad_norm": 0.09424114972352982,
      "learning_rate": 2.0500416478086932e-05,
      "loss": 0.0077,
      "step": 2430
    },
    {
      "epoch": 3.6657276995305166,
      "grad_norm": 0.3062797784805298,
      "learning_rate": 2.0078161501089954e-05,
      "loss": 0.0075,
      "step": 2440
    },
    {
      "epoch": 3.68075117370892,
      "grad_norm": 0.06614656746387482,
      "learning_rate": 1.9659205361050982e-05,
      "loss": 0.0034,
      "step": 2450
    },
    {
      "epoch": 3.68075117370892,
      "eval_loss": 0.04812372848391533,
      "eval_runtime": 197.4077,
      "eval_samples_per_second": 5.998,
      "eval_steps_per_second": 5.998,
      "step": 2450
    },
    {
      "epoch": 3.695774647887324,
      "grad_norm": 0.16158843040466309,
      "learning_rate": 1.924359424724408e-05,
      "loss": 0.0169,
      "step": 2460
    },
    {
      "epoch": 3.710798122065728,
      "grad_norm": 0.3151260018348694,
      "learning_rate": 1.8831373980159296e-05,
      "loss": 0.0127,
      "step": 2470
    },
    {
      "epoch": 3.7258215962441312,
      "grad_norm": 0.09106522053480148,
      "learning_rate": 1.8422590006450947e-05,
      "loss": 0.0042,
      "step": 2480
    },
    {
      "epoch": 3.740845070422535,
      "grad_norm": 0.07405770570039749,
      "learning_rate": 1.801728739392731e-05,
      "loss": 0.0058,
      "step": 2490
    },
    {
      "epoch": 3.755868544600939,
      "grad_norm": 0.16569849848747253,
      "learning_rate": 1.7615510826581904e-05,
      "loss": 0.0053,
      "step": 2500
    },
    {
      "epoch": 3.755868544600939,
      "eval_loss": 0.04956234619021416,
      "eval_runtime": 197.5247,
      "eval_samples_per_second": 5.994,
      "eval_steps_per_second": 5.994,
      "step": 2500
    },
    {
      "epoch": 3.770892018779343,
      "grad_norm": 0.271708220243454,
      "learning_rate": 1.7217304599667146e-05,
      "loss": 0.0087,
      "step": 2510
    },
    {
      "epoch": 3.7859154929577463,
      "grad_norm": 0.03417084738612175,
      "learning_rate": 1.6822712614810893e-05,
      "loss": 0.008,
      "step": 2520
    },
    {
      "epoch": 3.80093896713615,
      "grad_norm": 0.10270609706640244,
      "learning_rate": 1.643177837517631e-05,
      "loss": 0.0078,
      "step": 2530
    },
    {
      "epoch": 3.815962441314554,
      "grad_norm": 0.21668383479118347,
      "learning_rate": 1.6044544980665767e-05,
      "loss": 0.01,
      "step": 2540
    },
    {
      "epoch": 3.830985915492958,
      "grad_norm": 0.7087284922599792,
      "learning_rate": 1.5661055123169126e-05,
      "loss": 0.0052,
      "step": 2550
    },
    {
      "epoch": 3.830985915492958,
      "eval_loss": 0.04982556402683258,
      "eval_runtime": 197.2,
      "eval_samples_per_second": 6.004,
      "eval_steps_per_second": 6.004,
      "step": 2550
    },
    {
      "epoch": 3.8460093896713614,
      "grad_norm": 0.09885504096746445,
      "learning_rate": 1.5281351081856974e-05,
      "loss": 0.0065,
      "step": 2560
    },
    {
      "epoch": 3.8610328638497653,
      "grad_norm": 0.1990659385919571,
      "learning_rate": 1.4905474718519491e-05,
      "loss": 0.0032,
      "step": 2570
    },
    {
      "epoch": 3.876056338028169,
      "grad_norm": 0.2634921371936798,
      "learning_rate": 1.453346747295119e-05,
