{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7591455615246866,
  "eval_steps": 500,
  "global_step": 5935,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012790995139421847,
      "grad_norm": 623.6208359240255,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.5341346740722657,
      "step": 10
    },
    {
      "epoch": 0.0025581990278843694,
      "grad_norm": 463.91385336199573,
      "learning_rate": 3.921348314606742e-06,
      "loss": 0.9242143630981445,
      "step": 20
    },
    {
      "epoch": 0.003837298541826554,
      "grad_norm": 53.84492921608342,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.7964631080627441,
      "step": 30
    },
    {
      "epoch": 0.005116398055768739,
      "grad_norm": 61.330614407626,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.7263487815856934,
      "step": 40
    },
    {
      "epoch": 0.006395497569710923,
      "grad_norm": 44.648257436551425,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.6948539733886718,
      "step": 50
    },
    {
      "epoch": 0.007674597083653108,
      "grad_norm": 56.621196014184925,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.6822492122650147,
      "step": 60
    },
    {
      "epoch": 0.008953696597595293,
      "grad_norm": 37.47997089126972,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.6963601112365723,
      "step": 70
    },
    {
      "epoch": 0.010232796111537478,
      "grad_norm": 105.99404311228525,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.6244803428649902,
      "step": 80
    },
    {
      "epoch": 0.011511895625479662,
      "grad_norm": 24.305519470285297,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.6451755046844483,
      "step": 90
    },
    {
      "epoch": 0.012790995139421847,
      "grad_norm": 103.85716120098365,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.7046442031860352,
      "step": 100
    },
    {
      "epoch": 0.014070094653364031,
      "grad_norm": 8.153387119451759,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.7193686962127686,
      "step": 110
    },
    {
      "epoch": 0.015349194167306216,
      "grad_norm": 11.299752541462171,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.6899770259857178,
      "step": 120
    },
    {
      "epoch": 0.0166282936812484,
      "grad_norm": 51.069015129101714,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.6754970550537109,
      "step": 130
    },
    {
      "epoch": 0.017907393195190587,
      "grad_norm": 54.175973870128736,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.6890584468841553,
      "step": 140
    },
    {
      "epoch": 0.01918649270913277,
      "grad_norm": 20.136113415923976,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.7157495498657227,
      "step": 150
    },
    {
      "epoch": 0.020465592223074956,
      "grad_norm": 37.80605267456,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.7468127250671387,
      "step": 160
    },
    {
      "epoch": 0.02174469173701714,
      "grad_norm": 10.709328210046946,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.8013706207275391,
      "step": 170
    },
    {
      "epoch": 0.023023791250959325,
      "grad_norm": 13.849160742797585,
      "learning_rate": 2e-05,
      "loss": 0.7073573112487793,
      "step": 180
    },
    {
      "epoch": 0.02430289076490151,
      "grad_norm": 252.71253502007187,
      "learning_rate": 2e-05,
      "loss": 0.7326308727264405,
      "step": 190
    },
    {
      "epoch": 0.025581990278843694,
      "grad_norm": 4.329744526145032,
      "learning_rate": 2e-05,
      "loss": 0.7149838924407959,
      "step": 200
    },
    {
      "epoch": 0.02686108979278588,
      "grad_norm": 6.604126459594562,
      "learning_rate": 2e-05,
      "loss": 0.6777316093444824,
      "step": 210
    },
    {
      "epoch": 0.028140189306728063,
      "grad_norm": 9.52386806224655,
      "learning_rate": 2e-05,
      "loss": 0.6922791957855224,
      "step": 220
    },
    {
      "epoch": 0.02941928882067025,
      "grad_norm": 17.21807673598888,
      "learning_rate": 2e-05,
      "loss": 0.7441821575164795,
      "step": 230
    },
    {
      "epoch": 0.03069838833461243,
      "grad_norm": 9.18143504861921,
      "learning_rate": 2e-05,
      "loss": 0.7066239833831787,
      "step": 240
    },
    {
      "epoch": 0.031977487848554614,
      "grad_norm": 4.217817099649031,
      "learning_rate": 2e-05,
      "loss": 0.6791770458221436,
      "step": 250
    },
    {
      "epoch": 0.0332565873624968,
      "grad_norm": 17.965064875677463,
      "learning_rate": 2e-05,
      "loss": 0.6521565914154053,
      "step": 260
    },
    {
      "epoch": 0.03453568687643899,
      "grad_norm": 96.25438261050074,
      "learning_rate": 2e-05,
      "loss": 0.6563163757324219,
      "step": 270
    },
    {
      "epoch": 0.03581478639038117,
      "grad_norm": 14.572604906608621,
      "learning_rate": 2e-05,
      "loss": 0.6991176128387451,
      "step": 280
    },
    {
      "epoch": 0.03709388590432336,
      "grad_norm": 10.646801341452742,
      "learning_rate": 2e-05,
      "loss": 0.6827552318572998,
      "step": 290
    },
    {
      "epoch": 0.03837298541826554,
      "grad_norm": 5.100529139893381,
      "learning_rate": 2e-05,
      "loss": 0.6454049587249756,
      "step": 300
    },
    {
      "epoch": 0.039652084932207725,
      "grad_norm": 3.7542492731409425,
      "learning_rate": 2e-05,
      "loss": 0.6457695007324219,
      "step": 310
    },
    {
      "epoch": 0.04093118444614991,
      "grad_norm": 4.927798665797291,
      "learning_rate": 2e-05,
      "loss": 0.6363472938537598,
      "step": 320
    },
    {
      "epoch": 0.0422102839600921,
      "grad_norm": 4.6042140802832074,
      "learning_rate": 2e-05,
      "loss": 0.6547730922698974,
      "step": 330
    },
    {
      "epoch": 0.04348938347403428,
      "grad_norm": 3.9790435649800364,
      "learning_rate": 2e-05,
      "loss": 0.6693540573120117,
      "step": 340
    },
    {
      "epoch": 0.04476848298797646,
      "grad_norm": 6.056606530921015,
      "learning_rate": 2e-05,
      "loss": 0.6478125572204589,
      "step": 350
    },
    {
      "epoch": 0.04604758250191865,
      "grad_norm": 5.000014296778777,
      "learning_rate": 2e-05,
      "loss": 0.6575343132019043,
      "step": 360
    },
    {
      "epoch": 0.047326682015860835,
      "grad_norm": 170.7900036194794,
      "learning_rate": 2e-05,
      "loss": 0.6500011444091797,
      "step": 370
    },
    {
      "epoch": 0.04860578152980302,
      "grad_norm": 4.136859247278033,
      "learning_rate": 2e-05,
      "loss": 0.7492189407348633,
      "step": 380
    },
    {
      "epoch": 0.0498848810437452,
      "grad_norm": 4.9635865373650825,
      "learning_rate": 2e-05,
      "loss": 0.7007633209228515,
      "step": 390
    },
    {
      "epoch": 0.05116398055768739,
      "grad_norm": 7.184333434818073,
      "learning_rate": 2e-05,
      "loss": 0.6674163818359375,
      "step": 400
    },
    {
      "epoch": 0.05244308007162957,
      "grad_norm": 10.770350153540436,
      "learning_rate": 2e-05,
      "loss": 0.6679406642913819,
      "step": 410
    },
    {
      "epoch": 0.05372217958557176,
      "grad_norm": 7.497226452066126,
      "learning_rate": 2e-05,
      "loss": 0.7049285888671875,
      "step": 420
    },
    {
      "epoch": 0.05500127909951394,
      "grad_norm": 14.03042351222223,
      "learning_rate": 2e-05,
      "loss": 0.676944637298584,
      "step": 430
    },
    {
      "epoch": 0.056280378613456125,
      "grad_norm": 3.469593844461007,
      "learning_rate": 2e-05,
      "loss": 0.6626708507537842,
      "step": 440
    },
    {
      "epoch": 0.05755947812739831,
      "grad_norm": 3.410706827588382,
      "learning_rate": 2e-05,
      "loss": 0.65382399559021,
      "step": 450
    },
    {
      "epoch": 0.0588385776413405,
      "grad_norm": 1.980498220638673,
      "learning_rate": 2e-05,
      "loss": 0.7313810348510742,
      "step": 460
    },
    {
      "epoch": 0.060117677155282684,
      "grad_norm": 3.739263153638155,
      "learning_rate": 2e-05,
      "loss": 0.6562081336975097,
      "step": 470
    },
    {
      "epoch": 0.06139677666922486,
      "grad_norm": 2.0018607090496596,
      "learning_rate": 2e-05,
      "loss": 0.6454696178436279,
      "step": 480
    },
    {
      "epoch": 0.06267587618316706,
      "grad_norm": 3.4114991222813154,
      "learning_rate": 2e-05,
      "loss": 0.6556525230407715,
      "step": 490
    },
    {
      "epoch": 0.06395497569710923,
      "grad_norm": 3.205444422603881,
      "learning_rate": 2e-05,
      "loss": 0.6443347454071044,
      "step": 500
    },
    {
      "epoch": 0.06523407521105141,
      "grad_norm": 5.095129800632835,
      "learning_rate": 2e-05,
      "loss": 0.6250532150268555,
      "step": 510
    },
    {
      "epoch": 0.0665131747249936,
      "grad_norm": 2.5894252003422684,
      "learning_rate": 2e-05,
      "loss": 0.6742976188659668,
      "step": 520
    },
    {
      "epoch": 0.06779227423893579,
      "grad_norm": 2.8889178959957196,
      "learning_rate": 2e-05,
      "loss": 0.6449506282806396,
      "step": 530
    },
    {
      "epoch": 0.06907137375287797,
      "grad_norm": 3.5369834303098253,
      "learning_rate": 2e-05,
      "loss": 0.6863080024719238,
      "step": 540
    },
    {
      "epoch": 0.07035047326682016,
      "grad_norm": 1.6919645181272793,
      "learning_rate": 2e-05,
      "loss": 0.5828201293945312,
      "step": 550
    },
    {
      "epoch": 0.07162957278076235,
      "grad_norm": 2.7688275845052464,
      "learning_rate": 2e-05,
      "loss": 0.6269129753112793,
      "step": 560
    },
    {
      "epoch": 0.07290867229470453,
      "grad_norm": 9.01577429604241,
      "learning_rate": 2e-05,
      "loss": 0.674321174621582,
      "step": 570
    },
    {
      "epoch": 0.07418777180864672,
      "grad_norm": 1.3605319314128796,
      "learning_rate": 2e-05,
      "loss": 0.6332732200622558,
      "step": 580
    },
    {
      "epoch": 0.07546687132258889,
      "grad_norm": 2.3592211294500034,
      "learning_rate": 2e-05,
      "loss": 0.6280890941619873,
      "step": 590
    },
    {
      "epoch": 0.07674597083653108,
      "grad_norm": 2.1105007442103205,
      "learning_rate": 2e-05,
      "loss": 0.6547659873962403,
      "step": 600
    },
    {
      "epoch": 0.07802507035047326,
      "grad_norm": 1.9882775544536382,
      "learning_rate": 2e-05,
      "loss": 0.6447579860687256,
      "step": 610
    },
    {
      "epoch": 0.07930416986441545,
      "grad_norm": 2.1390953156204024,
      "learning_rate": 2e-05,
      "loss": 0.6662328243255615,
      "step": 620
    },
    {
      "epoch": 0.08058326937835764,
      "grad_norm": 1.822512642342021,
      "learning_rate": 2e-05,
      "loss": 0.6324103832244873,
      "step": 630
    },
    {
      "epoch": 0.08186236889229982,
      "grad_norm": 4.970675015253946,
      "learning_rate": 2e-05,
      "loss": 0.7083280563354493,
      "step": 640
    },
    {
      "epoch": 0.08314146840624201,
      "grad_norm": 1.618436134118923,
      "learning_rate": 2e-05,
      "loss": 0.6183808803558349,
      "step": 650
    },
    {
      "epoch": 0.0844205679201842,
      "grad_norm": 3.040019611591677,
      "learning_rate": 2e-05,
      "loss": 0.6266217231750488,
      "step": 660
    },
    {
      "epoch": 0.08569966743412638,
      "grad_norm": 2.9441294848550443,
      "learning_rate": 2e-05,
      "loss": 0.6195056438446045,
      "step": 670
    },
    {
      "epoch": 0.08697876694806855,
      "grad_norm": 2.2931610909553206,
      "learning_rate": 2e-05,
      "loss": 0.6402240753173828,
      "step": 680
    },
    {
      "epoch": 0.08825786646201074,
      "grad_norm": 1.5306455027972703,
      "learning_rate": 2e-05,
      "loss": 0.6292661666870117,
      "step": 690
    },
    {
      "epoch": 0.08953696597595293,
      "grad_norm": 3.1898947482255084,
      "learning_rate": 2e-05,
      "loss": 0.6502187728881836,
      "step": 700
    },
    {
      "epoch": 0.09081606548989511,
      "grad_norm": 2.4950893350851175,
      "learning_rate": 2e-05,
      "loss": 0.6811255931854248,
      "step": 710
    },
    {
      "epoch": 0.0920951650038373,
      "grad_norm": 11.041031364083185,
      "learning_rate": 2e-05,
      "loss": 0.6565009117126465,
      "step": 720
    },
    {
      "epoch": 0.09337426451777948,
      "grad_norm": 6.2379473566814365,
      "learning_rate": 2e-05,
      "loss": 0.6482836723327636,
      "step": 730
    },
    {
      "epoch": 0.09465336403172167,
      "grad_norm": 2.778467542831759,
      "learning_rate": 2e-05,
