{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.26150089951169364,
  "eval_steps": 500,
  "global_step": 2035,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0012850167052171678,
      "grad_norm": 856.6189574701012,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.1357397079467773,
      "step": 10
    },
    {
      "epoch": 0.0025700334104343356,
      "grad_norm": 15.552391128873273,
      "learning_rate": 3.921348314606742e-06,
      "loss": 0.7902592658996582,
      "step": 20
    },
    {
      "epoch": 0.0038550501156515036,
      "grad_norm": 15.294031112683584,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.7169486045837402,
      "step": 30
    },
    {
      "epoch": 0.005140066820868671,
      "grad_norm": 14.61793688455095,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.6879290580749512,
      "step": 40
    },
    {
      "epoch": 0.006425083526085839,
      "grad_norm": 17.865572565928804,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.6837553977966309,
      "step": 50
    },
    {
      "epoch": 0.007710100231303007,
      "grad_norm": 78.02976802685448,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.6830409049987793,
      "step": 60
    },
    {
      "epoch": 0.008995116936520175,
      "grad_norm": 13.787148293237584,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.6678519725799561,
      "step": 70
    },
    {
      "epoch": 0.010280133641737342,
      "grad_norm": 59.88701484053759,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.671458625793457,
      "step": 80
    },
    {
      "epoch": 0.01156515034695451,
      "grad_norm": 14.98452507287525,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.6946072578430176,
      "step": 90
    },
    {
      "epoch": 0.012850167052171678,
      "grad_norm": 9.949341410744578,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.6994287014007569,
      "step": 100
    },
    {
      "epoch": 0.014135183757388847,
      "grad_norm": 20.083005842913845,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.654233169555664,
      "step": 110
    },
    {
      "epoch": 0.015420200462606014,
      "grad_norm": 52.69058821223778,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.6701527595520019,
      "step": 120
    },
    {
      "epoch": 0.01670521716782318,
      "grad_norm": 33.06654632843674,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.6579795837402344,
      "step": 130
    },
    {
      "epoch": 0.01799023387304035,
      "grad_norm": 26.181216707950213,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.6791215419769288,
      "step": 140
    },
    {
      "epoch": 0.01927525057825752,
      "grad_norm": 25.305185028986585,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.6644459724426269,
      "step": 150
    },
    {
      "epoch": 0.020560267283474685,
      "grad_norm": 4.88393687591383,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.6924647808074951,
      "step": 160
    },
    {
      "epoch": 0.021845283988691854,
      "grad_norm": 7.644492046355705,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.6735517024993897,
      "step": 170
    },
    {
      "epoch": 0.02313030069390902,
      "grad_norm": 4.0610145968849345,
      "learning_rate": 2e-05,
      "loss": 0.6673793315887451,
      "step": 180
    },
    {
      "epoch": 0.02441531739912619,
      "grad_norm": 9.704924216779212,
      "learning_rate": 2e-05,
      "loss": 0.6797656059265137,
