{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12679127725856698,
  "eval_steps": 500,
  "global_step": 2035,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006230529595015577,
      "grad_norm": 63.15375442917419,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.372981071472168,
      "step": 10
    },
    {
      "epoch": 0.0012461059190031153,
      "grad_norm": 63.39470944322245,
      "learning_rate": 3.921348314606742e-06,
      "loss": 0.8310403823852539,
      "step": 20
    },
    {
      "epoch": 0.001869158878504673,
      "grad_norm": 11.96310343076372,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.7284352779388428,
      "step": 30
    },
    {
      "epoch": 0.0024922118380062306,
      "grad_norm": 159.35287956464992,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.6918130874633789,
      "step": 40
    },
    {
      "epoch": 0.003115264797507788,
      "grad_norm": 140.43690439070613,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.6914069652557373,
      "step": 50
    },
    {
      "epoch": 0.003738317757009346,
      "grad_norm": 20.04034887568955,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.6684707641601563,
      "step": 60
    },
    {
      "epoch": 0.004361370716510903,
      "grad_norm": 146.62158267929294,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.6722880363464355,
      "step": 70
    },
    {
      "epoch": 0.004984423676012461,
      "grad_norm": 48.33263649592231,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.7312785625457764,
      "step": 80
    },
    {
      "epoch": 0.005607476635514018,
      "grad_norm": 96.97131898292652,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.7157273292541504,
      "step": 90
    },
    {
      "epoch": 0.006230529595015576,
      "grad_norm": 95.5113244321525,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.7347418785095214,
      "step": 100
    },
    {
      "epoch": 0.006853582554517134,
      "grad_norm": 17.213242453649954,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.7522170066833496,
      "step": 110
    },
    {
      "epoch": 0.007476635514018692,
      "grad_norm": 31.300913667908624,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.7286031246185303,
      "step": 120
    },
    {
      "epoch": 0.00809968847352025,
      "grad_norm": 69.2417895023605,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.6891061305999756,
      "step": 130
    },
    {
      "epoch": 0.008722741433021807,
      "grad_norm": 20.324536486476994,
      "learning_rate": 1.60561797752809e-05,
      "loss": 1.3122027397155762,
      "step": 140
    },
    {
      "epoch": 0.009345794392523364,
      "grad_norm": 4.902224214231,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.7708694458007812,
      "step": 150
    },
    {
      "epoch": 0.009968847352024923,
      "grad_norm": 14.386566577478696,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.7179630279541016,
      "step": 160
    },
    {
      "epoch": 0.01059190031152648,
      "grad_norm": 7.998488720594685,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.684850025177002,
      "step": 170
    },
    {
      "epoch": 0.011214953271028037,
      "grad_norm": 10.20788593785828,
      "learning_rate": 2e-05,
      "loss": 0.7212257385253906,
      "step": 180
    },
    {
      "epoch": 0.011838006230529595,
      "grad_norm": 15.093739760534373,
      "learning_rate": 2e-05,
      "loss": 0.7274382591247559,
      "step": 190
    },
    {
      "epoch": 0.012461059190031152,
      "grad_norm": 7.624987181272481,
      "learning_rate": 2e-05,
      "loss": 0.6997462272644043,
      "step": 200
    },
    {
      "epoch": 0.013084112149532711,
      "grad_norm": 38.20889799071737,
      "learning_rate": 2e-05,
      "loss": 0.8538092613220215,
      "step": 210
    },
    {
      "epoch": 0.013707165109034268,
      "grad_norm": 5.610917778335814,
      "learning_rate": 2e-05,
      "loss": 0.7119334697723388,
      "step": 220
    },
    {
      "epoch": 0.014330218068535825,
      "grad_norm": 7.071002934278071,
      "learning_rate": 2e-05,
      "loss": 0.6945069313049317,
      "step": 230
    },
    {
      "epoch": 0.014953271028037384,
      "grad_norm": 1.9779972566834714,
      "learning_rate": 2e-05,
      "loss": 0.6839393615722656,
      "step": 240
    },
    {