      "loss": 0.6700276851654052,
      "step": 740
    },
    {
      "epoch": 0.09593246354566386,
      "grad_norm": 3.156097269104696,
      "learning_rate": 2e-05,
      "loss": 0.6324517726898193,
      "step": 750
    },
    {
      "epoch": 0.09721156305960604,
      "grad_norm": 5.284389107775213,
      "learning_rate": 2e-05,
      "loss": 0.6471449851989746,
      "step": 760
    },
    {
      "epoch": 0.09849066257354822,
      "grad_norm": 2.412749474777184,
      "learning_rate": 2e-05,
      "loss": 0.6330071449279785,
      "step": 770
    },
    {
      "epoch": 0.0997697620874904,
      "grad_norm": 5.1191263952404995,
      "learning_rate": 2e-05,
      "loss": 0.6730413436889648,
      "step": 780
    },
    {
      "epoch": 0.10104886160143259,
      "grad_norm": 2.6642492725506637,
      "learning_rate": 2e-05,
      "loss": 0.6396522521972656,
      "step": 790
    },
    {
      "epoch": 0.10232796111537477,
      "grad_norm": 3.1687128884117226,
      "learning_rate": 2e-05,
      "loss": 0.6374218940734864,
      "step": 800
    },
    {
      "epoch": 0.10360706062931696,
      "grad_norm": 1.3308499280343475,
      "learning_rate": 2e-05,
      "loss": 0.6415046691894531,
      "step": 810
    },
    {
      "epoch": 0.10488616014325915,
      "grad_norm": 3.3291664977384006,
      "learning_rate": 2e-05,
      "loss": 0.6248232841491699,
      "step": 820
    },
    {
      "epoch": 0.10616525965720133,
      "grad_norm": 4.859932361527834,
      "learning_rate": 2e-05,
      "loss": 0.6399627685546875,
      "step": 830
    },
    {
      "epoch": 0.10744435917114352,
      "grad_norm": 1.3955977789644918,
      "learning_rate": 2e-05,
      "loss": 0.6317929744720459,
      "step": 840
    },
    {
      "epoch": 0.1087234586850857,
      "grad_norm": 1.556340444654958,
      "learning_rate": 2e-05,
      "loss": 0.6560098648071289,
      "step": 850
    },
    {
      "epoch": 0.11000255819902788,
      "grad_norm": 2.5492831528808524,
      "learning_rate": 2e-05,
      "loss": 0.6061997890472413,
      "step": 860
    },
    {
      "epoch": 0.11128165771297006,
      "grad_norm": 1.4147821750495504,
      "learning_rate": 2e-05,
      "loss": 0.6266984462738037,
      "step": 870
    },
    {
      "epoch": 0.11256075722691225,
      "grad_norm": 2.8987948576009255,
      "learning_rate": 2e-05,
      "loss": 0.6040752410888672,
      "step": 880
    },
    {
      "epoch": 0.11383985674085444,
      "grad_norm": 5.819805029378425,
      "learning_rate": 2e-05,
      "loss": 0.6051603317260742,
      "step": 890
    },
    {
      "epoch": 0.11511895625479662,
      "grad_norm": 2.6668997063491373,
      "learning_rate": 2e-05,
      "loss": 0.6632390022277832,
      "step": 900
    },
    {
      "epoch": 0.11639805576873881,
      "grad_norm": 10.472386585595824,
      "learning_rate": 2e-05,
      "loss": 0.5980897903442383,
      "step": 910
    },
    {
      "epoch": 0.117677155282681,
      "grad_norm": 1.6504750774509547,
      "learning_rate": 2e-05,
      "loss": 0.6363026618957519,
      "step": 920
    },
    {
      "epoch": 0.11895625479662318,
      "grad_norm": 2.2303029332193622,
      "learning_rate": 2e-05,
      "loss": 0.6277225494384766,
      "step": 930
    },
    {
      "epoch": 0.12023535431056537,
      "grad_norm": 6.614007656325677,
      "learning_rate": 2e-05,
      "loss": 0.628856086730957,
      "step": 940
    },
    {
      "epoch": 0.12151445382450754,
      "grad_norm": 2.8583085848948158,
      "learning_rate": 2e-05,
      "loss": 0.645742130279541,
      "step": 950
    },
    {
      "epoch": 0.12279355333844973,
      "grad_norm": 4.562616521022261,
      "learning_rate": 2e-05,
      "loss": 0.6232656478881836,
      "step": 960
    },
    {
      "epoch": 0.12407265285239191,
      "grad_norm": 1.6307222970631774,
      "learning_rate": 2e-05,
      "loss": 0.6029533386230469,
      "step": 970
    },
    {
      "epoch": 0.1253517523663341,
      "grad_norm": 2.300518163213002,
      "learning_rate": 2e-05,
      "loss": 0.6161417961120605,
      "step": 980
    },
    {
      "epoch": 0.12663085188027629,
      "grad_norm": 6.2028343698878,
      "learning_rate": 2e-05,
      "loss": 0.6521723747253418,
      "step": 990
    },
    {
      "epoch": 0.12790995139421846,
      "grad_norm": 1.3776781323051128,
      "learning_rate": 2e-05,
      "loss": 0.6107587814331055,
      "step": 1000
    },
    {
      "epoch": 0.12918905090816066,
      "grad_norm": 20.073643216938986,
      "learning_rate": 2e-05,
      "loss": 0.6437624931335449,
      "step": 1010
    },
    {
      "epoch": 0.13046815042210283,
      "grad_norm": 1.9033248345662102,
      "learning_rate": 2e-05,
      "loss": 0.6639025688171387,
      "step": 1020
    },
    {
      "epoch": 0.13174724993604503,
      "grad_norm": 2.6068994484059096,
      "learning_rate": 2e-05,
      "loss": 0.6434185981750489,
      "step": 1030
    },
    {
      "epoch": 0.1330263494499872,
      "grad_norm": 3.3603545258908336,
      "learning_rate": 2e-05,
      "loss": 0.6232691764831543,
      "step": 1040
    },
    {
      "epoch": 0.1343054489639294,
      "grad_norm": 6.927639779770455,
      "learning_rate": 2e-05,
      "loss": 0.6380214691162109,
      "step": 1050
    },
    {
      "epoch": 0.13558454847787157,
      "grad_norm": 4.819538292193655,
      "learning_rate": 2e-05,
      "loss": 0.613338041305542,
      "step": 1060
    },
    {
      "epoch": 0.13686364799181377,
      "grad_norm": 2.5506324496776234,
      "learning_rate": 2e-05,
      "loss": 0.6329474449157715,
      "step": 1070
    },
    {
      "epoch": 0.13814274750575595,
      "grad_norm": 2.3190465926542494,
      "learning_rate": 2e-05,
      "loss": 0.6157898902893066,
      "step": 1080
    },
    {
      "epoch": 0.13942184701969812,
      "grad_norm": 1.9441146816676962,
      "learning_rate": 2e-05,
      "loss": 0.6196088314056396,
      "step": 1090
    },
    {
      "epoch": 0.14070094653364032,
      "grad_norm": 1.3698539182342104,
      "learning_rate": 2e-05,
      "loss": 0.6251972198486329,
      "step": 1100
    },
    {
      "epoch": 0.1419800460475825,
      "grad_norm": 1.6861606365575037,
      "learning_rate": 2e-05,
      "loss": 0.6314759731292725,
      "step": 1110
    },
    {
      "epoch": 0.1432591455615247,
      "grad_norm": 2.462889582220508,
      "learning_rate": 2e-05,
      "loss": 0.65068359375,
      "step": 1120
    },
    {
      "epoch": 0.14453824507546686,
      "grad_norm": 1.7742424329216033,
      "learning_rate": 2e-05,
      "loss": 0.6642440795898438,
      "step": 1130
    },
    {
      "epoch": 0.14581734458940906,
      "grad_norm": 2.320157848873566,
      "learning_rate": 2e-05,
      "loss": 0.6293258666992188,
      "step": 1140
    },
    {
      "epoch": 0.14709644410335124,
      "grad_norm": 2.1787448877032576,
      "learning_rate": 2e-05,
      "loss": 0.6295669555664063,
      "step": 1150
    },
    {
      "epoch": 0.14837554361729344,
      "grad_norm": 1.9357128893077755,
      "learning_rate": 2e-05,
      "loss": 0.6496104240417481,
      "step": 1160
    },
    {
      "epoch": 0.1496546431312356,
      "grad_norm": 4.110272654702471,
      "learning_rate": 2e-05,
      "loss": 0.6259574890136719,
      "step": 1170
    },
    {
      "epoch": 0.15093374264517778,
      "grad_norm": 2.12522137703711,
      "learning_rate": 2e-05,
      "loss": 0.6465348720550537,
      "step": 1180
    },
    {
      "epoch": 0.15221284215911998,
      "grad_norm": 7.509928739655393,
      "learning_rate": 2e-05,
      "loss": 0.6745663642883301,
      "step": 1190
    },
    {
      "epoch": 0.15349194167306215,
      "grad_norm": 2.1546043605789813,
      "learning_rate": 2e-05,
      "loss": 0.6271842956542969,
      "step": 1200
    },
    {
      "epoch": 0.15477104118700435,
      "grad_norm": 4.245976725772466,
      "learning_rate": 2e-05,
      "loss": 0.6347070217132569,
      "step": 1210
    },
    {
      "epoch": 0.15605014070094653,
      "grad_norm": 6.284489606731905,
      "learning_rate": 2e-05,
      "loss": 0.6175366878509522,
      "step": 1220
    },
    {
      "epoch": 0.15732924021488873,
      "grad_norm": 2.3968896694270403,
      "learning_rate": 2e-05,
      "loss": 0.6528811931610108,
      "step": 1230
    },
    {
      "epoch": 0.1586083397288309,
      "grad_norm": 4.44938644957285,
      "learning_rate": 2e-05,
      "loss": 0.6882292747497558,
      "step": 1240
    },
    {
      "epoch": 0.1598874392427731,
      "grad_norm": 1.8152041373141536,
      "learning_rate": 2e-05,
      "loss": 0.639340877532959,
      "step": 1250
    },
    {
      "epoch": 0.16116653875671527,
      "grad_norm": 1.788688964184629,
      "learning_rate": 2e-05,
      "loss": 0.6168077468872071,
      "step": 1260
    },
    {
      "epoch": 0.16244563827065744,
      "grad_norm": 3.011270442070237,
      "learning_rate": 2e-05,
      "loss": 0.6271864891052246,
      "step": 1270
    },
    {
      "epoch": 0.16372473778459964,
      "grad_norm": 3.5372237578631034,
      "learning_rate": 2e-05,
      "loss": 0.6144254684448243,
      "step": 1280
    },
    {
      "epoch": 0.16500383729854182,
      "grad_norm": 3.578563468270323,
      "learning_rate": 2e-05,
      "loss": 0.6395900249481201,
      "step": 1290
    },
    {
      "epoch": 0.16628293681248402,
      "grad_norm": 2.6762637970824787,
      "learning_rate": 2e-05,
      "loss": 0.63500075340271,
      "step": 1300
    },
    {
      "epoch": 0.1675620363264262,
      "grad_norm": 3.1705154085918235,
      "learning_rate": 2e-05,
      "loss": 0.62928786277771,
      "step": 1310
    },
    {
      "epoch": 0.1688411358403684,
      "grad_norm": 2.9443177248500025,
      "learning_rate": 2e-05,
      "loss": 0.6346927165985108,
      "step": 1320
    },
    {
      "epoch": 0.17012023535431056,
      "grad_norm": 2.7168177942420817,
      "learning_rate": 2e-05,
      "loss": 0.636958646774292,
      "step": 1330
    },
    {
      "epoch": 0.17139933486825276,
      "grad_norm": 5.3238376411308055,
      "learning_rate": 2e-05,
      "loss": 0.6074342250823974,
      "step": 1340
    },
    {
      "epoch": 0.17267843438219493,
      "grad_norm": 1.981751065904129,
      "learning_rate": 2e-05,
      "loss": 0.626393985748291,
      "step": 1350
    },
    {
      "epoch": 0.1739575338961371,
      "grad_norm": 6.208058419015704,
      "learning_rate": 2e-05,
      "loss": 0.6492095947265625,
      "step": 1360
    },
    {
      "epoch": 0.1752366334100793,
      "grad_norm": 3.2819184556188,
      "learning_rate": 2e-05,
      "loss": 0.6482612609863281,
      "step": 1370
    },
    {
      "epoch": 0.17651573292402148,
      "grad_norm": 2.59120863821285,
      "learning_rate": 2e-05,
      "loss": 0.6150693416595459,
      "step": 1380
    },
    {
      "epoch": 0.17779483243796368,
      "grad_norm": 1.8538499529211816,
      "learning_rate": 2e-05,
      "loss": 0.6163236618041992,
      "step": 1390
    },
    {
      "epoch": 0.17907393195190585,
      "grad_norm": 2.7084822537552924,
      "learning_rate": 2e-05,
      "loss": 0.6104673385620117,
      "step": 1400
    },
    {
      "epoch": 0.18035303146584805,
      "grad_norm": 4.439571622311464,
      "learning_rate": 2e-05,
      "loss": 0.6378762722015381,
      "step": 1410
    },
    {
      "epoch": 0.18163213097979022,
      "grad_norm": 1.3621008857551957,
      "learning_rate": 2e-05,
      "loss": 0.6388207912445069,
      "step": 1420
    },
    {
      "epoch": 0.18291123049373242,
      "grad_norm": 1.2226259864885263,
      "learning_rate": 2e-05,
      "loss": 0.6120697021484375,
      "step": 1430
    },
    {
      "epoch": 0.1841903300076746,
      "grad_norm": 2.1688106933527322,
      "learning_rate": 2e-05,
      "loss": 0.5997806549072265,
      "step": 1440
    },
    {
      "epoch": 0.18546942952161677,
      "grad_norm": 1.8126305762978137,
      "learning_rate": 2e-05,
      "loss": 0.6054566383361817,
      "step": 1450
    },
    {
      "epoch": 0.18674852903555897,
      "grad_norm": 1.5781503725015227,
      "learning_rate": 2e-05,
      "loss": 0.6335046291351318,
      "step": 1460
    },
    {
      "epoch": 0.18802762854950114,
      "grad_norm": 3.525019895765238,
      "learning_rate": 2e-05,
      "loss": 0.631711196899414,
      "step": 1470
    },
    {
      "epoch": 0.18930672806344334,
      "grad_norm": 2.713602868229821,
      "learning_rate": 2e-05,
      "loss": 0.6091978073120117,
      "step": 1480
    },
    {
      "epoch": 0.1905858275773855,
      "grad_norm": 2.16321468664903,
      "learning_rate": 2e-05,