      "step": 190
    },
    {
      "epoch": 0.025700334104343355,
      "grad_norm": 5.615408309945265,
      "learning_rate": 2e-05,
      "loss": 0.684023666381836,
      "step": 200
    },
    {
      "epoch": 0.026985350809560524,
      "grad_norm": 5.423083132935791,
      "learning_rate": 2e-05,
      "loss": 0.6746089935302735,
      "step": 210
    },
    {
      "epoch": 0.028270367514777694,
      "grad_norm": 8.79838661662345,
      "learning_rate": 2e-05,
      "loss": 0.6778485774993896,
      "step": 220
    },
    {
      "epoch": 0.02955538421999486,
      "grad_norm": 3.389056327858364,
      "learning_rate": 2e-05,
      "loss": 0.6816946029663086,
      "step": 230
    },
    {
      "epoch": 0.03084040092521203,
      "grad_norm": 3.2806365569635534,
      "learning_rate": 2e-05,
      "loss": 0.6739953517913818,
      "step": 240
    },
    {
      "epoch": 0.0321254176304292,
      "grad_norm": 2.714387951683758,
      "learning_rate": 2e-05,
      "loss": 0.6525035858154297,
      "step": 250
    },
    {
      "epoch": 0.03341043433564636,
      "grad_norm": 3.158516093565955,
      "learning_rate": 2e-05,
      "loss": 0.6718252182006836,
      "step": 260
    },
    {
      "epoch": 0.03469545104086353,
      "grad_norm": 5.772344202126518,
      "learning_rate": 2e-05,
      "loss": 0.647340965270996,
      "step": 270
    },
    {
      "epoch": 0.0359804677460807,
      "grad_norm": 5.502021173920203,
      "learning_rate": 2e-05,
      "loss": 0.6496793746948242,
      "step": 280
    },
    {
      "epoch": 0.03726548445129787,
      "grad_norm": 2.68008821860417,
      "learning_rate": 2e-05,
      "loss": 0.6709864616394043,
      "step": 290
    },
    {
      "epoch": 0.03855050115651504,
      "grad_norm": 1.7973533832528188,
      "learning_rate": 2e-05,
      "loss": 0.6566804885864258,
      "step": 300
    },
    {
      "epoch": 0.0398355178617322,
      "grad_norm": 6.51867504563315,
      "learning_rate": 2e-05,
      "loss": 0.6657544136047363,
      "step": 310
    },
    {
      "epoch": 0.04112053456694937,
      "grad_norm": 2.362613498811483,
      "learning_rate": 2e-05,
      "loss": 0.6502259731292724,
      "step": 320
    },
    {
      "epoch": 0.04240555127216654,
      "grad_norm": 3.907152138683563,
      "learning_rate": 2e-05,
      "loss": 0.6525882244110107,
      "step": 330
    },
    {
      "epoch": 0.04369056797738371,
      "grad_norm": 3.794880030440209,
      "learning_rate": 2e-05,
      "loss": 0.6532615661621094,
      "step": 340
    },
    {
      "epoch": 0.04497558468260087,
      "grad_norm": 8.769127576551398,
      "learning_rate": 2e-05,
      "loss": 0.6508918285369873,
      "step": 350
    },
    {
      "epoch": 0.04626060138781804,
      "grad_norm": 2.239128040127585,
      "learning_rate": 2e-05,
      "loss": 0.6526441097259521,
      "step": 360
    },
    {
      "epoch": 0.04754561809303521,
      "grad_norm": 3.6055555343647714,
      "learning_rate": 2e-05,
      "loss": 0.681905460357666,
      "step": 370
    },
    {
      "epoch": 0.04883063479825238,
      "grad_norm": 2.4554667315087624,
      "learning_rate": 2e-05,
      "loss": 0.6597890377044677,
      "step": 380
    },
    {
      "epoch": 0.05011565150346955,
      "grad_norm": 1.8867583686163962,
      "learning_rate": 2e-05,
      "loss": 0.6532648086547852,
      "step": 390
    },
    {
      "epoch": 0.05140066820868671,
      "grad_norm": 1.3820672524188946,