      "loss": 0.0116,
      "step": 2580
    },
    {
      "epoch": 3.8910798122065726,
      "grad_norm": 0.045282598584890366,
      "learning_rate": 1.4165370358382274e-05,
      "loss": 0.007,
      "step": 2590
    },
    {
      "epoch": 3.9061032863849765,
      "grad_norm": 0.04671657830476761,
      "learning_rate": 1.3801223956956994e-05,
      "loss": 0.0117,
      "step": 2600
    },
    {
      "epoch": 3.9061032863849765,
      "eval_loss": 0.05007306486368179,
      "eval_runtime": 196.5935,
      "eval_samples_per_second": 6.023,
      "eval_steps_per_second": 6.023,
      "step": 2600
    },
    {
      "epoch": 3.9211267605633804,
      "grad_norm": 0.2485828697681427,
      "learning_rate": 1.344106841525946e-05,
      "loss": 0.0108,
      "step": 2610
    },
    {
      "epoch": 3.936150234741784,
      "grad_norm": 0.542384147644043,
      "learning_rate": 1.3084943439887659e-05,
      "loss": 0.0105,
      "step": 2620
    },
    {
      "epoch": 3.9511737089201877,
      "grad_norm": 0.11230123043060303,
      "learning_rate": 1.273288829307579e-05,
      "loss": 0.005,
      "step": 2630
    },
    {
      "epoch": 3.9661971830985916,
      "grad_norm": 0.0764995664358139,
      "learning_rate": 1.2384941788365622e-05,
      "loss": 0.0055,
      "step": 2640
    },
    {
      "epoch": 3.981220657276995,
      "grad_norm": 0.2927321493625641,
      "learning_rate": 1.2041142286327477e-05,
      "loss": 0.0082,
      "step": 2650
    },
    {
      "epoch": 3.981220657276995,
      "eval_loss": 0.05000728368759155,
      "eval_runtime": 196.7082,
      "eval_samples_per_second": 6.019,
      "eval_steps_per_second": 6.019,
      "step": 2650
    },
    {
      "epoch": 3.996244131455399,
      "grad_norm": 0.8166933059692383,
      "learning_rate": 1.170152769033095e-05,
      "loss": 0.0062,
      "step": 2660
    },
    {
      "epoch": 4.011267605633803,
      "grad_norm": 0.20836931467056274,
      "learning_rate": 1.1366135442366127e-05,
      "loss": 0.0056,
      "step": 2670
    },
    {
      "epoch": 4.026291079812206,
      "grad_norm": 0.11336184293031693,
      "learning_rate": 1.103500251891571e-05,
      "loss": 0.0047,
      "step": 2680
    },
    {
      "epoch": 4.041314553990611,
      "grad_norm": 0.03469737619161606,
      "learning_rate": 1.0708165426878325e-05,
      "loss": 0.0028,
      "step": 2690
    },
    {
      "epoch": 4.056338028169014,
      "grad_norm": 0.30168983340263367,
      "learning_rate": 1.0385660199543812e-05,
      "loss": 0.0054,
      "step": 2700
    },
    {
      "epoch": 4.056338028169014,
      "eval_loss": 0.0516323521733284,
      "eval_runtime": 196.8548,
      "eval_samples_per_second": 6.015,
      "eval_steps_per_second": 6.015,
      "step": 2700
    },
    {
      "epoch": 4.0713615023474174,
      "grad_norm": 0.045852866023778915,
      "learning_rate": 1.0067522392620537e-05,
      "loss": 0.0031,
      "step": 2710
    },
    {
      "epoch": 4.086384976525822,
      "grad_norm": 0.2861749529838562,
      "learning_rate": 9.753787080315385e-06,
      "loss": 0.0089,
      "step": 2720
    },
    {
      "epoch": 4.101408450704225,
      "grad_norm": 0.029442287981510162,
      "learning_rate": 9.444488851467042e-06,
      "loss": 0.0025,
      "step": 2730
    },