      "epoch": 0.01557632398753894,
      "grad_norm": 4.542653465163337,
      "learning_rate": 2e-05,
      "loss": 0.6552486419677734,
      "step": 250
    },
    {
      "epoch": 0.0161993769470405,
      "grad_norm": 4.1326478566474645,
      "learning_rate": 2e-05,
      "loss": 0.7250523567199707,
      "step": 260
    },
    {
      "epoch": 0.016822429906542057,
      "grad_norm": 2.71061736504502,
      "learning_rate": 2e-05,
      "loss": 0.6365773677825928,
      "step": 270
    },
    {
      "epoch": 0.017445482866043614,
      "grad_norm": 2.4835202986671043,
      "learning_rate": 2e-05,
      "loss": 0.6632792472839355,
      "step": 280
    },
    {
      "epoch": 0.01806853582554517,
      "grad_norm": 22.09800173446232,
      "learning_rate": 2e-05,
      "loss": 0.663812780380249,
      "step": 290
    },
    {
      "epoch": 0.018691588785046728,
      "grad_norm": 15.489771776735218,
      "learning_rate": 2e-05,
      "loss": 0.650740385055542,
      "step": 300
    },
    {
      "epoch": 0.019314641744548288,
      "grad_norm": 1.8136045854235476,
      "learning_rate": 2e-05,
      "loss": 0.6493139266967773,
      "step": 310
    },
    {
      "epoch": 0.019937694704049845,
      "grad_norm": 1.4242810637346537,
      "learning_rate": 2e-05,
      "loss": 0.6703683376312256,
      "step": 320
    },
    {
      "epoch": 0.020560747663551402,
      "grad_norm": 2.01645414429854,
      "learning_rate": 2e-05,
      "loss": 0.6692399501800537,
      "step": 330
    },
    {
      "epoch": 0.02118380062305296,
      "grad_norm": 1.3748624267038327,
      "learning_rate": 2e-05,
      "loss": 0.6543716430664063,
      "step": 340
    },
    {
      "epoch": 0.021806853582554516,
      "grad_norm": 3.7632290163066595,
      "learning_rate": 2e-05,
      "loss": 0.6510223388671875,
      "step": 350
    },
    {
      "epoch": 0.022429906542056073,
      "grad_norm": 3.069062402265582,
      "learning_rate": 2e-05,
      "loss": 0.6401685237884521,
      "step": 360
    },
    {
      "epoch": 0.023052959501557634,
      "grad_norm": 3.9242151311515756,
      "learning_rate": 2e-05,
      "loss": 0.650885534286499,
      "step": 370
    },
    {
      "epoch": 0.02367601246105919,
      "grad_norm": 2.5645994240054146,
      "learning_rate": 2e-05,
      "loss": 0.6438897132873536,
      "step": 380
    },
    {
      "epoch": 0.024299065420560748,
      "grad_norm": 3.3264331329128587,
      "learning_rate": 2e-05,
      "loss": 0.6713788032531738,
      "step": 390
    },
    {
      "epoch": 0.024922118380062305,
      "grad_norm": 5.952680558627565,
      "learning_rate": 2e-05,
      "loss": 0.695007610321045,
      "step": 400
    },
    {
      "epoch": 0.02554517133956386,
      "grad_norm": 7.2640539794724965,
      "learning_rate": 2e-05,
      "loss": 0.6785341262817383,
      "step": 410
    },
    {
      "epoch": 0.026168224299065422,
      "grad_norm": 12.065849294819978,
      "learning_rate": 2e-05,
      "loss": 0.6561185836791992,
      "step": 420
    },
    {
      "epoch": 0.02679127725856698,
      "grad_norm": 2.9882502235892647,
      "learning_rate": 2e-05,
      "loss": 0.6439939498901367,
      "step": 430
    },
    {
      "epoch": 0.027414330218068536,
      "grad_norm": 2.919046299610538,
      "learning_rate": 2e-05,
      "loss": 0.6663602352142334,
      "step": 440
    },
    {
      "epoch": 0.028037383177570093,
      "grad_norm": 4.189627201982762,
      "learning_rate": 2e-05,
      "loss": 0.6693576812744141,
      "step": 450
    },
    {
      "epoch": 0.02866043613707165,
      "grad_norm": 7.890660674041811,
      "learning_rate": 2e-05,
      "loss": 0.6684383392333985,
      "step": 460
    },
    {
      "epoch": 0.029283489096573207,
      "grad_norm": 82.91478815091692,
      "learning_rate": 2e-05,
      "loss": 0.7290635585784913,
      "step": 470
    },
    {
      "epoch": 0.029906542056074768,
      "grad_norm": 3.333359402201608,
      "learning_rate": 2e-05,
      "loss": 0.6626537322998047,
      "step": 480
    },
    {
      "epoch": 0.030529595015576325,
      "grad_norm": 2.4351561694941313,
      "learning_rate": 2e-05,
      "loss": 0.6529546260833741,
      "step": 490
    },
    {
      "epoch": 0.03115264797507788,
      "grad_norm": 3.5168056840413233,
      "learning_rate": 2e-05,