      "learning_rate": 2e-05,
      "loss": 0.653763723373413,
      "step": 400
    },
    {
      "epoch": 0.05268568491390388,
      "grad_norm": 4.353588095279891,
      "learning_rate": 2e-05,
      "loss": 0.6449775695800781,
      "step": 410
    },
    {
      "epoch": 0.05397070161912105,
      "grad_norm": 4.764374381297121,
      "learning_rate": 2e-05,
      "loss": 0.6555165767669677,
      "step": 420
    },
    {
      "epoch": 0.05525571832433822,
      "grad_norm": 14.727196337345648,
      "learning_rate": 2e-05,
      "loss": 0.6529878139495849,
      "step": 430
    },
    {
      "epoch": 0.05654073502955539,
      "grad_norm": 3.744804024990321,
      "learning_rate": 2e-05,
      "loss": 0.6880834579467774,
      "step": 440
    },
    {
      "epoch": 0.05782575173477255,
      "grad_norm": 9.9949909991041,
      "learning_rate": 2e-05,
      "loss": 0.6793588161468506,
      "step": 450
    },
    {
      "epoch": 0.05911076843998972,
      "grad_norm": 4.6856351506601595,
      "learning_rate": 2e-05,
      "loss": 0.6606217384338379,
      "step": 460
    },
    {
      "epoch": 0.06039578514520689,
      "grad_norm": 2.5156498724259575,
      "learning_rate": 2e-05,
      "loss": 0.6711392879486084,
      "step": 470
    },
    {
      "epoch": 0.06168080185042406,
      "grad_norm": 3.205814816900315,
      "learning_rate": 2e-05,
      "loss": 0.6852891921997071,
      "step": 480
    },
    {
      "epoch": 0.06296581855564122,
      "grad_norm": 1.839545887284569,
      "learning_rate": 2e-05,
      "loss": 0.6825028896331787,
      "step": 490
    },
    {
      "epoch": 0.0642508352608584,
      "grad_norm": 2.4792665803649783,
      "learning_rate": 2e-05,
      "loss": 0.6519015789031982,
      "step": 500
    },
    {
      "epoch": 0.06553585196607556,
      "grad_norm": 1.9240510741369987,
      "learning_rate": 2e-05,
      "loss": 0.648968505859375,
      "step": 510
    },
    {
      "epoch": 0.06682086867129272,
      "grad_norm": 2.5695541275375082,
      "learning_rate": 2e-05,
      "loss": 0.6466852188110351,
      "step": 520
    },
    {
      "epoch": 0.0681058853765099,
      "grad_norm": 11.336408841436898,
      "learning_rate": 2e-05,
      "loss": 0.6370265483856201,
      "step": 530
    },
    {
      "epoch": 0.06939090208172706,
      "grad_norm": 2.7714988957574405,
      "learning_rate": 2e-05,
      "loss": 0.6360719680786133,
      "step": 540
    },
    {
      "epoch": 0.07067591878694424,
      "grad_norm": 2.5399185097295414,
      "learning_rate": 2e-05,
      "loss": 0.6560811996459961,
      "step": 550
    },
    {
      "epoch": 0.0719609354921614,
      "grad_norm": 4.439208424776135,
      "learning_rate": 2e-05,
      "loss": 0.644292163848877,
      "step": 560
    },
    {
      "epoch": 0.07324595219737856,
      "grad_norm": 5.696317025602095,
      "learning_rate": 2e-05,
      "loss": 0.6401437282562256,
      "step": 570
    },
    {
      "epoch": 0.07453096890259574,
      "grad_norm": 2.664013724679843,
      "learning_rate": 2e-05,
      "loss": 0.6491299629211426,
      "step": 580
    },
    {
      "epoch": 0.0758159856078129,
      "grad_norm": 2.1745058771672237,
      "learning_rate": 2e-05,
      "loss": 0.63185133934021,
      "step": 590
    },
    {
      "epoch": 0.07710100231303008,
      "grad_norm": 3.0153668076131943,
      "learning_rate": 2e-05,
      "loss": 0.6398284912109375,
      "step": 600