    {
      "epoch": 4.1164319248826295,
      "grad_norm": 0.16276951134204865,
      "learning_rate": 9.139661805732435e-06,
      "loss": 0.0021,
      "step": 2740
    },
    {
      "epoch": 4.131455399061033,
      "grad_norm": 0.2108067125082016,
      "learning_rate": 8.839339549827397e-06,
      "loss": 0.0019,
      "step": 2750
    },
    {
      "epoch": 4.131455399061033,
      "eval_loss": 0.054593775421381,
      "eval_runtime": 196.8035,
      "eval_samples_per_second": 6.016,
      "eval_steps_per_second": 6.016,
      "step": 2750
    },
    {
      "epoch": 4.146478873239436,
      "grad_norm": 0.02386028692126274,
      "learning_rate": 8.543555193821634e-06,
      "loss": 0.0046,
      "step": 2760
    },
    {
      "epoch": 4.161502347417841,
      "grad_norm": 0.24152906239032745,
      "learning_rate": 8.252341347488251e-06,
      "loss": 0.0025,
      "step": 2770
    },
    {
      "epoch": 4.176525821596244,
      "grad_norm": 0.12346199154853821,
      "learning_rate": 7.965730116708681e-06,
      "loss": 0.0038,
      "step": 2780
    },
    {
      "epoch": 4.191549295774648,
      "grad_norm": 0.0784059390425682,
      "learning_rate": 7.68375309993304e-06,
      "loss": 0.0019,
      "step": 2790
    },
    {
      "epoch": 4.206572769953052,
      "grad_norm": 0.08971384167671204,
      "learning_rate": 7.406441384696372e-06,
      "loss": 0.0048,
      "step": 2800
    },
    {
      "epoch": 4.206572769953052,
      "eval_loss": 0.05651959031820297,
      "eval_runtime": 196.0515,
      "eval_samples_per_second": 6.039,
      "eval_steps_per_second": 6.039,
      "step": 2800
    },
    {
      "epoch": 4.221596244131455,
      "grad_norm": 0.06048822030425072,
      "learning_rate": 7.133825544191464e-06,
      "loss": 0.0053,
      "step": 2810
    },
    {
      "epoch": 4.236619718309859,
      "grad_norm": 0.19797635078430176,
      "learning_rate": 6.865935633897996e-06,
      "loss": 0.002,
      "step": 2820
    },
    {
      "epoch": 4.251643192488263,
      "grad_norm": 0.13878192007541656,
      "learning_rate": 6.602801188269081e-06,
      "loss": 0.0043,
      "step": 2830
    },
    {
      "epoch": 4.266666666666667,
      "grad_norm": 0.12489641457796097,
      "learning_rate": 6.344451217475183e-06,
      "loss": 0.0015,
      "step": 2840
    },
    {
      "epoch": 4.28169014084507,
      "grad_norm": 0.11764870584011078,
      "learning_rate": 6.090914204205655e-06,
      "loss": 0.0026,
      "step": 2850
    },
    {
      "epoch": 4.28169014084507,
      "eval_loss": 0.05827711895108223,
      "eval_runtime": 195.8018,
      "eval_samples_per_second": 6.047,
      "eval_steps_per_second": 6.047,
      "step": 2850
    },
    {
      "epoch": 4.296713615023474,
      "grad_norm": 0.07948636263608932,
      "learning_rate": 5.842218100528679e-06,
      "loss": 0.0018,
      "step": 2860
    },
    {
      "epoch": 4.311737089201878,
      "grad_norm": 0.11953713744878769,
      "learning_rate": 5.598390324809555e-06,
      "loss": 0.0026,
      "step": 2870
    },
    {
      "epoch": 4.326760563380281,
      "grad_norm": 0.22272926568984985,
      "learning_rate": 5.359457758687841e-06,
      "loss": 0.0028,
      "step": 2880
    },
    {
      "epoch": 4.341784037558686,
      "grad_norm": 0.5969500541687012,