      "loss": 0.6785277843475341,
      "step": 500
    },
    {
      "epoch": 0.03177570093457944,
      "grad_norm": 64.52415387460657,
      "learning_rate": 2e-05,
      "loss": 0.6317290306091309,
      "step": 510
    },
    {
      "epoch": 0.032398753894081,
      "grad_norm": 1.9567345855965799,
      "learning_rate": 2e-05,
      "loss": 0.6504751205444336,
      "step": 520
    },
    {
      "epoch": 0.03302180685358255,
      "grad_norm": 2.4483916233070815,
      "learning_rate": 2e-05,
      "loss": 0.6714927673339843,
      "step": 530
    },
    {
      "epoch": 0.03364485981308411,
      "grad_norm": 6.84493838576302,
      "learning_rate": 2e-05,
      "loss": 0.6702317714691162,
      "step": 540
    },
    {
      "epoch": 0.03426791277258567,
      "grad_norm": 1.6302467458157637,
      "learning_rate": 2e-05,
      "loss": 0.6749137878417969,
      "step": 550
    },
    {
      "epoch": 0.03489096573208723,
      "grad_norm": 1.3830723147413793,
      "learning_rate": 2e-05,
      "loss": 0.6726428985595703,
      "step": 560
    },
    {
      "epoch": 0.03551401869158879,
      "grad_norm": 1.816565044561178,
      "learning_rate": 2e-05,
      "loss": 0.6633450984954834,
      "step": 570
    },
    {
      "epoch": 0.03613707165109034,
      "grad_norm": 1.4154803760531598,
      "learning_rate": 2e-05,
      "loss": 0.6670496940612793,
      "step": 580
    },
    {
      "epoch": 0.0367601246105919,
      "grad_norm": 1.441388864695392,
      "learning_rate": 2e-05,
      "loss": 0.66356520652771,
      "step": 590
    },
    {
      "epoch": 0.037383177570093455,
      "grad_norm": 5.297501644827849,
      "learning_rate": 2e-05,
      "loss": 0.6568848609924316,
      "step": 600
    },
    {
      "epoch": 0.038006230529595016,
      "grad_norm": 1.3268959282636765,
      "learning_rate": 2e-05,
      "loss": 0.6364224433898926,
      "step": 610
    },
    {
      "epoch": 0.038629283489096576,
      "grad_norm": 3.1357427218379725,
      "learning_rate": 2e-05,
      "loss": 0.6608361721038818,
      "step": 620
    },
    {
      "epoch": 0.03925233644859813,
      "grad_norm": 3.2870153380544496,
      "learning_rate": 2e-05,
      "loss": 0.7057211875915528,
      "step": 630
    },
    {
      "epoch": 0.03987538940809969,
      "grad_norm": 2.197431012725625,
      "learning_rate": 2e-05,
      "loss": 0.6897117614746093,
      "step": 640
    },
    {
      "epoch": 0.040498442367601244,
      "grad_norm": 8.502549125517568,
      "learning_rate": 2e-05,
      "loss": 0.6920805931091308,
      "step": 650
    },
    {
      "epoch": 0.041121495327102804,
      "grad_norm": 2.462320802295259,
      "learning_rate": 2e-05,
      "loss": 0.7063193321228027,
      "step": 660
    },
    {
      "epoch": 0.041744548286604365,
      "grad_norm": 5.683201827218238,
      "learning_rate": 2e-05,
      "loss": 0.6920748710632324,
      "step": 670
    },
    {
      "epoch": 0.04236760124610592,
      "grad_norm": 3.027115327083748,
      "learning_rate": 2e-05,
      "loss": 0.6580988883972168,
      "step": 680
    },
    {
      "epoch": 0.04299065420560748,
      "grad_norm": 9.069631138625143,
      "learning_rate": 2e-05,
      "loss": 0.6772819519042969,
      "step": 690
    },
    {
      "epoch": 0.04361370716510903,
      "grad_norm": 3.5222303907627026,
      "learning_rate": 2e-05,
      "loss": 0.6700127601623536,
      "step": 700
    },
    {
      "epoch": 0.04423676012461059,
      "grad_norm": 2.539625946046831,
      "learning_rate": 2e-05,
      "loss": 0.6825683116912842,
      "step": 710
    },
    {
      "epoch": 0.044859813084112146,
      "grad_norm": 7.530973621569606,
      "learning_rate": 2e-05,
      "loss": 0.6668391227722168,
      "step": 720
    },
    {
      "epoch": 0.04548286604361371,
      "grad_norm": 3.10826200653366,
      "learning_rate": 2e-05,
      "loss": 0.6900657176971435,
      "step": 730
    },
    {
      "epoch": 0.04610591900311527,
      "grad_norm": 3.021206626575616,
      "learning_rate": 2e-05,
      "loss": 0.666568374633789,
      "step": 740
    },
    {
      "epoch": 0.04672897196261682,
      "grad_norm": 3.4063444253551793,
      "learning_rate": 2e-05,
      "loss": 0.6786983489990235,
      "step": 750
    },
    {
      "epoch": 0.04735202492211838,
      "grad_norm": 5.674285755423112,