      "loss": 0.6511111259460449,
      "step": 1490
    },
    {
      "epoch": 0.1918649270913277,
      "grad_norm": 1.3249486358544464,
      "learning_rate": 2e-05,
      "loss": 0.6071283340454101,
      "step": 1500
    },
    {
      "epoch": 0.1931440266052699,
      "grad_norm": 1.4458246810171957,
      "learning_rate": 2e-05,
      "loss": 0.6332869529724121,
      "step": 1510
    },
    {
      "epoch": 0.1944231261192121,
      "grad_norm": 2.10165435585039,
      "learning_rate": 2e-05,
      "loss": 0.6224959850311279,
      "step": 1520
    },
    {
      "epoch": 0.19570222563315426,
      "grad_norm": 1.4509883881939123,
      "learning_rate": 2e-05,
      "loss": 0.5771361351013183,
      "step": 1530
    },
    {
      "epoch": 0.19698132514709643,
      "grad_norm": 4.326909238866829,
      "learning_rate": 2e-05,
      "loss": 0.637632942199707,
      "step": 1540
    },
    {
      "epoch": 0.19826042466103863,
      "grad_norm": 1.6950057566100827,
      "learning_rate": 2e-05,
      "loss": 0.6042330741882325,
      "step": 1550
    },
    {
      "epoch": 0.1995395241749808,
      "grad_norm": 1.417431774916637,
      "learning_rate": 2e-05,
      "loss": 0.6295817375183106,
      "step": 1560
    },
    {
      "epoch": 0.200818623688923,
      "grad_norm": 1.2405770238129155,
      "learning_rate": 2e-05,
      "loss": 0.5979316711425782,
      "step": 1570
    },
    {
      "epoch": 0.20209772320286518,
      "grad_norm": 1.1186847226524728,
      "learning_rate": 2e-05,
      "loss": 0.6549808502197265,
      "step": 1580
    },
    {
      "epoch": 0.20337682271680738,
      "grad_norm": 0.9806293817291409,
      "learning_rate": 2e-05,
      "loss": 0.6115755081176758,
      "step": 1590
    },
    {
      "epoch": 0.20465592223074955,
      "grad_norm": 0.9738594430445625,
      "learning_rate": 2e-05,
      "loss": 0.6076329231262207,
      "step": 1600
    },
    {
      "epoch": 0.20593502174469175,
      "grad_norm": 0.9886089005609494,
      "learning_rate": 2e-05,
      "loss": 0.5939586162567139,
      "step": 1610
    },
    {
      "epoch": 0.20721412125863392,
      "grad_norm": 1.3917061095403442,
      "learning_rate": 2e-05,
      "loss": 0.6180770874023438,
      "step": 1620
    },
    {
      "epoch": 0.2084932207725761,
      "grad_norm": 1.269503803416184,
      "learning_rate": 2e-05,
      "loss": 0.5889194488525391,
      "step": 1630
    },
    {
      "epoch": 0.2097723202865183,
      "grad_norm": 0.949283862884103,
      "learning_rate": 2e-05,
      "loss": 0.598993968963623,
      "step": 1640
    },
    {
      "epoch": 0.21105141980046047,
      "grad_norm": 1.3178112537921784,
      "learning_rate": 2e-05,
      "loss": 0.62374267578125,
      "step": 1650
    },
    {
      "epoch": 0.21233051931440267,
      "grad_norm": 3.3463082639228907,
      "learning_rate": 2e-05,
      "loss": 0.6416094779968262,
      "step": 1660
    },
    {
      "epoch": 0.21360961882834484,
      "grad_norm": 3.5250533260833783,
      "learning_rate": 2e-05,
      "loss": 0.5813175201416015,
      "step": 1670
    },
    {
      "epoch": 0.21488871834228704,
      "grad_norm": 3.1862615057182113,
      "learning_rate": 2e-05,
      "loss": 0.6030552864074707,
      "step": 1680
    },
    {
      "epoch": 0.2161678178562292,
      "grad_norm": 1.561881602369914,
      "learning_rate": 2e-05,
      "loss": 0.6324304580688477,
      "step": 1690
    },
    {
      "epoch": 0.2174469173701714,
      "grad_norm": 2.2515726293089977,
      "learning_rate": 2e-05,
      "loss": 0.6021174907684326,
      "step": 1700
    },
    {
      "epoch": 0.21872601688411358,
      "grad_norm": 2.882171305521348,
      "learning_rate": 2e-05,
      "loss": 0.6039133071899414,
      "step": 1710
    },
    {
      "epoch": 0.22000511639805576,
      "grad_norm": 1.7822903956791751,
      "learning_rate": 2e-05,
      "loss": 0.6312612056732178,
      "step": 1720
    },
    {
      "epoch": 0.22128421591199796,
      "grad_norm": 1.020240331434232,
      "learning_rate": 2e-05,
      "loss": 0.6174317836761475,
      "step": 1730
    },
    {
      "epoch": 0.22256331542594013,
      "grad_norm": 1.2001544084835765,
      "learning_rate": 2e-05,
      "loss": 0.6398989677429199,
      "step": 1740
    },
    {
      "epoch": 0.22384241493988233,
      "grad_norm": 1.47402791893427,
      "learning_rate": 2e-05,
      "loss": 0.6119761943817139,
      "step": 1750
    },
    {
      "epoch": 0.2251215144538245,
      "grad_norm": 2.055369256527985,
      "learning_rate": 2e-05,
      "loss": 0.6196990013122559,
      "step": 1760
    },
    {
      "epoch": 0.2264006139677667,
      "grad_norm": 1.5190061918217799,
      "learning_rate": 2e-05,
      "loss": 0.60753173828125,
      "step": 1770
    },
    {
      "epoch": 0.22767971348170887,
      "grad_norm": 1.0211065145486882,
      "learning_rate": 2e-05,
      "loss": 0.6249045372009278,
      "step": 1780
    },
    {
      "epoch": 0.22895881299565107,
      "grad_norm": 3.530948994051942,
      "learning_rate": 2e-05,
      "loss": 0.5920787811279297,
      "step": 1790
    },
    {
      "epoch": 0.23023791250959325,
      "grad_norm": 5.0504619669596185,
      "learning_rate": 2e-05,
      "loss": 0.6117641448974609,
      "step": 1800
    },
    {
      "epoch": 0.23151701202353542,
      "grad_norm": 4.544881002014028,
      "learning_rate": 2e-05,
      "loss": 0.6159785747528076,
      "step": 1810
    },
    {
      "epoch": 0.23279611153747762,
      "grad_norm": 5.447221215607949,
      "learning_rate": 2e-05,
      "loss": 0.5855494499206543,
      "step": 1820
    },
    {
      "epoch": 0.2340752110514198,
      "grad_norm": 1.354746137837009,
      "learning_rate": 2e-05,
      "loss": 0.6252618312835694,
      "step": 1830
    },
    {
      "epoch": 0.235354310565362,
      "grad_norm": 1.2175526202028102,
      "learning_rate": 2e-05,
      "loss": 0.6352687358856202,
      "step": 1840
    },
    {
      "epoch": 0.23663341007930416,
      "grad_norm": 1.5626912067449465,
      "learning_rate": 2e-05,
      "loss": 0.6236725807189941,
      "step": 1850
    },
    {
      "epoch": 0.23791250959324636,
      "grad_norm": 2.418620586616736,
      "learning_rate": 2e-05,
      "loss": 0.6202504634857178,
      "step": 1860
    },
    {
      "epoch": 0.23919160910718854,
      "grad_norm": 6.501326474729567,
      "learning_rate": 2e-05,
      "loss": 0.6046298980712891,
      "step": 1870
    },
    {
      "epoch": 0.24047070862113074,
      "grad_norm": 13.332115309703186,
      "learning_rate": 2e-05,
      "loss": 0.6176700592041016,
      "step": 1880
    },
    {
      "epoch": 0.2417498081350729,
      "grad_norm": 1.5564620734124854,
      "learning_rate": 2e-05,
      "loss": 0.6322036743164062,
      "step": 1890
    },
    {
      "epoch": 0.24302890764901508,
      "grad_norm": 1.3252946439407358,
      "learning_rate": 2e-05,
      "loss": 0.6052451610565186,
      "step": 1900
    },
    {
      "epoch": 0.24430800716295728,
      "grad_norm": 3.7754178720563902,
      "learning_rate": 2e-05,
      "loss": 0.621720838546753,
      "step": 1910
    },
    {
      "epoch": 0.24558710667689945,
      "grad_norm": 1.7709721126063882,
      "learning_rate": 2e-05,
      "loss": 0.6091246604919434,
      "step": 1920
    },
    {
      "epoch": 0.24686620619084165,
      "grad_norm": 1.3070411526030667,
      "learning_rate": 2e-05,
      "loss": 0.6220854759216309,
      "step": 1930
    },
    {
      "epoch": 0.24814530570478383,
      "grad_norm": 1.8401872322058064,
      "learning_rate": 2e-05,
      "loss": 0.6449881553649902,
      "step": 1940
    },
    {
      "epoch": 0.24942440521872603,
      "grad_norm": 2.4285418670456496,
      "learning_rate": 2e-05,
      "loss": 0.594057559967041,
      "step": 1950
    },
    {
      "epoch": 0.2507035047326682,
      "grad_norm": 1.7235239750037874,
      "learning_rate": 2e-05,
      "loss": 0.6182984828948974,
      "step": 1960
    },
    {
      "epoch": 0.2519826042466104,
      "grad_norm": 1.6285019660644895,
      "learning_rate": 2e-05,
      "loss": 0.5909505367279053,
      "step": 1970
    },
    {
      "epoch": 0.25326170376055257,
      "grad_norm": 1.2819655053500427,
      "learning_rate": 2e-05,
      "loss": 0.6325978755950927,
      "step": 1980
    },
    {
      "epoch": 0.25454080327449474,
      "grad_norm": 1.2159380347170259,
      "learning_rate": 2e-05,
      "loss": 0.5887606620788575,
      "step": 1990
    },
    {
      "epoch": 0.2558199027884369,
      "grad_norm": 0.9505599796559441,
      "learning_rate": 2e-05,
      "loss": 0.574717903137207,
      "step": 2000
    },
    {
      "epoch": 0.25709900230237914,
      "grad_norm": 1.2581472202385364,
      "learning_rate": 2e-05,
      "loss": 0.5856706142425537,
      "step": 2010
    },
    {
      "epoch": 0.2583781018163213,
      "grad_norm": 1.1363388383660025,
      "learning_rate": 2e-05,
      "loss": 0.5883913040161133,
      "step": 2020
    },
    {
      "epoch": 0.2596572013302635,
      "grad_norm": 1.560965561232806,
      "learning_rate": 2e-05,
      "loss": 0.5746139049530029,
      "step": 2030
    },
    {
      "epoch": 0.26093630084420566,
      "grad_norm": 2.1211040191156343,
      "learning_rate": 2e-05,
      "loss": 0.6083070755004882,
      "step": 2040
    },
    {
      "epoch": 0.2622154003581479,
      "grad_norm": 1.713441497273608,
      "learning_rate": 2e-05,
      "loss": 0.6275928020477295,
      "step": 2050
    },
    {
      "epoch": 0.26349449987209006,
      "grad_norm": 2.038098915097047,
      "learning_rate": 2e-05,
      "loss": 0.6330203533172607,
      "step": 2060
    },
    {
      "epoch": 0.26477359938603223,
      "grad_norm": 1.2264196348153609,
      "learning_rate": 2e-05,
      "loss": 0.6101226806640625,
      "step": 2070
    },
    {
      "epoch": 0.2660526988999744,
      "grad_norm": 1.8210121510445536,
      "learning_rate": 2e-05,
      "loss": 0.6420550346374512,
      "step": 2080
    },
    {
      "epoch": 0.2673317984139166,
      "grad_norm": 0.8963201860659803,
      "learning_rate": 2e-05,
      "loss": 0.5967807769775391,
      "step": 2090
    },
    {
      "epoch": 0.2686108979278588,
      "grad_norm": 2.0260697664647003,
      "learning_rate": 2e-05,
      "loss": 0.6149022579193115,
      "step": 2100
    },
    {
      "epoch": 0.269889997441801,
      "grad_norm": 4.914335014084526,
      "learning_rate": 2e-05,
      "loss": 0.6114978790283203,
      "step": 2110
    },
    {
      "epoch": 0.27116909695574315,
      "grad_norm": 2.106317779910681,
      "learning_rate": 2e-05,
      "loss": 0.6006545066833496,
      "step": 2120
    },
    {
      "epoch": 0.2724481964696853,
      "grad_norm": 2.7288315662600158,
      "learning_rate": 2e-05,
      "loss": 0.5773192882537842,
      "step": 2130
    },
    {
      "epoch": 0.27372729598362755,
      "grad_norm": 1.2064614446581332,
      "learning_rate": 2e-05,
      "loss": 0.5988784313201905,
      "step": 2140
    },
    {
      "epoch": 0.2750063954975697,
      "grad_norm": 1.4196592877324987,
      "learning_rate": 2e-05,
      "loss": 0.6252477169036865,
      "step": 2150
    },
    {
      "epoch": 0.2762854950115119,
      "grad_norm": 5.085511259042394,
      "learning_rate": 2e-05,
      "loss": 0.6232048034667969,
      "step": 2160
    },
    {
      "epoch": 0.27756459452545407,
      "grad_norm": 1.6573265068305214,
      "learning_rate": 2e-05,
      "loss": 0.5724109172821045,
      "step": 2170
    },
    {
      "epoch": 0.27884369403939624,
      "grad_norm": 0.8570570183137078,
      "learning_rate": 2e-05,
      "loss": 0.6122853755950928,
      "step": 2180
    },
    {
      "epoch": 0.28012279355333847,
      "grad_norm": 1.0409675006077839,
      "learning_rate": 2e-05,
      "loss": 0.5958579540252685,
      "step": 2190
    },
    {
      "epoch": 0.28140189306728064,
      "grad_norm": 2.850717441256514,
      "learning_rate": 2e-05,
      "loss": 0.61070556640625,
      "step": 2200
    },
    {
      "epoch": 0.2826809925812228,
      "grad_norm": 1.674482950677382,
      "learning_rate": 2e-05,
      "loss": 0.6583208084106446,
      "step": 2210
    },
    {
      "epoch": 0.283960092095165,
      "grad_norm": 1.031060356132696,
      "learning_rate": 2e-05,
      "loss": 0.5889376640319824,
      "step": 2220
    },
    {
      "epoch": 0.2852391916091072,
      "grad_norm": 1.2666316468928664,
      "learning_rate": 2e-05,
      "loss": 0.6129489421844483,
      "step": 2230
    },
    {
      "epoch": 0.2865182911230494,
      "grad_norm": 1.6411785567635628,
      "learning_rate": 2e-05,