    },
    {
      "epoch": 0.07838601901824724,
      "grad_norm": 4.211123263285935,
      "learning_rate": 2e-05,
      "loss": 0.645693016052246,
      "step": 610
    },
    {
      "epoch": 0.0796710357234644,
      "grad_norm": 9.078438471416925,
      "learning_rate": 2e-05,
      "loss": 0.6547775268554688,
      "step": 620
    },
    {
      "epoch": 0.08095605242868158,
      "grad_norm": 1.5458422123622622,
      "learning_rate": 2e-05,
      "loss": 0.6409534454345703,
      "step": 630
    },
    {
      "epoch": 0.08224106913389874,
      "grad_norm": 6.94798620371141,
      "learning_rate": 2e-05,
      "loss": 0.6655917644500733,
      "step": 640
    },
    {
      "epoch": 0.0835260858391159,
      "grad_norm": 1.271505509202168,
      "learning_rate": 2e-05,
      "loss": 0.6486823558807373,
      "step": 650
    },
    {
      "epoch": 0.08481110254433308,
      "grad_norm": 1.0317616781253849,
      "learning_rate": 2e-05,
      "loss": 0.6482730388641358,
      "step": 660
    },
    {
      "epoch": 0.08609611924955024,
      "grad_norm": 1.1601265249889448,
      "learning_rate": 2e-05,
      "loss": 0.6461831569671631,
      "step": 670
    },
    {
      "epoch": 0.08738113595476742,
      "grad_norm": 1.5955812719491171,
      "learning_rate": 2e-05,
      "loss": 0.6477439403533936,
      "step": 680
    },
    {
      "epoch": 0.08866615265998458,
      "grad_norm": 1.2974816270859992,
      "learning_rate": 2e-05,
      "loss": 0.630802059173584,
      "step": 690
    },
    {
      "epoch": 0.08995116936520174,
      "grad_norm": 1.802080446959691,
      "learning_rate": 2e-05,
      "loss": 0.6312804698944092,
      "step": 700
    },
    {
      "epoch": 0.09123618607041892,
      "grad_norm": 1.2734807542979136,
      "learning_rate": 2e-05,
      "loss": 0.6359667778015137,
      "step": 710
    },
    {
      "epoch": 0.09252120277563608,
      "grad_norm": 1.089715475560824,
      "learning_rate": 2e-05,
      "loss": 0.6190359115600585,
      "step": 720
    },
    {
      "epoch": 0.09380621948085326,
      "grad_norm": 1.4111528747217608,
      "learning_rate": 2e-05,
      "loss": 0.6462649822235107,
      "step": 730
    },
    {
      "epoch": 0.09509123618607042,
      "grad_norm": 2.3575893938615566,
      "learning_rate": 2e-05,
      "loss": 0.6176736354827881,
      "step": 740
    },
    {
      "epoch": 0.09637625289128758,
      "grad_norm": 4.099410091763762,
      "learning_rate": 2e-05,
      "loss": 0.6239109039306641,
      "step": 750
    },
    {
      "epoch": 0.09766126959650476,
      "grad_norm": 4.853333713828358,
      "learning_rate": 2e-05,
      "loss": 0.6444282531738281,
      "step": 760
    },
    {
      "epoch": 0.09894628630172192,
      "grad_norm": 2.5867597289295405,
      "learning_rate": 2e-05,
      "loss": 0.6411311626434326,
      "step": 770
    },
    {
      "epoch": 0.1002313030069391,
      "grad_norm": 3.048638070554145,
      "learning_rate": 2e-05,
      "loss": 0.6209658622741699,
      "step": 780
    },
    {
      "epoch": 0.10151631971215626,
      "grad_norm": 1.9668189975150188,
      "learning_rate": 2e-05,
      "loss": 0.6313361167907715,
      "step": 790
    },
    {
      "epoch": 0.10280133641737342,
      "grad_norm": 2.7558941074380496,
      "learning_rate": 2e-05,
      "loss": 0.6456878185272217,
      "step": 800
    },
    {
      "epoch": 0.1040863531225906,
      "grad_norm": 4.931251393074305,