      "learning_rate": 2e-05,
      "loss": 0.6631364822387695,
      "step": 760
    },
    {
      "epoch": 0.047975077881619935,
      "grad_norm": 5.447979039303706,
      "learning_rate": 2e-05,
      "loss": 0.6851268768310547,
      "step": 770
    },
    {
      "epoch": 0.048598130841121495,
      "grad_norm": 2.0548995784325275,
      "learning_rate": 2e-05,
      "loss": 0.6662421703338623,
      "step": 780
    },
    {
      "epoch": 0.049221183800623056,
      "grad_norm": 1.2276649895732483,
      "learning_rate": 2e-05,
      "loss": 0.6671768188476562,
      "step": 790
    },
    {
      "epoch": 0.04984423676012461,
      "grad_norm": 1.5987199277283939,
      "learning_rate": 2e-05,
      "loss": 0.6489299774169922,
      "step": 800
    },
    {
      "epoch": 0.05046728971962617,
      "grad_norm": 22.9031846877308,
      "learning_rate": 2e-05,
      "loss": 0.6675164699554443,
      "step": 810
    },
    {
      "epoch": 0.05109034267912772,
      "grad_norm": 2.9034364102338985,
      "learning_rate": 2e-05,
      "loss": 0.6826793670654296,
      "step": 820
    },
    {
      "epoch": 0.051713395638629284,
      "grad_norm": 1.8432534502921918,
      "learning_rate": 2e-05,
      "loss": 0.6763291835784913,
      "step": 830
    },
    {
      "epoch": 0.052336448598130844,
      "grad_norm": 2.0417799918298596,
      "learning_rate": 2e-05,
      "loss": 0.6497424125671387,
      "step": 840
    },
    {
      "epoch": 0.0529595015576324,
      "grad_norm": 22.32540564160995,
      "learning_rate": 2e-05,
      "loss": 0.6354857444763183,
      "step": 850
    },
    {
      "epoch": 0.05358255451713396,
      "grad_norm": 6.229762616389806,
      "learning_rate": 2e-05,
      "loss": 0.657370662689209,
      "step": 860
    },
    {
      "epoch": 0.05420560747663551,
      "grad_norm": 3.23998905516123,
      "learning_rate": 2e-05,
      "loss": 0.6524142265319824,
      "step": 870
    },
    {
      "epoch": 0.05482866043613707,
      "grad_norm": 2.8784021094456604,
      "learning_rate": 2e-05,
      "loss": 0.653789472579956,
      "step": 880
    },
    {
      "epoch": 0.05545171339563863,
      "grad_norm": 1.40857663500608,
      "learning_rate": 2e-05,
      "loss": 0.6703225612640381,
      "step": 890
    },
    {
      "epoch": 0.056074766355140186,
      "grad_norm": 5.080917667561178,
      "learning_rate": 2e-05,
      "loss": 0.6439192771911622,
      "step": 900
    },
    {
      "epoch": 0.05669781931464175,
      "grad_norm": 4.5764665849587445,
      "learning_rate": 2e-05,
      "loss": 0.6634146690368652,
      "step": 910
    },
    {
      "epoch": 0.0573208722741433,
      "grad_norm": 3.2037710370417023,
      "learning_rate": 2e-05,
      "loss": 0.6824374198913574,
      "step": 920
    },
    {
      "epoch": 0.05794392523364486,
      "grad_norm": 6.053033149131505,
      "learning_rate": 2e-05,
      "loss": 0.6423258781433105,
      "step": 930
    },
    {
      "epoch": 0.058566978193146414,
      "grad_norm": 13.436486964861611,
      "learning_rate": 2e-05,
      "loss": 0.6521054744720459,
      "step": 940
    },
    {
      "epoch": 0.059190031152647975,
      "grad_norm": 1.4820242941537698,
      "learning_rate": 2e-05,
      "loss": 0.6490562438964844,
      "step": 950
    },
    {
      "epoch": 0.059813084112149535,
      "grad_norm": 3.970984796336633,
      "learning_rate": 2e-05,
      "loss": 0.6425233364105225,
      "step": 960
    },
    {
      "epoch": 0.06043613707165109,
      "grad_norm": 2.606688835969579,
      "learning_rate": 2e-05,
      "loss": 0.6780929088592529,
      "step": 970
    },
    {
      "epoch": 0.06105919003115265,
      "grad_norm": 1.6441407042964533,
      "learning_rate": 2e-05,
      "loss": 0.6381833553314209,
      "step": 980
    },
    {
      "epoch": 0.0616822429906542,
      "grad_norm": 1.9338014114509088,
      "learning_rate": 2e-05,
      "loss": 0.671638298034668,
      "step": 990
    },
    {
      "epoch": 0.06230529595015576,
      "grad_norm": 2.39864419114088,
      "learning_rate": 2e-05,
      "loss": 0.659054183959961,
      "step": 1000
    },
    {
      "epoch": 0.06292834890965732,
      "grad_norm": 4.683446585267705,
      "learning_rate": 2e-05,
      "loss": 0.6786028861999511,
      "step": 1010
    },
    {
      "epoch": 0.06355140186915888,