      "loss": 0.5777743816375732,
      "step": 2240
    },
    {
      "epoch": 0.28779739063699156,
      "grad_norm": 1.8797050463236573,
      "learning_rate": 2e-05,
      "loss": 0.6185701370239258,
      "step": 2250
    },
    {
      "epoch": 0.28907649015093373,
      "grad_norm": 1.1158819337514871,
      "learning_rate": 2e-05,
      "loss": 0.6073112487792969,
      "step": 2260
    },
    {
      "epoch": 0.2903555896648759,
      "grad_norm": 1.260587701397972,
      "learning_rate": 2e-05,
      "loss": 0.604709243774414,
      "step": 2270
    },
    {
      "epoch": 0.29163468917881813,
      "grad_norm": 2.394044170108493,
      "learning_rate": 2e-05,
      "loss": 0.595456886291504,
      "step": 2280
    },
    {
      "epoch": 0.2929137886927603,
      "grad_norm": 1.857401244405304,
      "learning_rate": 2e-05,
      "loss": 0.6224223136901855,
      "step": 2290
    },
    {
      "epoch": 0.2941928882067025,
      "grad_norm": 2.2784297913171336,
      "learning_rate": 2e-05,
      "loss": 0.6300773620605469,
      "step": 2300
    },
    {
      "epoch": 0.29547198772064465,
      "grad_norm": 0.9471614831952674,
      "learning_rate": 2e-05,
      "loss": 0.6309947013854981,
      "step": 2310
    },
    {
      "epoch": 0.2967510872345869,
      "grad_norm": 1.1542869310513502,
      "learning_rate": 2e-05,
      "loss": 0.6161383628845215,
      "step": 2320
    },
    {
      "epoch": 0.29803018674852905,
      "grad_norm": 0.9695713668697032,
      "learning_rate": 2e-05,
      "loss": 0.6323893547058106,
      "step": 2330
    },
    {
      "epoch": 0.2993092862624712,
      "grad_norm": 0.9319436233814953,
      "learning_rate": 2e-05,
      "loss": 0.5904321193695068,
      "step": 2340
    },
    {
      "epoch": 0.3005883857764134,
      "grad_norm": 1.227810386425401,
      "learning_rate": 2e-05,
      "loss": 0.6253926277160644,
      "step": 2350
    },
    {
      "epoch": 0.30186748529035556,
      "grad_norm": 1.7096380269139588,
      "learning_rate": 2e-05,
      "loss": 0.5987714290618896,
      "step": 2360
    },
    {
      "epoch": 0.3031465848042978,
      "grad_norm": 3.3222742211615315,
      "learning_rate": 2e-05,
      "loss": 0.609607458114624,
      "step": 2370
    },
    {
      "epoch": 0.30442568431823996,
      "grad_norm": 1.0979241510256794,
      "learning_rate": 2e-05,
      "loss": 0.5992640495300293,
      "step": 2380
    },
    {
      "epoch": 0.30570478383218214,
      "grad_norm": 3.532618720860108,
      "learning_rate": 2e-05,
      "loss": 0.6203838348388672,
      "step": 2390
    },
    {
      "epoch": 0.3069838833461243,
      "grad_norm": 1.22304004334949,
      "learning_rate": 2e-05,
      "loss": 0.5954967021942139,
      "step": 2400
    },
    {
      "epoch": 0.30826298286006654,
      "grad_norm": 2.410293308883467,
      "learning_rate": 2e-05,
      "loss": 0.6060263633728027,
      "step": 2410
    },
    {
      "epoch": 0.3095420823740087,
      "grad_norm": 1.2560451534079777,
      "learning_rate": 2e-05,
      "loss": 0.6317928314208985,
      "step": 2420
    },
    {
      "epoch": 0.3108211818879509,
      "grad_norm": 1.0455056153829623,
      "learning_rate": 2e-05,
      "loss": 0.5984413146972656,
      "step": 2430
    },
    {
      "epoch": 0.31210028140189305,
      "grad_norm": 1.650132689260639,
      "learning_rate": 2e-05,
      "loss": 0.6033206939697265,
      "step": 2440
    },
    {
      "epoch": 0.3133793809158352,
      "grad_norm": 2.5592592325487105,
      "learning_rate": 2e-05,
      "loss": 0.5678141593933106,
      "step": 2450
    },
    {
      "epoch": 0.31465848042977745,
      "grad_norm": 1.562608097897649,
      "learning_rate": 2e-05,
      "loss": 0.5921920776367188,
      "step": 2460
    },
    {
      "epoch": 0.3159375799437196,
      "grad_norm": 2.192241892940132,
      "learning_rate": 2e-05,
      "loss": 0.598881721496582,
      "step": 2470
    },
    {
      "epoch": 0.3172166794576618,
      "grad_norm": 0.9967509265789414,
      "learning_rate": 2e-05,
      "loss": 0.575558090209961,
      "step": 2480
    },
    {
      "epoch": 0.31849577897160397,
      "grad_norm": 4.787238755740847,
      "learning_rate": 2e-05,
      "loss": 0.5959879875183105,
      "step": 2490
    },
    {
      "epoch": 0.3197748784855462,
      "grad_norm": 1.1428325653163887,
      "learning_rate": 2e-05,
      "loss": 0.5783063411712647,
      "step": 2500
    },
    {
      "epoch": 0.32105397799948837,
      "grad_norm": 2.9202327745384973,
      "learning_rate": 2e-05,
      "loss": 0.5774745941162109,
      "step": 2510
    },
    {
      "epoch": 0.32233307751343054,
      "grad_norm": 2.0279978562828735,
      "learning_rate": 2e-05,
      "loss": 0.6202166557312012,
      "step": 2520
    },
    {
      "epoch": 0.3236121770273727,
      "grad_norm": 1.5802874918294092,
      "learning_rate": 2e-05,
      "loss": 0.6294447898864746,
      "step": 2530
    },
    {
      "epoch": 0.3248912765413149,
      "grad_norm": 0.9906512101701456,
      "learning_rate": 2e-05,
      "loss": 0.5678004264831543,
      "step": 2540
    },
    {
      "epoch": 0.3261703760552571,
      "grad_norm": 1.746760187852664,
      "learning_rate": 2e-05,
      "loss": 0.6238263607025146,
      "step": 2550
    },
    {
      "epoch": 0.3274494755691993,
      "grad_norm": 2.8108078554248594,
      "learning_rate": 2e-05,
      "loss": 0.6203858375549316,
      "step": 2560
    },
    {
      "epoch": 0.32872857508314146,
      "grad_norm": 4.806066554618259,
      "learning_rate": 2e-05,
      "loss": 0.6184251308441162,
      "step": 2570
    },
    {
      "epoch": 0.33000767459708363,
      "grad_norm": 2.0113483642991965,
      "learning_rate": 2e-05,
      "loss": 0.6477134704589844,
      "step": 2580
    },
    {
      "epoch": 0.33128677411102586,
      "grad_norm": 1.3830922692363252,
      "learning_rate": 2e-05,
      "loss": 0.6008767127990723,
      "step": 2590
    },
    {
      "epoch": 0.33256587362496803,
      "grad_norm": 1.2349526734744556,
      "learning_rate": 2e-05,
      "loss": 0.624567699432373,
      "step": 2600
    },
    {
      "epoch": 0.3338449731389102,
      "grad_norm": 1.271755174590234,
      "learning_rate": 2e-05,
      "loss": 0.6194814205169678,
      "step": 2610
    },
    {
      "epoch": 0.3351240726528524,
      "grad_norm": 1.4926328848591621,
      "learning_rate": 2e-05,
      "loss": 0.5909404754638672,
      "step": 2620
    },
    {
      "epoch": 0.33640317216679455,
      "grad_norm": 1.7225071189531411,
      "learning_rate": 2e-05,
      "loss": 0.6224935054779053,
      "step": 2630
    },
    {
      "epoch": 0.3376822716807368,
      "grad_norm": 3.5736648482858246,
      "learning_rate": 2e-05,
      "loss": 0.6273334503173829,
      "step": 2640
    },
    {
      "epoch": 0.33896137119467895,
      "grad_norm": 1.6199005667039337,
      "learning_rate": 2e-05,
      "loss": 0.6213912010192871,
      "step": 2650
    },
    {
      "epoch": 0.3402404707086211,
      "grad_norm": 1.2784430597805259,
      "learning_rate": 2e-05,
      "loss": 0.6026556968688965,
      "step": 2660
    },
    {
      "epoch": 0.3415195702225633,
      "grad_norm": 3.8766837411975454,
      "learning_rate": 2e-05,
      "loss": 0.5997205257415772,
      "step": 2670
    },
    {
      "epoch": 0.3427986697365055,
      "grad_norm": 1.522149639598022,
      "learning_rate": 2e-05,
      "loss": 0.6241069793701172,
      "step": 2680
    },
    {
      "epoch": 0.3440777692504477,
      "grad_norm": 2.7584445375179047,
      "learning_rate": 2e-05,
      "loss": 0.6011125564575195,
      "step": 2690
    },
    {
      "epoch": 0.34535686876438987,
      "grad_norm": 4.487541114475743,
      "learning_rate": 2e-05,
      "loss": 0.6028533458709717,
      "step": 2700
    },
    {
      "epoch": 0.34663596827833204,
      "grad_norm": 1.1344741201338224,
      "learning_rate": 2e-05,
      "loss": 0.6265992164611817,
      "step": 2710
    },
    {
      "epoch": 0.3479150677922742,
      "grad_norm": 1.264736856329463,
      "learning_rate": 2e-05,
      "loss": 0.63673095703125,
      "step": 2720
    },
    {
      "epoch": 0.34919416730621644,
      "grad_norm": 0.8044421663432847,
      "learning_rate": 2e-05,
      "loss": 0.602764892578125,
      "step": 2730
    },
    {
      "epoch": 0.3504732668201586,
      "grad_norm": 1.0593200577466688,
      "learning_rate": 2e-05,
      "loss": 0.5976439952850342,
      "step": 2740
    },
    {
      "epoch": 0.3517523663341008,
      "grad_norm": 1.1434465645557081,
      "learning_rate": 2e-05,
      "loss": 0.5878979682922363,
      "step": 2750
    },
    {
      "epoch": 0.35303146584804296,
      "grad_norm": 1.3903295183809865,
      "learning_rate": 2e-05,
      "loss": 0.6098363399505615,
      "step": 2760
    },
    {
      "epoch": 0.3543105653619852,
      "grad_norm": 1.282377166346566,
      "learning_rate": 2e-05,
      "loss": 0.5990135192871093,
      "step": 2770
    },
    {
      "epoch": 0.35558966487592736,
      "grad_norm": 1.1435337542967667,
      "learning_rate": 2e-05,
      "loss": 0.5908602714538574,
      "step": 2780
    },
    {
      "epoch": 0.35686876438986953,
      "grad_norm": 1.170245442472612,
      "learning_rate": 2e-05,
      "loss": 0.6214659690856934,
      "step": 2790
    },
    {
      "epoch": 0.3581478639038117,
      "grad_norm": 1.0601914113434607,
      "learning_rate": 2e-05,
      "loss": 0.5969385147094727,
      "step": 2800
    },
    {
      "epoch": 0.3594269634177539,
      "grad_norm": 0.8880429422636937,
      "learning_rate": 2e-05,
      "loss": 0.5900382041931153,
      "step": 2810
    },
    {
      "epoch": 0.3607060629316961,
      "grad_norm": 1.3291734362533796,
      "learning_rate": 2e-05,
      "loss": 0.5984704971313477,
      "step": 2820
    },
    {
      "epoch": 0.3619851624456383,
      "grad_norm": 2.3196798317877803,
      "learning_rate": 2e-05,
      "loss": 0.6031060695648194,
      "step": 2830
    },
    {
      "epoch": 0.36326426195958045,
      "grad_norm": 1.4909813232977367,
      "learning_rate": 2e-05,
      "loss": 0.5790597438812256,
      "step": 2840
    },
    {
      "epoch": 0.3645433614735226,
      "grad_norm": 2.7690017473286033,
      "learning_rate": 2e-05,
      "loss": 0.577203369140625,
      "step": 2850
    },
    {
      "epoch": 0.36582246098746485,
      "grad_norm": 1.2195966201599757,
      "learning_rate": 2e-05,
      "loss": 0.6123817443847657,
      "step": 2860
    },
    {
      "epoch": 0.367101560501407,
      "grad_norm": 3.2580341346851034,
      "learning_rate": 2e-05,
      "loss": 0.5920998573303222,
      "step": 2870
    },
    {
      "epoch": 0.3683806600153492,
      "grad_norm": 1.191401593758783,
      "learning_rate": 2e-05,
      "loss": 0.6038727760314941,
      "step": 2880
    },
    {
      "epoch": 0.36965975952929137,
      "grad_norm": 1.3743110118592492,
      "learning_rate": 2e-05,
      "loss": 0.6003240585327149,
      "step": 2890
    },
    {
      "epoch": 0.37093885904323354,
      "grad_norm": 4.509715704890444,
      "learning_rate": 2e-05,
      "loss": 0.5917400360107422,
      "step": 2900
    },
    {
      "epoch": 0.37221795855717577,
      "grad_norm": 0.9032779057730358,
      "learning_rate": 2e-05,
      "loss": 0.6044122695922851,
      "step": 2910
    },
    {
      "epoch": 0.37349705807111794,
      "grad_norm": 0.9594536006104978,
      "learning_rate": 2e-05,
      "loss": 0.6204854488372803,
      "step": 2920
    },
    {
      "epoch": 0.3747761575850601,
      "grad_norm": 2.9729199422688404,
      "learning_rate": 2e-05,
      "loss": 0.6077807426452637,
      "step": 2930
    },
    {
      "epoch": 0.3760552570990023,
      "grad_norm": 2.0925145891159795,
      "learning_rate": 2e-05,
      "loss": 0.5951478004455566,
      "step": 2940
    },
    {
      "epoch": 0.3773343566129445,
      "grad_norm": 21.76963806945384,
      "learning_rate": 2e-05,
      "loss": 0.5952364921569824,
      "step": 2950
    },
    {
      "epoch": 0.3786134561268867,
      "grad_norm": 1.5329185958740825,
      "learning_rate": 2e-05,
      "loss": 0.6125658988952637,
      "step": 2960
    },
    {
      "epoch": 0.37989255564082886,
      "grad_norm": 1.4903676249564757,
      "learning_rate": 2e-05,
      "loss": 0.5745217800140381,
      "step": 2970
    },
    {
      "epoch": 0.381171655154771,
      "grad_norm": 1.8185242853324988,
      "learning_rate": 2e-05,
      "loss": 0.587047004699707,
      "step": 2980
    },
    {
      "epoch": 0.3824507546687132,
      "grad_norm": 1.1127655033537205,
      "learning_rate": 2e-05,