      "learning_rate": 5.125446744113743e-06,
      "loss": 0.0028,
      "step": 2890
    },
    {
      "epoch": 4.356807511737089,
      "grad_norm": 0.025193296372890472,
      "learning_rate": 4.896383080443934e-06,
      "loss": 0.001,
      "step": 2900
    },
    {
      "epoch": 4.356807511737089,
      "eval_loss": 0.06079353392124176,
      "eval_runtime": 196.1461,
      "eval_samples_per_second": 6.036,
      "eval_steps_per_second": 6.036,
      "step": 2900
    },
    {
      "epoch": 4.371830985915493,
      "grad_norm": 0.2489527016878128,
      "learning_rate": 4.672292021597174e-06,
      "loss": 0.004,
      "step": 2910
    },
    {
      "epoch": 4.386854460093897,
      "grad_norm": 0.10525476932525635,
      "learning_rate": 4.4531982732702145e-06,
      "loss": 0.0037,
      "step": 2920
    },
    {
      "epoch": 4.4018779342723,
      "grad_norm": 0.08666111528873444,
      "learning_rate": 4.239125990213883e-06,
      "loss": 0.0019,
      "step": 2930
    },
    {
      "epoch": 4.416901408450705,
      "grad_norm": 0.15897853672504425,
      "learning_rate": 4.030098773570174e-06,
      "loss": 0.0006,
      "step": 2940
    },
    {
      "epoch": 4.431924882629108,
      "grad_norm": 0.05585940182209015,
      "learning_rate": 3.826139668270234e-06,
      "loss": 0.0022,
      "step": 2950
    },
    {
      "epoch": 4.431924882629108,
      "eval_loss": 0.06087281182408333,
      "eval_runtime": 195.9838,
      "eval_samples_per_second": 6.041,
      "eval_steps_per_second": 6.041,
      "step": 2950
    },
    {
      "epoch": 4.446948356807511,
      "grad_norm": 0.27004674077033997,
      "learning_rate": 3.6272711604936504e-06,
      "loss": 0.0022,
      "step": 2960
    },
    {
      "epoch": 4.461971830985916,
      "grad_norm": 0.02100216969847679,
      "learning_rate": 3.433515175189428e-06,
      "loss": 0.0021,
      "step": 2970
    },
    {
      "epoch": 4.476995305164319,
      "grad_norm": 0.023518303409218788,
      "learning_rate": 3.2448930736588e-06,
      "loss": 0.0063,
      "step": 2980
    },
    {
      "epoch": 4.492018779342723,
      "grad_norm": 0.04742588847875595,
      "learning_rate": 3.061425651200117e-06,
      "loss": 0.0021,
      "step": 2990
    },
    {
      "epoch": 4.507042253521127,
      "grad_norm": 0.19590170681476593,
      "learning_rate": 2.883133134816296e-06,
      "loss": 0.0036,
      "step": 3000
    },
    {
      "epoch": 4.507042253521127,
      "eval_loss": 0.06154530867934227,
      "eval_runtime": 196.0968,
      "eval_samples_per_second": 6.038,
      "eval_steps_per_second": 6.038,
      "step": 3000
    },
    {
      "epoch": 4.52206572769953,
      "grad_norm": 0.02329372987151146,
      "learning_rate": 2.7100351809847326e-06,
      "loss": 0.0018,
      "step": 3010
    },
    {
      "epoch": 4.537089201877935,
      "grad_norm": 0.034720126539468765,
      "learning_rate": 2.542150873490251e-06,
      "loss": 0.0013,
      "step": 3020
    },
    {
      "epoch": 4.552112676056338,
      "grad_norm": 0.1508730798959732,
      "learning_rate": 2.3794987213211383e-06,
      "loss": 0.0025,
      "step": 3030
    },
    {
      "epoch": 4.567136150234742,
      "grad_norm": 0.06496744602918625,
      "learning_rate": 2.222096656628547e-06,
      "loss": 0.0036,