      "grad_norm": 2.719894383949491,
      "learning_rate": 2e-05,
      "loss": 0.6658808708190918,
      "step": 1020
    },
    {
      "epoch": 0.06417445482866044,
      "grad_norm": 1.1588652494157246,
      "learning_rate": 2e-05,
      "loss": 0.6696397304534912,
      "step": 1030
    },
    {
      "epoch": 0.064797507788162,
      "grad_norm": 1.0813640353054965,
      "learning_rate": 2e-05,
      "loss": 0.6676293849945069,
      "step": 1040
    },
    {
      "epoch": 0.06542056074766354,
      "grad_norm": 1.2271064002015928,
      "learning_rate": 2e-05,
      "loss": 0.6389595985412597,
      "step": 1050
    },
    {
      "epoch": 0.0660436137071651,
      "grad_norm": 2.8431458204110074,
      "learning_rate": 2e-05,
      "loss": 0.6555405616760254,
      "step": 1060
    },
    {
      "epoch": 0.06666666666666667,
      "grad_norm": 3.2188265816780692,
      "learning_rate": 2e-05,
      "loss": 0.6418774604797364,
      "step": 1070
    },
    {
      "epoch": 0.06728971962616823,
      "grad_norm": 9.465426552560904,
      "learning_rate": 2e-05,
      "loss": 0.6477942943572998,
      "step": 1080
    },
    {
      "epoch": 0.06791277258566979,
      "grad_norm": 1.1426314291875073,
      "learning_rate": 2e-05,
      "loss": 0.6486814022064209,
      "step": 1090
    },
    {
      "epoch": 0.06853582554517133,
      "grad_norm": 1.3614517863008955,
      "learning_rate": 2e-05,
      "loss": 0.6399056911468506,
      "step": 1100
    },
    {
      "epoch": 0.0691588785046729,
      "grad_norm": 1.616958777603617,
      "learning_rate": 2e-05,
      "loss": 0.625498104095459,
      "step": 1110
    },
    {
      "epoch": 0.06978193146417445,
      "grad_norm": 1.700623675066532,
      "learning_rate": 2e-05,
      "loss": 0.620833158493042,
      "step": 1120
    },
    {
      "epoch": 0.07040498442367601,
      "grad_norm": 2.5588029176909948,
      "learning_rate": 2e-05,
      "loss": 0.6351099967956543,
      "step": 1130
    },
    {
      "epoch": 0.07102803738317758,
      "grad_norm": 6.556254605023658,
      "learning_rate": 2e-05,
      "loss": 0.6419850826263428,
      "step": 1140
    },
    {
      "epoch": 0.07165109034267912,
      "grad_norm": 2.4751544762969946,
      "learning_rate": 2e-05,
      "loss": 0.6722092628479004,
      "step": 1150
    },
    {
      "epoch": 0.07227414330218068,
      "grad_norm": 1.5727047413222857,
      "learning_rate": 2e-05,
      "loss": 0.6558211326599122,
      "step": 1160
    },
    {
      "epoch": 0.07289719626168224,
      "grad_norm": 9.438322770095112,
      "learning_rate": 2e-05,
      "loss": 0.6519765853881836,
      "step": 1170
    },
    {
      "epoch": 0.0735202492211838,
      "grad_norm": 2.839029932122009,
      "learning_rate": 2e-05,
      "loss": 0.6472910881042481,
      "step": 1180
    },
    {
      "epoch": 0.07414330218068536,
      "grad_norm": 1.3804888984788395,
      "learning_rate": 2e-05,
      "loss": 0.6604276180267334,
      "step": 1190
    },
    {
      "epoch": 0.07476635514018691,
      "grad_norm": 1.0934274555239194,
      "learning_rate": 2e-05,
      "loss": 0.6465075969696045,
      "step": 1200
    },
    {
      "epoch": 0.07538940809968847,
      "grad_norm": 1.8841069578655136,
      "learning_rate": 2e-05,
      "loss": 0.6400872230529785,
      "step": 1210
    },
    {
      "epoch": 0.07601246105919003,
      "grad_norm": 16.12011257195063,
      "learning_rate": 2e-05,
      "loss": 0.655942440032959,
      "step": 1220
    },
    {
      "epoch": 0.07663551401869159,
      "grad_norm": 1.9392045965213456,
      "learning_rate": 2e-05,
      "loss": 0.6418064117431641,
      "step": 1230
    },
    {
      "epoch": 0.07725856697819315,
      "grad_norm": 1.2879388752480592,
      "learning_rate": 2e-05,
      "loss": 0.6534547328948974,
      "step": 1240
    },
    {
      "epoch": 0.0778816199376947,
      "grad_norm": 1.3491269021694952,
      "learning_rate": 2e-05,
      "loss": 0.6325421333312988,
      "step": 1250
    },
    {
      "epoch": 0.07850467289719626,
      "grad_norm": 1.9538179100083624,
      "learning_rate": 2e-05,
      "loss": 0.6555644989013671,
      "step": 1260
    },
    {
      "epoch": 0.07912772585669782,
      "grad_norm": 1.686205721804821,
      "learning_rate": 2e-05,
      "loss": 0.6385934352874756,