      "loss": 0.5825289726257324,
      "step": 2990
    },
    {
      "epoch": 0.3837298541826554,
      "grad_norm": 2.494189103068932,
      "learning_rate": 2e-05,
      "loss": 0.6626487731933594,
      "step": 3000
    },
    {
      "epoch": 0.3850089536965976,
      "grad_norm": 1.1353770749693715,
      "learning_rate": 2e-05,
      "loss": 0.5981975555419922,
      "step": 3010
    },
    {
      "epoch": 0.3862880532105398,
      "grad_norm": 4.0866893647819476,
      "learning_rate": 2e-05,
      "loss": 0.6148322582244873,
      "step": 3020
    },
    {
      "epoch": 0.38756715272448194,
      "grad_norm": 1.4328489841066114,
      "learning_rate": 2e-05,
      "loss": 0.5979141235351563,
      "step": 3030
    },
    {
      "epoch": 0.3888462522384242,
      "grad_norm": 0.9580306575907652,
      "learning_rate": 2e-05,
      "loss": 0.5795217514038086,
      "step": 3040
    },
    {
      "epoch": 0.39012535175236634,
      "grad_norm": 1.027657289655969,
      "learning_rate": 2e-05,
      "loss": 0.593060827255249,
      "step": 3050
    },
    {
      "epoch": 0.3914044512663085,
      "grad_norm": 2.867852167814524,
      "learning_rate": 2e-05,
      "loss": 0.649545955657959,
      "step": 3060
    },
    {
      "epoch": 0.3926835507802507,
      "grad_norm": 1.7315278137528691,
      "learning_rate": 2e-05,
      "loss": 0.5818486690521241,
      "step": 3070
    },
    {
      "epoch": 0.39396265029419286,
      "grad_norm": 18.51948790045596,
      "learning_rate": 2e-05,
      "loss": 0.6028565883636474,
      "step": 3080
    },
    {
      "epoch": 0.3952417498081351,
      "grad_norm": 1.0814514467211724,
      "learning_rate": 2e-05,
      "loss": 0.6369311809539795,
      "step": 3090
    },
    {
      "epoch": 0.39652084932207726,
      "grad_norm": 0.8365819549840845,
      "learning_rate": 2e-05,
      "loss": 0.5891385555267334,
      "step": 3100
    },
    {
      "epoch": 0.39779994883601943,
      "grad_norm": 1.1643014580060915,
      "learning_rate": 2e-05,
      "loss": 0.620572280883789,
      "step": 3110
    },
    {
      "epoch": 0.3990790483499616,
      "grad_norm": 1.3853634092376967,
      "learning_rate": 2e-05,
      "loss": 0.5695016384124756,
      "step": 3120
    },
    {
      "epoch": 0.40035814786390383,
      "grad_norm": 0.941601759431241,
      "learning_rate": 2e-05,
      "loss": 0.5876264095306396,
      "step": 3130
    },
    {
      "epoch": 0.401637247377846,
      "grad_norm": 0.9057064185186952,
      "learning_rate": 2e-05,
      "loss": 0.635837173461914,
      "step": 3140
    },
    {
      "epoch": 0.4029163468917882,
      "grad_norm": 1.0678199968255322,
      "learning_rate": 2e-05,
      "loss": 0.5842601776123046,
      "step": 3150
    },
    {
      "epoch": 0.40419544640573035,
      "grad_norm": 2.353208944812402,
      "learning_rate": 2e-05,
      "loss": 0.6364971160888672,
      "step": 3160
    },
    {
      "epoch": 0.4054745459196725,
      "grad_norm": 2.2056815946147945,
      "learning_rate": 2e-05,
      "loss": 0.5831233024597168,
      "step": 3170
    },
    {
      "epoch": 0.40675364543361475,
      "grad_norm": 1.0432213491886457,
      "learning_rate": 2e-05,
      "loss": 0.6237380981445313,
      "step": 3180
    },
    {
      "epoch": 0.4080327449475569,
      "grad_norm": 1.7016360850124244,
      "learning_rate": 2e-05,
      "loss": 0.5941524505615234,
      "step": 3190
    },
    {
      "epoch": 0.4093118444614991,
      "grad_norm": 17.46618090659867,
      "learning_rate": 2e-05,
      "loss": 0.6070680141448974,
      "step": 3200
    },
    {
      "epoch": 0.41059094397544127,
      "grad_norm": 1.9361701078093683,
      "learning_rate": 2e-05,
      "loss": 0.5642125129699707,
      "step": 3210
    },
    {
      "epoch": 0.4118700434893835,
      "grad_norm": 1.9629421677603713,
      "learning_rate": 2e-05,
      "loss": 0.5952259063720703,
      "step": 3220
    },
    {
      "epoch": 0.41314914300332567,
      "grad_norm": 3.65571115063658,
      "learning_rate": 2e-05,
      "loss": 0.5965877056121827,
      "step": 3230
    },
    {
      "epoch": 0.41442824251726784,
      "grad_norm": 1.3605239806487905,
      "learning_rate": 2e-05,
      "loss": 0.5746373653411865,
      "step": 3240
    },
    {
      "epoch": 0.41570734203121,
      "grad_norm": 1.0451824030026722,
      "learning_rate": 2e-05,
      "loss": 0.5859414577484131,
      "step": 3250
    },
    {
      "epoch": 0.4169864415451522,
      "grad_norm": 1.7463120327939081,
      "learning_rate": 2e-05,
      "loss": 0.5620139122009278,
      "step": 3260
    },
    {
      "epoch": 0.4182655410590944,
      "grad_norm": 2.138413071925528,
      "learning_rate": 2e-05,
      "loss": 0.5906031608581543,
      "step": 3270
    },
    {
      "epoch": 0.4195446405730366,
      "grad_norm": 2.7608510899833103,
      "learning_rate": 2e-05,
      "loss": 0.5822212219238281,
      "step": 3280
    },
    {
      "epoch": 0.42082374008697876,
      "grad_norm": 0.9893407948872976,
      "learning_rate": 2e-05,
      "loss": 0.6076113700866699,
      "step": 3290
    },
    {
      "epoch": 0.42210283960092093,
      "grad_norm": 3.415050230981063,
      "learning_rate": 2e-05,
      "loss": 0.5842854499816894,
      "step": 3300
    },
    {
      "epoch": 0.42338193911486316,
      "grad_norm": 1.2831244261530679,
      "learning_rate": 2e-05,
      "loss": 0.6099458694458008,
      "step": 3310
    },
    {
      "epoch": 0.42466103862880533,
      "grad_norm": 1.463679574003703,
      "learning_rate": 2e-05,
      "loss": 0.5927666664123535,
      "step": 3320
    },
    {
      "epoch": 0.4259401381427475,
      "grad_norm": 1.2948575443800137,
      "learning_rate": 2e-05,
      "loss": 0.6183608055114747,
      "step": 3330
    },
    {
      "epoch": 0.4272192376566897,
      "grad_norm": 1.7719649112060292,
      "learning_rate": 2e-05,
      "loss": 0.5920117855072021,
      "step": 3340
    },
    {
      "epoch": 0.42849833717063185,
      "grad_norm": 2.5338780912419314,
      "learning_rate": 2e-05,
      "loss": 0.5885765075683593,
      "step": 3350
    },
    {
      "epoch": 0.4297774366845741,
      "grad_norm": 2.1104264661059933,
      "learning_rate": 2e-05,
      "loss": 0.5881740093231201,
      "step": 3360
    },
    {
      "epoch": 0.43105653619851625,
      "grad_norm": 3.062281823976241,
      "learning_rate": 2e-05,
      "loss": 0.5795594692230225,
      "step": 3370
    },
    {
      "epoch": 0.4323356357124584,
      "grad_norm": 1.1334170743304055,
      "learning_rate": 2e-05,
      "loss": 0.6271108150482178,
      "step": 3380
    },
    {
      "epoch": 0.4336147352264006,
      "grad_norm": 1.3336335739798473,
      "learning_rate": 2e-05,
      "loss": 0.602003812789917,
      "step": 3390
    },
    {
      "epoch": 0.4348938347403428,
      "grad_norm": 1.3227068900163543,
      "learning_rate": 2e-05,
      "loss": 0.6033557891845703,
      "step": 3400
    },
    {
      "epoch": 0.436172934254285,
      "grad_norm": 1.1445072235885272,
      "learning_rate": 2e-05,
      "loss": 0.6056077003479003,
      "step": 3410
    },
    {
      "epoch": 0.43745203376822717,
      "grad_norm": 2.8759382031453025,
      "learning_rate": 2e-05,
      "loss": 0.5847073554992676,
      "step": 3420
    },
    {
      "epoch": 0.43873113328216934,
      "grad_norm": 2.40286580704749,
      "learning_rate": 2e-05,
      "loss": 0.6021082878112793,
      "step": 3430
    },
    {
      "epoch": 0.4400102327961115,
      "grad_norm": 1.058490597268109,
      "learning_rate": 2e-05,
      "loss": 0.5841891765594482,
      "step": 3440
    },
    {
      "epoch": 0.44128933231005374,
      "grad_norm": 1.304059233417537,
      "learning_rate": 2e-05,
      "loss": 0.6094838619232178,
      "step": 3450
    },
    {
      "epoch": 0.4425684318239959,
      "grad_norm": 0.8060832967578174,
      "learning_rate": 2e-05,
      "loss": 0.5862702846527099,
      "step": 3460
    },
    {
      "epoch": 0.4438475313379381,
      "grad_norm": 0.840500573095063,
      "learning_rate": 2e-05,
      "loss": 0.5949713706970214,
      "step": 3470
    },
    {
      "epoch": 0.44512663085188026,
      "grad_norm": 4.542032451860691,
      "learning_rate": 2e-05,
      "loss": 0.6270683288574219,
      "step": 3480
    },
    {
      "epoch": 0.4464057303658225,
      "grad_norm": 1.0368756465623938,
      "learning_rate": 2e-05,
      "loss": 0.606722640991211,
      "step": 3490
    },
    {
      "epoch": 0.44768482987976466,
      "grad_norm": 0.947207216589235,
      "learning_rate": 2e-05,
      "loss": 0.5707377433776856,
      "step": 3500
    },
    {
      "epoch": 0.44896392939370683,
      "grad_norm": 1.62305712819688,
      "learning_rate": 2e-05,
      "loss": 0.5990262508392334,
      "step": 3510
    },
    {
      "epoch": 0.450243028907649,
      "grad_norm": 1.4535680237402864,
      "learning_rate": 2e-05,
      "loss": 0.6101785182952881,
      "step": 3520
    },
    {
      "epoch": 0.4515221284215912,
      "grad_norm": 1.0679867768799687,
      "learning_rate": 2e-05,
      "loss": 0.5582299709320069,
      "step": 3530
    },
    {
      "epoch": 0.4528012279355334,
      "grad_norm": 2.029254450147691,
      "learning_rate": 2e-05,
      "loss": 0.6040618896484375,
      "step": 3540
    },
    {
      "epoch": 0.4540803274494756,
      "grad_norm": 3.4693073257949103,
      "learning_rate": 2e-05,
      "loss": 0.5946227073669433,
      "step": 3550
    },
    {
      "epoch": 0.45535942696341775,
      "grad_norm": 1.8264567925108544,
      "learning_rate": 2e-05,
      "loss": 0.6080288887023926,
      "step": 3560
    },
    {
      "epoch": 0.4566385264773599,
      "grad_norm": 1.0695415170895906,
      "learning_rate": 2e-05,
      "loss": 0.6208341598510743,
      "step": 3570
    },
    {
      "epoch": 0.45791762599130215,
      "grad_norm": 1.1090984471019585,
      "learning_rate": 2e-05,
      "loss": 0.5863708019256592,
      "step": 3580
    },
    {
      "epoch": 0.4591967255052443,
      "grad_norm": 1.6688268457920834,
      "learning_rate": 2e-05,
      "loss": 0.591333532333374,
      "step": 3590
    },
    {
      "epoch": 0.4604758250191865,
      "grad_norm": 1.8755299547963225,
      "learning_rate": 2e-05,
      "loss": 0.5972795486450195,
      "step": 3600
    },
    {
      "epoch": 0.46175492453312866,
      "grad_norm": 1.991568535455219,
      "learning_rate": 2e-05,
      "loss": 0.5916287899017334,
      "step": 3610
    },
    {
      "epoch": 0.46303402404707084,
      "grad_norm": 5.633434865013647,
      "learning_rate": 2e-05,
      "loss": 0.5936531066894531,
      "step": 3620
    },
    {
      "epoch": 0.46431312356101306,
      "grad_norm": 2.7813966842720927,
      "learning_rate": 2e-05,
      "loss": 0.6172687530517578,
      "step": 3630
    },
    {
      "epoch": 0.46559222307495524,
      "grad_norm": 1.1795683170771842,
      "learning_rate": 2e-05,
      "loss": 0.608796501159668,
      "step": 3640
    },
    {
      "epoch": 0.4668713225888974,
      "grad_norm": 2.1898796237116938,
      "learning_rate": 2e-05,
      "loss": 0.5805678367614746,
      "step": 3650
    },
    {
      "epoch": 0.4681504221028396,
      "grad_norm": 1.5027857186055826,
      "learning_rate": 2e-05,
      "loss": 0.5937762260437012,
      "step": 3660
    },
    {
      "epoch": 0.4694295216167818,
      "grad_norm": 2.2064227907053877,
      "learning_rate": 2e-05,
      "loss": 0.5972755432128907,
      "step": 3670
    },
    {
      "epoch": 0.470708621130724,
      "grad_norm": 3.307339050548398,
      "learning_rate": 2e-05,
      "loss": 0.6013876914978027,
      "step": 3680
    },
    {
      "epoch": 0.47198772064466615,
      "grad_norm": 1.7538816105627706,
      "learning_rate": 2e-05,
      "loss": 0.5830701351165771,
      "step": 3690
    },
    {
      "epoch": 0.4732668201586083,
      "grad_norm": 1.1541330208039735,
      "learning_rate": 2e-05,
      "loss": 0.5992613792419433,
      "step": 3700
    },
    {
      "epoch": 0.4745459196725505,
      "grad_norm": 2.433370390394683,
      "learning_rate": 2e-05,
      "loss": 0.6018333435058594,
      "step": 3710
    },
    {
      "epoch": 0.4758250191864927,
      "grad_norm": 8.05880675122322,
      "learning_rate": 2e-05,
      "loss": 0.5928830146789551,
      "step": 3720
    },
    {
      "epoch": 0.4771041187004349,
      "grad_norm": 1.4689043424612738,
      "learning_rate": 2e-05,
      "loss": 0.5966341972351075,
      "step": 3730
    },
    {
      "epoch": 0.47838321821437707,
      "grad_norm": 1.3562110418457431,
      "learning_rate": 2e-05,