      "step": 3040
    },
    {
      "epoch": 4.582159624413146,
      "grad_norm": 0.024426713585853577,
      "learning_rate": 2.0699620327495174e-06,
      "loss": 0.0006,
      "step": 3050
    },
    {
      "epoch": 4.582159624413146,
      "eval_loss": 0.06214800477027893,
      "eval_runtime": 194.9732,
      "eval_samples_per_second": 6.073,
      "eval_steps_per_second": 6.073,
      "step": 3050
    },
    {
      "epoch": 4.597183098591549,
      "grad_norm": 0.006804203614592552,
      "learning_rate": 1.9231116222937996e-06,
      "loss": 0.0021,
      "step": 3060
    },
    {
      "epoch": 4.612206572769953,
      "grad_norm": 0.043833401054143906,
      "learning_rate": 1.7815616152946523e-06,
      "loss": 0.0024,
      "step": 3070
    },
    {
      "epoch": 4.627230046948357,
      "grad_norm": 0.04456528648734093,
      "learning_rate": 1.6453276174240195e-06,
      "loss": 0.0018,
      "step": 3080
    },
    {
      "epoch": 4.642253521126761,
      "grad_norm": 0.06189090386033058,
      "learning_rate": 1.5144246482719114e-06,
      "loss": 0.0032,
      "step": 3090
    },
    {
      "epoch": 4.657276995305164,
      "grad_norm": 0.009907769970595837,
      "learning_rate": 1.3888671396905805e-06,
      "loss": 0.0019,
      "step": 3100
    },
    {
      "epoch": 4.657276995305164,
      "eval_loss": 0.06242601200938225,
      "eval_runtime": 195.4706,
      "eval_samples_per_second": 6.057,
      "eval_steps_per_second": 6.057,
      "step": 3100
    },
    {
      "epoch": 4.672300469483568,
      "grad_norm": 0.3474283516407013,
      "learning_rate": 1.2686689342034431e-06,
      "loss": 0.006,
      "step": 3110
    },
    {
      "epoch": 4.687323943661972,
      "grad_norm": 0.6320505738258362,
      "learning_rate": 1.1538432834789227e-06,
      "loss": 0.0032,
      "step": 3120
    },
    {
      "epoch": 4.702347417840375,
      "grad_norm": 0.22293131053447723,
      "learning_rate": 1.044402846869491e-06,
      "loss": 0.0021,
      "step": 3130
    },
    {
      "epoch": 4.71737089201878,
      "grad_norm": 0.3564072549343109,
      "learning_rate": 9.403596900160073e-07,
      "loss": 0.0013,
      "step": 3140
    },
    {
      "epoch": 4.732394366197183,
      "grad_norm": 0.2784675359725952,
      "learning_rate": 8.417252835174749e-07,
      "loss": 0.0027,
      "step": 3150
    },
    {
      "epoch": 4.732394366197183,
      "eval_loss": 0.06278736144304276,
      "eval_runtime": 195.5233,
      "eval_samples_per_second": 6.056,
      "eval_steps_per_second": 6.056,
      "step": 3150
    },
    {
      "epoch": 4.7474178403755865,
      "grad_norm": 0.29395076632499695,
      "learning_rate": 7.48510501666455e-07,
      "loss": 0.0023,
      "step": 3160
    },
    {
      "epoch": 4.762441314553991,
      "grad_norm": 0.17587997019290924,
      "learning_rate": 6.607256212501578e-07,
      "loss": 0.002,
      "step": 3170
    },
    {
      "epoch": 4.777464788732394,
      "grad_norm": 0.40092355012893677,
      "learning_rate": 5.783803204174654e-07,
      "loss": 0.0043,
      "step": 3180
    },
    {
      "epoch": 4.792488262910798,
      "grad_norm": 0.10170930624008179,
      "learning_rate": 5.014836776119358e-07,
      "loss": 0.0026,
      "step": 3190
    },
    {
      "epoch": 4.807511737089202,