      "step": 1270
    },
    {
      "epoch": 0.07975077881619938,
      "grad_norm": 1.4326411502688534,
      "learning_rate": 2e-05,
      "loss": 0.6342977046966553,
      "step": 1280
    },
    {
      "epoch": 0.08037383177570094,
      "grad_norm": 4.860024808022459,
      "learning_rate": 2e-05,
      "loss": 0.6609707832336426,
      "step": 1290
    },
    {
      "epoch": 0.08099688473520249,
      "grad_norm": 1.2371686758942715,
      "learning_rate": 2e-05,
      "loss": 0.6476799964904785,
      "step": 1300
    },
    {
      "epoch": 0.08161993769470405,
      "grad_norm": 1.1597203515801051,
      "learning_rate": 2e-05,
      "loss": 0.6473023891448975,
      "step": 1310
    },
    {
      "epoch": 0.08224299065420561,
      "grad_norm": 1.8273978345410087,
      "learning_rate": 2e-05,
      "loss": 0.6425920009613038,
      "step": 1320
    },
    {
      "epoch": 0.08286604361370717,
      "grad_norm": 1.6088638146203655,
      "learning_rate": 2e-05,
      "loss": 0.6325952053070069,
      "step": 1330
    },
    {
      "epoch": 0.08348909657320873,
      "grad_norm": 1.0395526162178266,
      "learning_rate": 2e-05,
      "loss": 0.6280010223388672,
      "step": 1340
    },
    {
      "epoch": 0.08411214953271028,
      "grad_norm": 1.1776111039430426,
      "learning_rate": 2e-05,
      "loss": 0.6420888900756836,
      "step": 1350
    },
    {
      "epoch": 0.08473520249221184,
      "grad_norm": 1.144362156326936,
      "learning_rate": 2e-05,
      "loss": 0.6447340965270996,
      "step": 1360
    },
    {
      "epoch": 0.0853582554517134,
      "grad_norm": 1.2899619747320272,
      "learning_rate": 2e-05,
      "loss": 0.6319891929626464,
      "step": 1370
    },
    {
      "epoch": 0.08598130841121496,
      "grad_norm": 1.1503180888416502,
      "learning_rate": 2e-05,
      "loss": 0.6263866424560547,
      "step": 1380
    },
    {
      "epoch": 0.08660436137071652,
      "grad_norm": 0.9492449615285058,
      "learning_rate": 2e-05,
      "loss": 0.6156892776489258,
      "step": 1390
    },
    {
      "epoch": 0.08722741433021806,
      "grad_norm": 0.9840501517202644,
      "learning_rate": 2e-05,
      "loss": 0.6201627731323243,
      "step": 1400
    },
    {
      "epoch": 0.08785046728971962,
      "grad_norm": 0.9663965412569069,
      "learning_rate": 2e-05,
      "loss": 0.618841552734375,
      "step": 1410
    },
    {
      "epoch": 0.08847352024922119,
      "grad_norm": 1.2024904063187882,
      "learning_rate": 2e-05,
      "loss": 0.6406625270843506,
      "step": 1420
    },
    {
      "epoch": 0.08909657320872275,
      "grad_norm": 0.9573647223711624,
      "learning_rate": 2e-05,
      "loss": 0.6550570487976074,
      "step": 1430
    },
    {
      "epoch": 0.08971962616822429,
      "grad_norm": 1.0307197954823704,
      "learning_rate": 2e-05,
      "loss": 0.6338574886322021,
      "step": 1440
    },
    {
      "epoch": 0.09034267912772585,
      "grad_norm": 2.305281565846579,
      "learning_rate": 2e-05,
      "loss": 0.6363538265228271,
      "step": 1450
    },
    {
      "epoch": 0.09096573208722741,
      "grad_norm": 1.4667417153142859,
      "learning_rate": 2e-05,
      "loss": 0.6235406875610352,
      "step": 1460
    },
    {
      "epoch": 0.09158878504672897,
      "grad_norm": 2.9584530551772437,
      "learning_rate": 2e-05,
      "loss": 0.6403841018676758,
      "step": 1470
    },
    {
      "epoch": 0.09221183800623053,
      "grad_norm": 1.112253956223209,
      "learning_rate": 2e-05,
      "loss": 0.6414080619812011,
      "step": 1480
    },
    {
      "epoch": 0.09283489096573208,
      "grad_norm": 1.197396557670252,
      "learning_rate": 2e-05,
      "loss": 0.6136775970458984,
      "step": 1490
    },
    {
      "epoch": 0.09345794392523364,
      "grad_norm": 0.930725260856421,
      "learning_rate": 2e-05,
      "loss": 0.6281547546386719,
      "step": 1500
    },
    {
      "epoch": 0.0940809968847352,
      "grad_norm": 2.246282067394849,
      "learning_rate": 2e-05,
      "loss": 0.6232186317443847,
      "step": 1510
    },
    {
      "epoch": 0.09470404984423676,
      "grad_norm": 1.811695916729897,
      "learning_rate": 2e-05,
      "loss": 0.6236707210540772,
      "step": 1520
    },
    {
      "epoch": 0.09532710280373832,
      "grad_norm": 1.8453018493553404,