      "learning_rate": 2e-05,
      "loss": 0.616514778137207,
      "step": 810
    },
    {
      "epoch": 0.10537136982780776,
      "grad_norm": 1.9224203127091826,
      "learning_rate": 2e-05,
      "loss": 0.6402089595794678,
      "step": 820
    },
    {
      "epoch": 0.10665638653302494,
      "grad_norm": 2.6967273056980985,
      "learning_rate": 2e-05,
      "loss": 0.6169169425964356,
      "step": 830
    },
    {
      "epoch": 0.1079414032382421,
      "grad_norm": 2.935113359723491,
      "learning_rate": 2e-05,
      "loss": 0.6237776756286622,
      "step": 840
    },
    {
      "epoch": 0.10922641994345926,
      "grad_norm": 5.369118446973871,
      "learning_rate": 2e-05,
      "loss": 0.6399550437927246,
      "step": 850
    },
    {
      "epoch": 0.11051143664867644,
      "grad_norm": 2.44005948506668,
      "learning_rate": 2e-05,
      "loss": 0.6399797439575196,
      "step": 860
    },
    {
      "epoch": 0.1117964533538936,
      "grad_norm": 9.032870562997907,
      "learning_rate": 2e-05,
      "loss": 0.6228564739227295,
      "step": 870
    },
    {
      "epoch": 0.11308147005911078,
      "grad_norm": 2.2222412773073255,
      "learning_rate": 2e-05,
      "loss": 0.6560338497161865,
      "step": 880
    },
    {
      "epoch": 0.11436648676432794,
      "grad_norm": 3.2511137004113686,
      "learning_rate": 2e-05,
      "loss": 0.637030029296875,
      "step": 890
    },
    {
      "epoch": 0.1156515034695451,
      "grad_norm": 2.19881020755354,
      "learning_rate": 2e-05,
      "loss": 0.6965466499328613,
      "step": 900
    },
    {
      "epoch": 0.11693652017476228,
      "grad_norm": 1.5817637004939435,
      "learning_rate": 2e-05,
      "loss": 0.6543152809143067,
      "step": 910
    },
    {
      "epoch": 0.11822153687997944,
      "grad_norm": 1.1276990840531103,
      "learning_rate": 2e-05,
      "loss": 0.6320334434509277,
      "step": 920
    },
    {
      "epoch": 0.1195065535851966,
      "grad_norm": 1.5922154872413146,
      "learning_rate": 2e-05,
      "loss": 0.617594051361084,
      "step": 930
    },
    {
      "epoch": 0.12079157029041378,
      "grad_norm": 2.504533267963736,
      "learning_rate": 2e-05,
      "loss": 0.609941816329956,
      "step": 940
    },
    {
      "epoch": 0.12207658699563094,
      "grad_norm": 1.0654835934253888,
      "learning_rate": 2e-05,
      "loss": 0.6268711090087891,
      "step": 950
    },
    {
      "epoch": 0.12336160370084812,
      "grad_norm": 1.3754530051132428,
      "learning_rate": 2e-05,
      "loss": 0.6483760833740234,
      "step": 960
    },
    {
      "epoch": 0.12464662040606528,
      "grad_norm": 2.0675680343817304,
      "learning_rate": 2e-05,
      "loss": 0.625995397567749,
      "step": 970
    },
    {
      "epoch": 0.12593163711128244,
      "grad_norm": 0.9779567253189434,
      "learning_rate": 2e-05,
      "loss": 0.6271099090576172,
      "step": 980
    },
    {
      "epoch": 0.1272166538164996,
      "grad_norm": 0.9982099110220216,
      "learning_rate": 2e-05,
      "loss": 0.61366286277771,
      "step": 990
    },
    {
      "epoch": 0.1285016705217168,
      "grad_norm": 1.0658286710551574,
      "learning_rate": 2e-05,
      "loss": 0.6421555519104004,
      "step": 1000
    },
    {
      "epoch": 0.12978668722693396,
      "grad_norm": 1.2925354348794225,
      "learning_rate": 2e-05,
      "loss": 0.6284059524536133,
      "step": 1010
    },
    {