      "grad_norm": 0.29708781838417053,
      "learning_rate": 4.300441705708924e-07,
      "loss": 0.0022,
      "step": 3200
    },
    {
      "epoch": 4.807511737089202,
      "eval_loss": 0.06290177255868912,
      "eval_runtime": 195.5195,
      "eval_samples_per_second": 6.056,
      "eval_steps_per_second": 6.056,
      "step": 3200
    },
    {
      "epoch": 4.822535211267605,
      "grad_norm": 0.205259770154953,
      "learning_rate": 3.6406967539078796e-07,
      "loss": 0.0017,
      "step": 3210
    },
    {
      "epoch": 4.83755868544601,
      "grad_norm": 0.03722928464412689,
      "learning_rate": 3.0356746565887715e-07,
      "loss": 0.0038,
      "step": 3220
    },
    {
      "epoch": 4.852582159624413,
      "grad_norm": 0.17495347559452057,
      "learning_rate": 2.485442116513026e-07,
      "loss": 0.0015,
      "step": 3230
    },
    {
      "epoch": 4.867605633802817,
      "grad_norm": 0.0666784942150116,
      "learning_rate": 1.9900597959770507e-07,
      "loss": 0.005,
      "step": 3240
    },
    {
      "epoch": 4.882629107981221,
      "grad_norm": 0.014443851076066494,
      "learning_rate": 1.5495823101245866e-07,
      "loss": 0.002,
      "step": 3250
    },
    {
      "epoch": 4.882629107981221,
      "eval_loss": 0.06296313554048538,
      "eval_runtime": 194.8864,
      "eval_samples_per_second": 6.075,
      "eval_steps_per_second": 6.075,
      "step": 3250
    },
    {
      "epoch": 4.897652582159624,
      "grad_norm": 0.10799351334571838,
      "learning_rate": 1.164058220925135e-07,
      "loss": 0.0024,
      "step": 3260
    },
    {
      "epoch": 4.912676056338028,
      "grad_norm": 0.12190629541873932,
      "learning_rate": 8.335300318201844e-08,
      "loss": 0.0054,
      "step": 3270
    },
    {
      "epoch": 4.927699530516432,
      "grad_norm": 0.4779601991176605,
      "learning_rate": 5.5803418303745917e-08,
      "loss": 0.0051,
      "step": 3280
    },
    {
      "epoch": 4.942723004694836,
      "grad_norm": 0.051619164645671844,
      "learning_rate": 3.3760104757313284e-08,
      "loss": 0.0009,
      "step": 3290
    },
    {
      "epoch": 4.957746478873239,
      "grad_norm": 0.03362768143415451,
      "learning_rate": 1.7225492784345156e-08,
      "loss": 0.0063,
      "step": 3300
    },
    {
      "epoch": 4.957746478873239,
      "eval_loss": 0.06280913203954697,
      "eval_runtime": 194.8282,
      "eval_samples_per_second": 6.077,
      "eval_steps_per_second": 6.077,
      "step": 3300
    },
    {
      "epoch": 4.972769953051643,
      "grad_norm": 0.059018392115831375,
      "learning_rate": 6.201405300532148e-09,
      "loss": 0.0035,
      "step": 3310
    },
    {
      "epoch": 4.987793427230047,
      "grad_norm": 0.08097545057535172,
      "learning_rate": 6.890576946805282e-10,
      "loss": 0.0008,
      "step": 3320
    },
    {
      "epoch": 4.995305164319249,
      "step": 3325,
      "total_flos": 8.345060490986127e+17,
      "train_loss": 0.041049694813843955,
      "train_runtime": 47877.1291,
      "train_samples_per_second": 1.112,
      "train_steps_per_second": 0.069
    }
  ],
  "logging_steps": 10,
  "max_steps": 3325,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 8.345060490986127e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}