      "learning_rate": 2e-05,
      "loss": 0.6396072387695313,
      "step": 1530
    },
    {
      "epoch": 0.09595015576323987,
      "grad_norm": 1.4810091845213549,
      "learning_rate": 2e-05,
      "loss": 0.6346172332763672,
      "step": 1540
    },
    {
      "epoch": 0.09657320872274143,
      "grad_norm": 1.1558342150707985,
      "learning_rate": 2e-05,
      "loss": 0.6340350151062012,
      "step": 1550
    },
    {
      "epoch": 0.09719626168224299,
      "grad_norm": 1.1277930144534245,
      "learning_rate": 2e-05,
      "loss": 0.6178334712982178,
      "step": 1560
    },
    {
      "epoch": 0.09781931464174455,
      "grad_norm": 1.6697265802855836,
      "learning_rate": 2e-05,
      "loss": 0.6273056030273437,
      "step": 1570
    },
    {
      "epoch": 0.09844236760124611,
      "grad_norm": 1.1996090977492178,
      "learning_rate": 2e-05,
      "loss": 0.6388472557067871,
      "step": 1580
    },
    {
      "epoch": 0.09906542056074766,
      "grad_norm": 4.193851985080715,
      "learning_rate": 2e-05,
      "loss": 0.6388599395751953,
      "step": 1590
    },
    {
      "epoch": 0.09968847352024922,
      "grad_norm": 1.3073390628519994,
      "learning_rate": 2e-05,
      "loss": 0.6429434776306152,
      "step": 1600
    },
    {
      "epoch": 0.10031152647975078,
      "grad_norm": 1.5006350300856774,
      "learning_rate": 2e-05,
      "loss": 0.6440896034240723,
      "step": 1610
    },
    {
      "epoch": 0.10093457943925234,
      "grad_norm": 1.2275982378982346,
      "learning_rate": 2e-05,
      "loss": 0.6285616874694824,
      "step": 1620
    },
    {
      "epoch": 0.1015576323987539,
      "grad_norm": 1.1626753929589895,
      "learning_rate": 2e-05,
      "loss": 0.6467322826385498,
      "step": 1630
    },
    {
      "epoch": 0.10218068535825545,
      "grad_norm": 1.8546526164062445,
      "learning_rate": 2e-05,
      "loss": 0.6216934204101563,
      "step": 1640
    },
    {
      "epoch": 0.102803738317757,
      "grad_norm": 4.204227006721846,
      "learning_rate": 2e-05,
      "loss": 0.6037077903747559,
      "step": 1650
    },
    {
      "epoch": 0.10342679127725857,
      "grad_norm": 2.4575617089230892,
      "learning_rate": 2e-05,
      "loss": 0.6549998760223389,
      "step": 1660
    },
    {
      "epoch": 0.10404984423676013,
      "grad_norm": 0.9831225747692173,
      "learning_rate": 2e-05,
      "loss": 0.6362233638763428,
      "step": 1670
    },
    {
      "epoch": 0.10467289719626169,
      "grad_norm": 37.336116705288426,
      "learning_rate": 2e-05,
      "loss": 0.6452502250671387,
      "step": 1680
    },
    {
      "epoch": 0.10529595015576323,
      "grad_norm": 10.151830562748597,
      "learning_rate": 2e-05,
      "loss": 0.6224545478820801,
      "step": 1690
    },
    {
      "epoch": 0.1059190031152648,
      "grad_norm": 1.0474870590574359,
      "learning_rate": 2e-05,
      "loss": 0.6610437393188476,
      "step": 1700
    },
    {
      "epoch": 0.10654205607476636,
      "grad_norm": 3.461782785767983,
      "learning_rate": 2e-05,
      "loss": 0.62818922996521,
      "step": 1710
    },
    {
      "epoch": 0.10716510903426792,
      "grad_norm": 1.0436258381858183,
      "learning_rate": 2e-05,
      "loss": 0.6486543178558349,
      "step": 1720
    },
    {
      "epoch": 0.10778816199376948,
      "grad_norm": 3.870674655703194,
      "learning_rate": 2e-05,
      "loss": 0.6329143524169922,
      "step": 1730
    },
    {
      "epoch": 0.10841121495327102,
      "grad_norm": 0.8873679451298351,
      "learning_rate": 2e-05,
      "loss": 0.6266915798187256,
      "step": 1740
    },
    {
      "epoch": 0.10903426791277258,
      "grad_norm": 1.934393771485732,
      "learning_rate": 2e-05,
      "loss": 0.6206174373626709,
      "step": 1750
    },
    {
      "epoch": 0.10965732087227414,
      "grad_norm": 0.9788497030191609,
      "learning_rate": 2e-05,
      "loss": 0.617664098739624,
      "step": 1760
    },
    {
      "epoch": 0.1102803738317757,
      "grad_norm": 1.3294647839474172,
      "learning_rate": 2e-05,
      "loss": 0.6125317573547363,
      "step": 1770
    },
    {
      "epoch": 0.11090342679127727,
      "grad_norm": 0.9114337186956374,
      "learning_rate": 2e-05,
      "loss": 0.6136180877685546,
      "step": 1780
    },
    {