      "loss": 0.5890920162200928,
      "step": 3740
    },
    {
      "epoch": 0.47966231772831924,
      "grad_norm": 1.878872514143756,
      "learning_rate": 2e-05,
      "loss": 0.5998680114746093,
      "step": 3750
    },
    {
      "epoch": 0.48094141724226147,
      "grad_norm": 1.1196343435761942,
      "learning_rate": 2e-05,
      "loss": 0.5796008110046387,
      "step": 3760
    },
    {
      "epoch": 0.48222051675620364,
      "grad_norm": 1.1156802634346448,
      "learning_rate": 2e-05,
      "loss": 0.5972362518310547,
      "step": 3770
    },
    {
      "epoch": 0.4834996162701458,
      "grad_norm": 2.1220659692240647,
      "learning_rate": 2e-05,
      "loss": 0.5801189422607422,
      "step": 3780
    },
    {
      "epoch": 0.484778715784088,
      "grad_norm": 0.868961971462684,
      "learning_rate": 2e-05,
      "loss": 0.5691156387329102,
      "step": 3790
    },
    {
      "epoch": 0.48605781529803016,
      "grad_norm": 0.9336123344027439,
      "learning_rate": 2e-05,
      "loss": 0.5900723934173584,
      "step": 3800
    },
    {
      "epoch": 0.4873369148119724,
      "grad_norm": 0.9906753203156491,
      "learning_rate": 2e-05,
      "loss": 0.5864486694335938,
      "step": 3810
    },
    {
      "epoch": 0.48861601432591456,
      "grad_norm": 1.0851288805526387,
      "learning_rate": 2e-05,
      "loss": 0.5898387908935547,
      "step": 3820
    },
    {
      "epoch": 0.48989511383985673,
      "grad_norm": 3.5916418699724537,
      "learning_rate": 2e-05,
      "loss": 0.5957374572753906,
      "step": 3830
    },
    {
      "epoch": 0.4911742133537989,
      "grad_norm": 0.7967894190370447,
      "learning_rate": 2e-05,
      "loss": 0.5781919002532959,
      "step": 3840
    },
    {
      "epoch": 0.49245331286774113,
      "grad_norm": 1.2068188085171687,
      "learning_rate": 2e-05,
      "loss": 0.5645539283752441,
      "step": 3850
    },
    {
      "epoch": 0.4937324123816833,
      "grad_norm": 0.8042966972737443,
      "learning_rate": 2e-05,
      "loss": 0.5781976222991944,
      "step": 3860
    },
    {
      "epoch": 0.4950115118956255,
      "grad_norm": 2.9731362460553186,
      "learning_rate": 2e-05,
      "loss": 0.6096286296844482,
      "step": 3870
    },
    {
      "epoch": 0.49629061140956765,
      "grad_norm": 1.8855978838078042,
      "learning_rate": 2e-05,
      "loss": 0.6072882652282715,
      "step": 3880
    },
    {
      "epoch": 0.4975697109235098,
      "grad_norm": 2.3279484427764148,
      "learning_rate": 2e-05,
      "loss": 0.593841552734375,
      "step": 3890
    },
    {
      "epoch": 0.49884881043745205,
      "grad_norm": 10.335316468198988,
      "learning_rate": 2e-05,
      "loss": 0.5867321968078614,
      "step": 3900
    },
    {
      "epoch": 0.5001279099513942,
      "grad_norm": 1.151612922573664,
      "learning_rate": 2e-05,
      "loss": 0.5656989574432373,
      "step": 3910
    },
    {
      "epoch": 0.5014070094653365,
      "grad_norm": 1.285187483922441,
      "learning_rate": 2e-05,
      "loss": 0.5880422115325927,
      "step": 3920
    },
    {
      "epoch": 0.5026861089792786,
      "grad_norm": 2.7311402622059107,
      "learning_rate": 2e-05,
      "loss": 0.6022061824798584,
      "step": 3930
    },
    {
      "epoch": 0.5039652084932208,
      "grad_norm": 1.0709412627660333,
      "learning_rate": 2e-05,
      "loss": 0.5973441123962402,
      "step": 3940
    },
    {
      "epoch": 0.5052443080071629,
      "grad_norm": 14.145026244622716,
      "learning_rate": 2e-05,
      "loss": 0.6096267700195312,
      "step": 3950
    },
    {
      "epoch": 0.5065234075211051,
      "grad_norm": 3.9916995890687144,
      "learning_rate": 2e-05,
      "loss": 0.5941912651062011,
      "step": 3960
    },
    {
      "epoch": 0.5078025070350474,
      "grad_norm": 1.536315111968428,
      "learning_rate": 2e-05,
      "loss": 0.5798390865325928,
      "step": 3970
    },
    {
      "epoch": 0.5090816065489895,
      "grad_norm": 3.038710043449597,
      "learning_rate": 2e-05,
      "loss": 0.5881974220275878,
      "step": 3980
    },
    {
      "epoch": 0.5103607060629317,
      "grad_norm": 1.7437382040110692,
      "learning_rate": 2e-05,
      "loss": 0.597667646408081,
      "step": 3990
    },
    {
      "epoch": 0.5116398055768738,
      "grad_norm": 2.2086375022016718,
      "learning_rate": 2e-05,
      "loss": 0.6139577388763428,
      "step": 4000
    },
    {
      "epoch": 0.5129189050908161,
      "grad_norm": 1.1094758092405541,
      "learning_rate": 2e-05,
      "loss": 0.5899992942810058,
      "step": 4010
    },
    {
      "epoch": 0.5141980046047583,
      "grad_norm": 3.191819345415452,
      "learning_rate": 2e-05,
      "loss": 0.5848158836364746,
      "step": 4020
    },
    {
      "epoch": 0.5154771041187004,
      "grad_norm": 18.343464601651103,
      "learning_rate": 2e-05,
      "loss": 0.6113421440124511,
      "step": 4030
    },
    {
      "epoch": 0.5167562036326426,
      "grad_norm": 8.151646256379413,
      "learning_rate": 2e-05,
      "loss": 0.6080557823181152,
      "step": 4040
    },
    {
      "epoch": 0.5180353031465849,
      "grad_norm": 1.4743567384823875,
      "learning_rate": 2e-05,
      "loss": 0.5929433345794678,
      "step": 4050
    },
    {
      "epoch": 0.519314402660527,
      "grad_norm": 1.6244297406185133,
      "learning_rate": 2e-05,
      "loss": 0.6228420257568359,
      "step": 4060
    },
    {
      "epoch": 0.5205935021744692,
      "grad_norm": 1.8964666797465644,
      "learning_rate": 2e-05,
      "loss": 0.6942201137542725,
      "step": 4070
    },
    {
      "epoch": 0.5218726016884113,
      "grad_norm": 1.5186283239632539,
      "learning_rate": 2e-05,
      "loss": 0.5778283596038818,
      "step": 4080
    },
    {
      "epoch": 0.5231517012023535,
      "grad_norm": 1.5658761832878125,
      "learning_rate": 2e-05,
      "loss": 0.6066390991210937,
      "step": 4090
    },
    {
      "epoch": 0.5244308007162958,
      "grad_norm": 1.3736386086641268,
      "learning_rate": 2e-05,
      "loss": 0.6242761135101318,
      "step": 4100
    },
    {
      "epoch": 0.5257099002302379,
      "grad_norm": 0.9262824430834768,
      "learning_rate": 2e-05,
      "loss": 0.599934196472168,
      "step": 4110
    },
    {
      "epoch": 0.5269889997441801,
      "grad_norm": 4.02490493429956,
      "learning_rate": 2e-05,
      "loss": 0.5782945156097412,
      "step": 4120
    },
    {
      "epoch": 0.5282680992581222,
      "grad_norm": 0.9676104257162095,
      "learning_rate": 2e-05,
      "loss": 0.6002942085266113,
      "step": 4130
    },
    {
      "epoch": 0.5295471987720645,
      "grad_norm": 0.8478665014039519,
      "learning_rate": 2e-05,
      "loss": 0.5808352470397949,
      "step": 4140
    },
    {
      "epoch": 0.5308262982860067,
      "grad_norm": 1.2619085918070871,
      "learning_rate": 2e-05,
      "loss": 0.573380994796753,
      "step": 4150
    },
    {
      "epoch": 0.5321053977999488,
      "grad_norm": 1.05574847518852,
      "learning_rate": 2e-05,
      "loss": 0.5796599388122559,
      "step": 4160
    },
    {
      "epoch": 0.533384497313891,
      "grad_norm": 1.0418655691663419,
      "learning_rate": 2e-05,
      "loss": 0.5615795135498047,
      "step": 4170
    },
    {
      "epoch": 0.5346635968278332,
      "grad_norm": 2.21038532959651,
      "learning_rate": 2e-05,
      "loss": 0.5999737739562988,
      "step": 4180
    },
    {
      "epoch": 0.5359426963417754,
      "grad_norm": 0.9439425400900301,
      "learning_rate": 2e-05,
      "loss": 0.5941405773162842,
      "step": 4190
    },
    {
      "epoch": 0.5372217958557176,
      "grad_norm": 1.6515809495540743,
      "learning_rate": 2e-05,
      "loss": 0.5667489528656006,
      "step": 4200
    },
    {
      "epoch": 0.5385008953696597,
      "grad_norm": 2.3318134698231283,
      "learning_rate": 2e-05,
      "loss": 0.5927960395812988,
      "step": 4210
    },
    {
      "epoch": 0.539779994883602,
      "grad_norm": 1.0377047326833577,
      "learning_rate": 2e-05,
      "loss": 0.5670997619628906,
      "step": 4220
    },
    {
      "epoch": 0.5410590943975442,
      "grad_norm": 1.077343508063325,
      "learning_rate": 2e-05,
      "loss": 0.583483076095581,
      "step": 4230
    },
    {
      "epoch": 0.5423381939114863,
      "grad_norm": 1.4911627128588618,
      "learning_rate": 2e-05,
      "loss": 0.5575456142425537,
      "step": 4240
    },
    {
      "epoch": 0.5436172934254285,
      "grad_norm": 1.0158421368691988,
      "learning_rate": 2e-05,
      "loss": 0.5972830772399902,
      "step": 4250
    },
    {
      "epoch": 0.5448963929393706,
      "grad_norm": 3.431073508965014,
      "learning_rate": 2e-05,
      "loss": 0.581250524520874,
      "step": 4260
    },
    {
      "epoch": 0.5461754924533129,
      "grad_norm": 1.8742004338445502,
      "learning_rate": 2e-05,
      "loss": 0.5883965969085694,
      "step": 4270
    },
    {
      "epoch": 0.5474545919672551,
      "grad_norm": 23.413683788939032,
      "learning_rate": 2e-05,
      "loss": 0.6253197669982911,
      "step": 4280
    },
    {
      "epoch": 0.5487336914811972,
      "grad_norm": 1.3611415143040233,
      "learning_rate": 2e-05,
      "loss": 0.571039342880249,
      "step": 4290
    },
    {
      "epoch": 0.5500127909951394,
      "grad_norm": 5.717662676715657,
      "learning_rate": 2e-05,
      "loss": 0.5469499111175538,
      "step": 4300
    },
    {
      "epoch": 0.5512918905090816,
      "grad_norm": 1.3827731318399812,
      "learning_rate": 2e-05,
      "loss": 0.5600070953369141,
      "step": 4310
    },
    {
      "epoch": 0.5525709900230238,
      "grad_norm": 1.2757442388514462,
      "learning_rate": 2e-05,
      "loss": 0.556539249420166,
      "step": 4320
    },
    {
      "epoch": 0.553850089536966,
      "grad_norm": 1.3527303490809621,
      "learning_rate": 2e-05,
      "loss": 0.620005989074707,
      "step": 4330
    },
    {
      "epoch": 0.5551291890509081,
      "grad_norm": 1.388222299690858,
      "learning_rate": 2e-05,
      "loss": 0.5785350322723388,
      "step": 4340
    },
    {
      "epoch": 0.5564082885648504,
      "grad_norm": 2.1801295381295502,
      "learning_rate": 2e-05,
      "loss": 0.5734769344329834,
      "step": 4350
    },
    {
      "epoch": 0.5576873880787925,
      "grad_norm": 2.4169523750802675,
      "learning_rate": 2e-05,
      "loss": 0.5844883918762207,
      "step": 4360
    },
    {
      "epoch": 0.5589664875927347,
      "grad_norm": 0.9705106323146543,
      "learning_rate": 2e-05,
      "loss": 0.5755261898040771,
      "step": 4370
    },
    {
      "epoch": 0.5602455871066769,
      "grad_norm": 1.227734351147137,
      "learning_rate": 2e-05,
      "loss": 0.5806452751159668,
      "step": 4380
    },
    {
      "epoch": 0.561524686620619,
      "grad_norm": 1.472801187168156,
      "learning_rate": 2e-05,
      "loss": 0.5938130378723144,
      "step": 4390
    },
    {
      "epoch": 0.5628037861345613,
      "grad_norm": 1.248661300619643,
      "learning_rate": 2e-05,
      "loss": 0.606907844543457,
      "step": 4400
    },
    {
      "epoch": 0.5640828856485035,
      "grad_norm": 1.015258501473048,
      "learning_rate": 2e-05,
      "loss": 0.5792621612548828,
      "step": 4410
    },
    {
      "epoch": 0.5653619851624456,
      "grad_norm": 3.7932070530676456,
      "learning_rate": 2e-05,
      "loss": 0.5865047454833985,
      "step": 4420
    },
    {
      "epoch": 0.5666410846763879,
      "grad_norm": 0.9418649037742004,
      "learning_rate": 2e-05,
      "loss": 0.5964035987854004,
      "step": 4430
    },
    {
      "epoch": 0.56792018419033,
      "grad_norm": 1.5100375997876072,
      "learning_rate": 2e-05,
      "loss": 0.5906060695648193,
      "step": 4440
    },
    {
      "epoch": 0.5691992837042722,
      "grad_norm": 4.0821023542551,
      "learning_rate": 2e-05,
      "loss": 0.5996895790100097,
      "step": 4450
    },
    {
      "epoch": 0.5704783832182144,
      "grad_norm": 0.9754984544620716,
      "learning_rate": 2e-05,
      "loss": 0.5960330009460449,
      "step": 4460
    },
    {
      "epoch": 0.5717574827321565,
      "grad_norm": 1.2091605398753102,
      "learning_rate": 2e-05,
      "loss": 0.5999294281005859,
      "step": 4470
    },
    {
      "epoch": 0.5730365822460988,
      "grad_norm": 0.9154880559845541,
      "learning_rate": 2e-05,
      "loss": 0.5727841854095459,
      "step": 4480
    },
    {
      "epoch": 0.5743156817600409,
      "grad_norm": 0.7205151402163276,
      "learning_rate": 2e-05,
      "loss": 0.5909416198730468,