      "epoch": 0.11152647975077881,
      "grad_norm": 0.9168910070389206,
      "learning_rate": 2e-05,
      "loss": 0.6254988670349121,
      "step": 1790
    },
    {
      "epoch": 0.11214953271028037,
      "grad_norm": 1.2163942422286602,
      "learning_rate": 2e-05,
      "loss": 0.6339495658874512,
      "step": 1800
    },
    {
      "epoch": 0.11277258566978193,
      "grad_norm": 1.0031026506964655,
      "learning_rate": 2e-05,
      "loss": 0.6163856506347656,
      "step": 1810
    },
    {
      "epoch": 0.1133956386292835,
      "grad_norm": 1.4487390671485545,
      "learning_rate": 2e-05,
      "loss": 0.6172023296356202,
      "step": 1820
    },
    {
      "epoch": 0.11401869158878504,
      "grad_norm": 0.971335799491775,
      "learning_rate": 2e-05,
      "loss": 0.603542423248291,
      "step": 1830
    },
    {
      "epoch": 0.1146417445482866,
      "grad_norm": 1.0645808741308516,
      "learning_rate": 2e-05,
      "loss": 0.6314529418945313,
      "step": 1840
    },
    {
      "epoch": 0.11526479750778816,
      "grad_norm": 1.1564622773995759,
      "learning_rate": 2e-05,
      "loss": 0.6358022212982177,
      "step": 1850
    },
    {
      "epoch": 0.11588785046728972,
      "grad_norm": 6.250954738770854,
      "learning_rate": 2e-05,
      "loss": 0.6266725063323975,
      "step": 1860
    },
    {
      "epoch": 0.11651090342679128,
      "grad_norm": 2.485950620772887,
      "learning_rate": 2e-05,
      "loss": 0.6347331047058106,
      "step": 1870
    },
    {
      "epoch": 0.11713395638629283,
      "grad_norm": 1.0491315047430319,
      "learning_rate": 2e-05,
      "loss": 0.6353681087493896,
      "step": 1880
    },
    {
      "epoch": 0.11775700934579439,
      "grad_norm": 0.9344118553673864,
      "learning_rate": 2e-05,
      "loss": 0.6243095397949219,
      "step": 1890
    },
    {
      "epoch": 0.11838006230529595,
      "grad_norm": 1.166751507604473,
      "learning_rate": 2e-05,
      "loss": 0.624453353881836,
      "step": 1900
    },
    {
      "epoch": 0.11900311526479751,
      "grad_norm": 1.1609173008743447,
      "learning_rate": 2e-05,
      "loss": 0.6156741142272949,
      "step": 1910
    },
    {
      "epoch": 0.11962616822429907,
      "grad_norm": 1.2650824767917728,
      "learning_rate": 2e-05,
      "loss": 0.6225091457366944,
      "step": 1920
    },
    {
      "epoch": 0.12024922118380062,
      "grad_norm": 1.323039674561904,
      "learning_rate": 2e-05,
      "loss": 0.6126754760742188,
      "step": 1930
    },
    {
      "epoch": 0.12087227414330218,
      "grad_norm": 10.576149324131405,
      "learning_rate": 2e-05,
      "loss": 0.6002795219421386,
      "step": 1940
    },
    {
      "epoch": 0.12149532710280374,
      "grad_norm": 1.1408899621175108,
      "learning_rate": 2e-05,
      "loss": 0.6268309593200684,
      "step": 1950
    },
    {
      "epoch": 0.1221183800623053,
      "grad_norm": 1.04557318312074,
      "learning_rate": 2e-05,
      "loss": 0.6239619255065918,
      "step": 1960
    },
    {
      "epoch": 0.12274143302180686,
      "grad_norm": 1.2159263913891547,
      "learning_rate": 2e-05,
      "loss": 0.6297917842864991,
      "step": 1970
    },
    {
      "epoch": 0.1233644859813084,
      "grad_norm": 2.784941022644952,
      "learning_rate": 2e-05,
      "loss": 0.6152359008789062,
      "step": 1980
    },
    {
      "epoch": 0.12398753894080997,
      "grad_norm": 0.9318872113204291,
      "learning_rate": 2e-05,
      "loss": 0.6123355388641357,
      "step": 1990
    },
    {
      "epoch": 0.12461059190031153,
      "grad_norm": 1.6289119810248258,
      "learning_rate": 2e-05,
      "loss": 0.6148693084716796,
      "step": 2000
    },
    {
      "epoch": 0.1252336448598131,
      "grad_norm": 6.458481915206926,
      "learning_rate": 2e-05,
      "loss": 0.6158293724060059,
      "step": 2010
    },
    {
      "epoch": 0.12585669781931463,
      "grad_norm": 1.9163812666195066,
      "learning_rate": 2e-05,
      "loss": 0.6355224132537842,
      "step": 2020
    },
    {
      "epoch": 0.1264797507788162,
      "grad_norm": 5.359098164590273,
      "learning_rate": 2e-05,
      "loss": 0.6285907745361328,
      "step": 2030
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 963613493821440.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}