      "step": 4490
    },
    {
      "epoch": 0.5755947812739831,
      "grad_norm": 1.2220796035555954,
      "learning_rate": 2e-05,
      "loss": 0.5693288803100586,
      "step": 4500
    },
    {
      "epoch": 0.5768738807879253,
      "grad_norm": 1.0619950103390499,
      "learning_rate": 2e-05,
      "loss": 0.5669358730316162,
      "step": 4510
    },
    {
      "epoch": 0.5781529803018675,
      "grad_norm": 0.9438439496920831,
      "learning_rate": 2e-05,
      "loss": 0.5788814067840576,
      "step": 4520
    },
    {
      "epoch": 0.5794320798158097,
      "grad_norm": 1.3637522696415814,
      "learning_rate": 2e-05,
      "loss": 0.5696746349334717,
      "step": 4530
    },
    {
      "epoch": 0.5807111793297518,
      "grad_norm": 1.104471035353041,
      "learning_rate": 2e-05,
      "loss": 0.615854549407959,
      "step": 4540
    },
    {
      "epoch": 0.581990278843694,
      "grad_norm": 2.1880095182583985,
      "learning_rate": 2e-05,
      "loss": 0.6321958065032959,
      "step": 4550
    },
    {
      "epoch": 0.5832693783576363,
      "grad_norm": 1.5508900350635002,
      "learning_rate": 2e-05,
      "loss": 0.5872769832611084,
      "step": 4560
    },
    {
      "epoch": 0.5845484778715784,
      "grad_norm": 0.913929968764322,
      "learning_rate": 2e-05,
      "loss": 0.5903765678405761,
      "step": 4570
    },
    {
      "epoch": 0.5858275773855206,
      "grad_norm": 1.5506874052946917,
      "learning_rate": 2e-05,
      "loss": 0.5771470546722413,
      "step": 4580
    },
    {
      "epoch": 0.5871066768994628,
      "grad_norm": 1.3405237997199888,
      "learning_rate": 2e-05,
      "loss": 0.627193546295166,
      "step": 4590
    },
    {
      "epoch": 0.588385776413405,
      "grad_norm": 1.131873746838468,
      "learning_rate": 2e-05,
      "loss": 0.5685714721679688,
      "step": 4600
    },
    {
      "epoch": 0.5896648759273472,
      "grad_norm": 0.9941722968260163,
      "learning_rate": 2e-05,
      "loss": 0.586639404296875,
      "step": 4610
    },
    {
      "epoch": 0.5909439754412893,
      "grad_norm": 1.0854285952546154,
      "learning_rate": 2e-05,
      "loss": 0.5921260356903076,
      "step": 4620
    },
    {
      "epoch": 0.5922230749552315,
      "grad_norm": 0.9273547382340928,
      "learning_rate": 2e-05,
      "loss": 0.5947554588317872,
      "step": 4630
    },
    {
      "epoch": 0.5935021744691737,
      "grad_norm": 1.2485551525495309,
      "learning_rate": 2e-05,
      "loss": 0.5865970611572265,
      "step": 4640
    },
    {
      "epoch": 0.5947812739831159,
      "grad_norm": 2.64035119270219,
      "learning_rate": 2e-05,
      "loss": 0.5649829864501953,
      "step": 4650
    },
    {
      "epoch": 0.5960603734970581,
      "grad_norm": 2.2264713784338177,
      "learning_rate": 2e-05,
      "loss": 0.5737303733825684,
      "step": 4660
    },
    {
      "epoch": 0.5973394730110002,
      "grad_norm": 5.11480061380236,
      "learning_rate": 2e-05,
      "loss": 0.5618254661560058,
      "step": 4670
    },
    {
      "epoch": 0.5986185725249424,
      "grad_norm": 1.5610706101720855,
      "learning_rate": 2e-05,
      "loss": 0.5706601619720459,
      "step": 4680
    },
    {
      "epoch": 0.5998976720388847,
      "grad_norm": 1.2708405823405424,
      "learning_rate": 2e-05,
      "loss": 0.5990711212158203,
      "step": 4690
    },
    {
      "epoch": 0.6011767715528268,
      "grad_norm": 2.0672195589982105,
      "learning_rate": 2e-05,
      "loss": 0.5991153717041016,
      "step": 4700
    },
    {
      "epoch": 0.602455871066769,
      "grad_norm": 1.5293908526762696,
      "learning_rate": 2e-05,
      "loss": 0.5510924339294434,
      "step": 4710
    },
    {
      "epoch": 0.6037349705807111,
      "grad_norm": 1.030805748856227,
      "learning_rate": 2e-05,
      "loss": 0.5742188930511475,
      "step": 4720
    },
    {
      "epoch": 0.6050140700946534,
      "grad_norm": 1.11507392591496,
      "learning_rate": 2e-05,
      "loss": 0.6045660972595215,
      "step": 4730
    },
    {
      "epoch": 0.6062931696085956,
      "grad_norm": 1.108089046744097,
      "learning_rate": 2e-05,
      "loss": 0.5572391510009765,
      "step": 4740
    },
    {
      "epoch": 0.6075722691225377,
      "grad_norm": 0.9596019314887728,
      "learning_rate": 2e-05,
      "loss": 0.5867826461791992,
      "step": 4750
    },
    {
      "epoch": 0.6088513686364799,
      "grad_norm": 2.2309423120020666,
      "learning_rate": 2e-05,
      "loss": 0.5885453701019288,
      "step": 4760
    },
    {
      "epoch": 0.6101304681504222,
      "grad_norm": 0.8122051055107862,
      "learning_rate": 2e-05,
      "loss": 0.5756945610046387,
      "step": 4770
    },
    {
      "epoch": 0.6114095676643643,
      "grad_norm": 0.7973634800311817,
      "learning_rate": 2e-05,
      "loss": 0.5984803676605225,
      "step": 4780
    },
    {
      "epoch": 0.6126886671783065,
      "grad_norm": 0.9811665022666481,
      "learning_rate": 2e-05,
      "loss": 0.5812541961669921,
      "step": 4790
    },
    {
      "epoch": 0.6139677666922486,
      "grad_norm": 0.8459850551659377,
      "learning_rate": 2e-05,
      "loss": 0.5552265167236328,
      "step": 4800
    },
    {
      "epoch": 0.6152468662061908,
      "grad_norm": 1.355516991977353,
      "learning_rate": 2e-05,
      "loss": 0.5846190452575684,
      "step": 4810
    },
    {
      "epoch": 0.6165259657201331,
      "grad_norm": 0.8399072480506798,
      "learning_rate": 2e-05,
      "loss": 0.602293872833252,
      "step": 4820
    },
    {
      "epoch": 0.6178050652340752,
      "grad_norm": 2.175988043437718,
      "learning_rate": 2e-05,
      "loss": 0.5567227363586426,
      "step": 4830
    },
    {
      "epoch": 0.6190841647480174,
      "grad_norm": 1.021804458594567,
      "learning_rate": 2e-05,
      "loss": 0.5912508964538574,
      "step": 4840
    },
    {
      "epoch": 0.6203632642619595,
      "grad_norm": 1.7173867718948452,
      "learning_rate": 2e-05,
      "loss": 0.5785565376281738,
      "step": 4850
    },
    {
      "epoch": 0.6216423637759018,
      "grad_norm": 1.1642729025096377,
      "learning_rate": 2e-05,
      "loss": 0.573655652999878,
      "step": 4860
    },
    {
      "epoch": 0.622921463289844,
      "grad_norm": 1.0152206639585626,
      "learning_rate": 2e-05,
      "loss": 0.600522518157959,
      "step": 4870
    },
    {
      "epoch": 0.6242005628037861,
      "grad_norm": 9.212059034328284,
      "learning_rate": 2e-05,
      "loss": 0.5742599487304687,
      "step": 4880
    },
    {
      "epoch": 0.6254796623177283,
      "grad_norm": 2.772497081272569,
      "learning_rate": 2e-05,
      "loss": 0.5782221794128418,
      "step": 4890
    },
    {
      "epoch": 0.6267587618316705,
      "grad_norm": 1.2266384866601492,
      "learning_rate": 2e-05,
      "loss": 0.6024472713470459,
      "step": 4900
    },
    {
      "epoch": 0.6280378613456127,
      "grad_norm": 1.5680142811027098,
      "learning_rate": 2e-05,
      "loss": 0.5703952312469482,
      "step": 4910
    },
    {
      "epoch": 0.6293169608595549,
      "grad_norm": 2.4224053712716516,
      "learning_rate": 2e-05,
      "loss": 0.5668153285980224,
      "step": 4920
    },
    {
      "epoch": 0.630596060373497,
      "grad_norm": 0.9412679194656801,
      "learning_rate": 2e-05,
      "loss": 0.5642160415649414,
      "step": 4930
    },
    {
      "epoch": 0.6318751598874393,
      "grad_norm": 1.6676548716952435,
      "learning_rate": 2e-05,
      "loss": 0.5556129455566406,
      "step": 4940
    },
    {
      "epoch": 0.6331542594013815,
      "grad_norm": 1.0913250915088346,
      "learning_rate": 2e-05,
      "loss": 0.5532308101654053,
      "step": 4950
    },
    {
      "epoch": 0.6344333589153236,
      "grad_norm": 0.8650100790088543,
      "learning_rate": 2e-05,
      "loss": 0.5819149017333984,
      "step": 4960
    },
    {
      "epoch": 0.6357124584292658,
      "grad_norm": 0.8561610497428719,
      "learning_rate": 2e-05,
      "loss": 0.5867252349853516,
      "step": 4970
    },
    {
      "epoch": 0.6369915579432079,
      "grad_norm": 0.8594777285935665,
      "learning_rate": 2e-05,
      "loss": 0.5625604152679443,
      "step": 4980
    },
    {
      "epoch": 0.6382706574571502,
      "grad_norm": 0.8081928805463781,
      "learning_rate": 2e-05,
      "loss": 0.57229585647583,
      "step": 4990
    },
    {
      "epoch": 0.6395497569710924,
      "grad_norm": 1.0962352741198798,
      "learning_rate": 2e-05,
      "loss": 0.5781308650970459,
      "step": 5000
    },
    {
      "epoch": 0.6408288564850345,
      "grad_norm": 0.8704133308867311,
      "learning_rate": 2e-05,
      "loss": 0.5762115478515625,
      "step": 5010
    },
    {
      "epoch": 0.6421079559989767,
      "grad_norm": 0.8906681137674894,
      "learning_rate": 2e-05,
      "loss": 0.5532569885253906,
      "step": 5020
    },
    {
      "epoch": 0.6433870555129189,
      "grad_norm": 3.5205291532714167,
      "learning_rate": 2e-05,
      "loss": 0.5602557182312011,
      "step": 5030
    },
    {
      "epoch": 0.6446661550268611,
      "grad_norm": 0.7880356154344685,
      "learning_rate": 2e-05,
      "loss": 0.582758092880249,
      "step": 5040
    },
    {
      "epoch": 0.6459452545408033,
      "grad_norm": 1.0494160383292277,
      "learning_rate": 2e-05,
      "loss": 0.5776301383972168,
      "step": 5050
    },
    {
      "epoch": 0.6472243540547454,
      "grad_norm": 0.8895822391817331,
      "learning_rate": 2e-05,
      "loss": 0.614192533493042,
      "step": 5060
    },
    {
      "epoch": 0.6485034535686877,
      "grad_norm": 1.173638043739992,
      "learning_rate": 2e-05,
      "loss": 0.6129615783691407,
      "step": 5070
    },
    {
      "epoch": 0.6497825530826298,
      "grad_norm": 1.4985544338626486,
      "learning_rate": 2e-05,
      "loss": 0.5763076305389404,
      "step": 5080
    },
    {
      "epoch": 0.651061652596572,
      "grad_norm": 0.9183496698629439,
      "learning_rate": 2e-05,
      "loss": 0.5809043407440185,
      "step": 5090
    },
    {
      "epoch": 0.6523407521105142,
      "grad_norm": 0.945248670588832,
      "learning_rate": 2e-05,
      "loss": 0.5711579322814941,
      "step": 5100
    },
    {
      "epoch": 0.6536198516244563,
      "grad_norm": 0.8263087226949879,
      "learning_rate": 2e-05,
      "loss": 0.5542846202850342,
      "step": 5110
    },
    {
      "epoch": 0.6548989511383986,
      "grad_norm": 1.4716049243155063,
      "learning_rate": 2e-05,
      "loss": 0.5838287353515625,
      "step": 5120
    },
    {
      "epoch": 0.6561780506523408,
      "grad_norm": 1.1282707596126975,
      "learning_rate": 2e-05,
      "loss": 0.5974137783050537,
      "step": 5130
    },
    {
      "epoch": 0.6574571501662829,
      "grad_norm": 0.8986907280666849,
      "learning_rate": 2e-05,
      "loss": 0.5346490859985351,
      "step": 5140
    },
    {
      "epoch": 0.6587362496802251,
      "grad_norm": 0.8101502209110832,
      "learning_rate": 2e-05,
      "loss": 0.5379027843475341,
      "step": 5150
    },
    {
      "epoch": 0.6600153491941673,
      "grad_norm": 1.2170438869870648,
      "learning_rate": 2e-05,
      "loss": 0.6006532669067383,
      "step": 5160
    },
    {
      "epoch": 0.6612944487081095,
      "grad_norm": 0.9459596560732497,
      "learning_rate": 2e-05,
      "loss": 0.5909769058227539,
      "step": 5170
    },
    {
      "epoch": 0.6625735482220517,
      "grad_norm": 0.9486200386219625,
      "learning_rate": 2e-05,
      "loss": 0.5767647743225097,
      "step": 5180
    },
    {
      "epoch": 0.6638526477359938,
      "grad_norm": 0.8844645270216186,
      "learning_rate": 2e-05,
      "loss": 0.6064223289489746,
      "step": 5190
    },
    {
      "epoch": 0.6651317472499361,
      "grad_norm": 1.1118874664756488,
      "learning_rate": 2e-05,
      "loss": 0.5366742610931396,
      "step": 5200
    },
    {
      "epoch": 0.6664108467638782,
      "grad_norm": 0.868634870218082,
      "learning_rate": 2e-05,
      "loss": 0.544734001159668,
      "step": 5210
    },
    {
      "epoch": 0.6676899462778204,
      "grad_norm": 0.8362326935410392,
      "learning_rate": 2e-05,
      "loss": 0.5742597103118896,
      "step": 5220
    },
    {
      "epoch": 0.6689690457917626,
      "grad_norm": 0.7661601835974579,
      "learning_rate": 2e-05,
      "loss": 0.589471435546875,
      "step": 5230
    },
    {
      "epoch": 0.6702481453057048,
      "grad_norm": 0.7835396706468216,
      "learning_rate": 2e-05,
      "loss": 0.5805575370788574,
      "step": 5240
    },
    {
      "epoch": 0.671527244819647,
      "grad_norm": 0.8823981631761211,
      "learning_rate": 2e-05,
      "loss": 0.5844844818115235,
      "step": 5250
    },
    {
      "epoch": 0.6728063443335891,
      "grad_norm": 0.8505388656474377,
      "learning_rate": 2e-05,
      "loss": 0.5834897518157959,
      "step": 5260
    },
    {
      "epoch": 0.6740854438475313,
      "grad_norm": 0.8534796171377586,
      "learning_rate": 2e-05,
      "loss": 0.5822929382324219,
      "step": 5270
    },
    {
      "epoch": 0.6753645433614736,
      "grad_norm": 0.89681982388502,
      "learning_rate": 2e-05,
      "loss": 0.5529897689819336,
      "step": 5280
    },
    {
      "epoch": 0.6766436428754157,
      "grad_norm": 0.8349533942038879,
      "learning_rate": 2e-05,
      "loss": 0.5589690685272217,
      "step": 5290
    },
    {
      "epoch": 0.6779227423893579,
      "grad_norm": 1.5671954749167862,
      "learning_rate": 2e-05,
      "loss": 0.5590453147888184,
      "step": 5300
    },
    {
      "epoch": 0.6792018419033001,
      "grad_norm": 0.8431941159099524,
      "learning_rate": 2e-05,
      "loss": 0.5770968437194824,
      "step": 5310
    },
    {
      "epoch": 0.6804809414172422,
      "grad_norm": 1.4911413616703288,
      "learning_rate": 2e-05,
      "loss": 0.5589974880218506,
      "step": 5320
    },
    {
      "epoch": 0.6817600409311845,
      "grad_norm": 0.9661248431328731,
      "learning_rate": 2e-05,
      "loss": 0.5906218528747559,
      "step": 5330
    },
    {
      "epoch": 0.6830391404451266,
      "grad_norm": 0.8692444217769545,
      "learning_rate": 2e-05,
      "loss": 0.5714686393737793,
      "step": 5340
    },
    {
      "epoch": 0.6843182399590688,
      "grad_norm": 1.0375095987437317,
      "learning_rate": 1.999194520388485e-05,
      "loss": 0.5932032585144043,
      "step": 5350
    },
    {
      "epoch": 0.685597339473011,
      "grad_norm": 1.003265235826933,
      "learning_rate": 1.9959247303157763e-05,
      "loss": 0.5622076034545899,
      "step": 5360
    },
    {
      "epoch": 0.6868764389869532,
      "grad_norm": 0.8843618523933533,
      "learning_rate": 1.990149424459817e-05,
      "loss": 0.5790035247802734,
      "step": 5370
    },
    {
      "epoch": 0.6881555385008954,
      "grad_norm": 4.084688608466956,
      "learning_rate": 1.9818847538648688e-05,
      "loss": 0.5558701992034912,
      "step": 5380
    },
    {
      "epoch": 0.6894346380148375,
      "grad_norm": 1.662594715956831,
      "learning_rate": 1.971153831256721e-05,
      "loss": 0.5631487846374512,
      "step": 5390
    },
    {
      "epoch": 0.6907137375287797,
      "grad_norm": 0.8693878605611629,
      "learning_rate": 1.9579866664063435e-05,
      "loss": 0.6056262016296386,
      "step": 5400
    },
    {
      "epoch": 0.691992837042722,
      "grad_norm": 0.9280530393550792,
      "learning_rate": 1.9424200822054782e-05,
      "loss": 0.5815097332000733,
      "step": 5410
    },
    {
      "epoch": 0.6932719365566641,
      "grad_norm": 1.4869043399097692,
      "learning_rate": 1.9244976116888626e-05,
      "loss": 0.5936915397644043,
      "step": 5420
    },
    {
      "epoch": 0.6945510360706063,
      "grad_norm": 0.8796971967347503,
      "learning_rate": 1.904269376291071e-05,
      "loss": 0.5816187858581543,
      "step": 5430
    },
    {
      "epoch": 0.6958301355845484,
      "grad_norm": 2.230087436913863,
      "learning_rate": 1.88179194567844e-05,
      "loss": 0.564141035079956,
      "step": 5440
    },
    {
      "epoch": 0.6971092350984907,
      "grad_norm": 3.087592356360664,
      "learning_rate": 1.8571281795480632e-05,
      "loss": 0.5804192543029785,
      "step": 5450
    },
    {
      "epoch": 0.6983883346124329,
      "grad_norm": 1.2651389758871605,
      "learning_rate": 1.8303470518362808e-05,
      "loss": 0.5868245601654053,
      "step": 5460
    },
    {
      "epoch": 0.699667434126375,
      "grad_norm": 0.8100251076204393,
      "learning_rate": 1.801523457828271e-05,
      "loss": 0.5611304759979248,
      "step": 5470
    },
    {
      "epoch": 0.7009465336403172,
      "grad_norm": 1.2934518508714499,
      "learning_rate": 1.7707380047081796e-05,
      "loss": 0.5905750274658204,
      "step": 5480
    },
    {
      "epoch": 0.7022256331542595,
      "grad_norm": 0.8030452831153584,
      "learning_rate": 1.7380767861355252e-05,
      "loss": 0.5805261611938477,
      "step": 5490
    },
    {
      "epoch": 0.7035047326682016,
      "grad_norm": 1.870863271729129,
      "learning_rate": 1.703631141478299e-05,
      "loss": 0.5857024192810059,
      "step": 5500
    },
    {
      "epoch": 0.7047838321821438,
      "grad_norm": 0.7653372068891764,
      "learning_rate": 1.6674974003760705e-05,
      "loss": 0.564706563949585,
      "step": 5510
    },
    {
      "epoch": 0.7060629316960859,
      "grad_norm": 1.380053635623579,
      "learning_rate": 1.6297766133474556e-05,
      "loss": 0.5737529754638672,
      "step": 5520
    },
    {
      "epoch": 0.7073420312100281,
      "grad_norm": 1.0000549748312881,
      "learning_rate": 1.5905742691953186e-05,
      "loss": 0.5773307800292968,
      "step": 5530
    },
    {
      "epoch": 0.7086211307239704,
      "grad_norm": 3.8926440753495335,
      "learning_rate": 1.55e-05,
      "loss": 0.5528308868408203,
      "step": 5540
    },
    {
      "epoch": 0.7099002302379125,
      "grad_norm": 1.5313117169146349,
      "learning_rate": 1.5081672745255862e-05,
      "loss": 0.6001224040985107,
      "step": 5550
    },
    {
      "epoch": 0.7111793297518547,
      "grad_norm": 1.2314308754735297,
      "learning_rate": 1.4651930808966206e-05,
      "loss": 0.5886702537536621,
      "step": 5560
    },
    {
      "epoch": 0.7124584292657968,
      "grad_norm": 0.930473270042156,
      "learning_rate": 1.4211975994326848e-05,
      "loss": 0.5943416118621826,
      "step": 5570
    },
    {
      "epoch": 0.7137375287797391,
      "grad_norm": 6.942932207612464,
      "learning_rate": 1.376303866555779e-05,
      "loss": 0.5635397911071778,
      "step": 5580
    },
    {
      "epoch": 0.7150166282936813,
      "grad_norm": 1.1593887746248703,
      "learning_rate": 1.3306374307104164e-05,
      "loss": 0.5360561370849609,
      "step": 5590
    },
    {
      "epoch": 0.7162957278076234,
      "grad_norm": 1.7139303356152429,
      "learning_rate": 1.2843260012586718e-05,
      "loss": 0.5889227867126465,
      "step": 5600
    },
    {
      "epoch": 0.7175748273215656,
      "grad_norm": 1.2942209037678882,
      "learning_rate": 1.2374990913320726e-05,
      "loss": 0.5862586975097657,
      "step": 5610
    },
    {
      "epoch": 0.7188539268355078,
      "grad_norm": 1.8320086711319004,
      "learning_rate": 1.1902876556391247e-05,
      "loss": 0.560733413696289,
      "step": 5620
    },
    {
      "epoch": 0.72013302634945,
      "grad_norm": 1.4490240711897515,
      "learning_rate": 1.1428237242413685e-05,
      "loss": 0.5700700283050537,
      "step": 5630
    },
    {
      "epoch": 0.7214121258633922,
      "grad_norm": 1.0928483173494343,
      "learning_rate": 1.095240033322132e-05,
      "loss": 0.53792724609375,
      "step": 5640
    },
    {
      "epoch": 0.7226912253773343,
      "grad_norm": 2.666768347905484,
      "learning_rate": 1.047669653980572e-05,
      "loss": 0.5752121925354003,
      "step": 5650
    },
    {
      "epoch": 0.7239703248912766,
      "grad_norm": 1.8821632144856029,
      "learning_rate": 1.0002456200890902e-05,
      "loss": 0.5423181056976318,
      "step": 5660
    },
    {
      "epoch": 0.7252494244052188,
      "grad_norm": 1.2844210347698184,
      "learning_rate": 9.531005562548668e-06,
      "loss": 0.5630322456359863,
      "step": 5670
    },
    {
      "epoch": 0.7265285239191609,
      "grad_norm": 2.239734646282082,
      "learning_rate": 9.06366306925925e-06,
      "loss": 0.568562126159668,
      "step": 5680
    },
    {
      "epoch": 0.7278076234331031,
      "grad_norm": 1.1132730006123257,
      "learning_rate": 8.601735676789686e-06,
      "loss": 0.6024523735046386,
      "step": 5690
    },
    {
      "epoch": 0.7290867229470452,
      "grad_norm": 1.0129300532679848,
      "learning_rate": 8.14651519720118e-06,
      "loss": 0.5717919826507568,
      "step": 5700
    },
    {
      "epoch": 0.7303658224609875,
      "grad_norm": 0.9394716201694613,
      "learning_rate": 7.699274686206866e-06,
      "loss": 0.5789447784423828,
      "step": 5710
    },
    {
      "epoch": 0.7316449219749297,
      "grad_norm": 1.1536081436240524,
      "learning_rate": 7.261264882983024e-06,
      "loss": 0.5560969352722168,
      "step": 5720
    },
    {
      "epoch": 0.7329240214888718,
      "grad_norm": 1.340372148285593,
      "learning_rate": 6.8337107123900386e-06,
      "loss": 0.55464186668396,
      "step": 5730
    },
    {
      "epoch": 0.734203121002814,
      "grad_norm": 1.0032194015370932,
      "learning_rate": 6.417807859384881e-06,
      "loss": 0.5238447666168213,
      "step": 5740
    },
    {
      "epoch": 0.7354822205167562,
      "grad_norm": 1.2259582109161506,
      "learning_rate": 6.014719425205009e-06,
      "loss": 0.5588017463684082,
      "step": 5750
    },
    {
      "epoch": 0.7367613200306984,
      "grad_norm": 2.435191846982898,
      "learning_rate": 5.6255726746749275e-06,
      "loss": 0.5562623500823974,
      "step": 5760
    },
    {
      "epoch": 0.7380404195446406,
      "grad_norm": 2.80865720993386,
      "learning_rate": 5.251455883731728e-06,
      "loss": 0.5464850425720215,
      "step": 5770
    },
    {
      "epoch": 0.7393195190585827,
      "grad_norm": 1.0687779980982417,
      "learning_rate": 4.893415295985812e-06,
      "loss": 0.5721700191497803,
      "step": 5780
    },
    {
      "epoch": 0.740598618572525,
      "grad_norm": 1.385350447071666,
      "learning_rate": 4.552452196827885e-06,
      "loss": 0.5811313152313232,
      "step": 5790
    },
    {
      "epoch": 0.7418777180864671,
      "grad_norm": 2.2195842379940474,
      "learning_rate": 4.229520113264785e-06,
      "loss": 0.5505681514739991,
      "step": 5800
    },
    {
      "epoch": 0.7431568176004093,
      "grad_norm": 0.945571151276205,
      "learning_rate": 3.925522147314915e-06,
      "loss": 0.5433897018432617,
      "step": 5810
    },
    {
      "epoch": 0.7444359171143515,
      "grad_norm": 1.284444125012454,
      "learning_rate": 3.641308450420675e-06,
      "loss": 0.5399296760559082,
      "step": 5820
    },
    {
      "epoch": 0.7457150166282936,
      "grad_norm": 0.9916111464325258,
      "learning_rate": 3.37767384594087e-06,
      "loss": 0.5575818538665771,
      "step": 5830
    },
    {
      "epoch": 0.7469941161422359,
      "grad_norm": 2.002666191869293,
      "learning_rate": 3.1353556063719357e-06,
      "loss": 0.5583402156829834,
      "step": 5840
    },
    {
      "epoch": 0.7482732156561781,
      "grad_norm": 1.0647467464926939,
      "learning_rate": 2.9150313915141474e-06,
      "loss": 0.568100881576538,
      "step": 5850
    },
    {
      "epoch": 0.7495523151701202,
      "grad_norm": 1.3399986512185353,
      "learning_rate": 2.717317353348891e-06,
      "loss": 0.5538657665252685,
      "step": 5860
    },
    {
      "epoch": 0.7508314146840624,
      "grad_norm": 0.9830757747750302,
      "learning_rate": 2.5427664129268253e-06,
      "loss": 0.5606857299804687,
      "step": 5870
    },
    {
      "epoch": 0.7521105141980046,
      "grad_norm": 1.6442338983834675,
      "learning_rate": 2.391866714085717e-06,
      "loss": 0.5571887493133545,
      "step": 5880
    },
    {
      "epoch": 0.7533896137119468,
      "grad_norm": 1.541187374388024,
      "learning_rate": 2.2650402583222753e-06,
      "loss": 0.551693344116211,
      "step": 5890
    },
    {
      "epoch": 0.754668713225889,
      "grad_norm": 1.2926329424423075,
      "learning_rate": 2.1626417246356398e-06,
      "loss": 0.5294806957244873,
      "step": 5900
    },
    {
      "epoch": 0.7559478127398311,
      "grad_norm": 1.1714042213695304,
      "learning_rate": 2.084957477642894e-06,
      "loss": 0.5456143856048584,
      "step": 5910
    },
    {
      "epoch": 0.7572269122537734,
      "grad_norm": 1.3260001825763719,
      "learning_rate": 2.0322047667405147e-06,
      "loss": 0.5545669078826905,
      "step": 5920
    },
    {
      "epoch": 0.7585060117677155,
      "grad_norm": 1.094544165497821,
      "learning_rate": 2.0045311185513342e-06,
      "loss": 0.5544045925140381,
      "step": 5930
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2810342056919040.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}