      "epoch": 0.13107170393215112,
      "grad_norm": 1.0082444749487007,
      "learning_rate": 2e-05,
      "loss": 0.6287686824798584,
      "step": 1020
    },
    {
      "epoch": 0.13235672063736828,
      "grad_norm": 1.4045341222523071,
      "learning_rate": 2e-05,
      "loss": 0.6289882659912109,
      "step": 1030
    },
    {
      "epoch": 0.13364173734258544,
      "grad_norm": 1.1867690266586628,
      "learning_rate": 2e-05,
      "loss": 0.6120185852050781,
      "step": 1040
    },
    {
      "epoch": 0.13492675404780263,
      "grad_norm": 1.0489340200173105,
      "learning_rate": 2e-05,
      "loss": 0.6052242279052734,
      "step": 1050
    },
    {
      "epoch": 0.1362117707530198,
      "grad_norm": 1.3799884020710995,
      "learning_rate": 2e-05,
      "loss": 0.6165138244628906,
      "step": 1060
    },
    {
      "epoch": 0.13749678745823696,
      "grad_norm": 2.9370340336765803,
      "learning_rate": 2e-05,
      "loss": 0.6380349159240722,
      "step": 1070
    },
    {
      "epoch": 0.13878180416345412,
      "grad_norm": 1.0910366276944916,
      "learning_rate": 2e-05,
      "loss": 0.6134048461914062,
      "step": 1080
    },
    {
      "epoch": 0.14006682086867128,
      "grad_norm": 1.099783169446652,
      "learning_rate": 2e-05,
      "loss": 0.6116578578948975,
      "step": 1090
    },
    {
      "epoch": 0.14135183757388847,
      "grad_norm": 6.3506737723576645,
      "learning_rate": 2e-05,
      "loss": 0.6087802410125732,
      "step": 1100
    },
    {
      "epoch": 0.14263685427910563,
      "grad_norm": 1.830326258724304,
      "learning_rate": 2e-05,
      "loss": 0.6111043453216553,
      "step": 1110
    },
    {
      "epoch": 0.1439218709843228,
      "grad_norm": 1.4371773206636391,
      "learning_rate": 2e-05,
      "loss": 0.6332690238952636,
      "step": 1120
    },
    {
      "epoch": 0.14520688768953996,
      "grad_norm": 1.1346896882724038,
      "learning_rate": 2e-05,
      "loss": 0.6117291450500488,
      "step": 1130
    },
    {
      "epoch": 0.14649190439475712,
      "grad_norm": 1.229043968440211,
      "learning_rate": 2e-05,
      "loss": 0.634147834777832,
      "step": 1140
    },
    {
      "epoch": 0.1477769210999743,
      "grad_norm": 1.0643282673936258,
      "learning_rate": 2e-05,
      "loss": 0.6050711631774902,
      "step": 1150
    },
    {
      "epoch": 0.14906193780519147,
      "grad_norm": 0.9899183553255214,
      "learning_rate": 2e-05,
      "loss": 0.6159072875976562,
      "step": 1160
    },
    {
      "epoch": 0.15034695451040864,
      "grad_norm": 0.9756944895379619,
      "learning_rate": 2e-05,
      "loss": 0.6437289714813232,
      "step": 1170
    },
    {
      "epoch": 0.1516319712156258,
      "grad_norm": 1.5804530368583096,
      "learning_rate": 2e-05,
      "loss": 0.6059542655944824,
      "step": 1180
    },
    {
      "epoch": 0.15291698792084296,
      "grad_norm": 2.653732074916257,
      "learning_rate": 2e-05,
      "loss": 0.6155671119689942,
      "step": 1190
    },
    {
      "epoch": 0.15420200462606015,
      "grad_norm": 1.0526988484175226,
      "learning_rate": 2e-05,
      "loss": 0.607187032699585,
      "step": 1200
    },
    {
      "epoch": 0.15548702133127731,
      "grad_norm": 1.4066545066883747,
      "learning_rate": 2e-05,
      "loss": 0.6067052841186523,
      "step": 1210
    },
    {
      "epoch": 0.15677203803649448,
      "grad_norm": 1.14445816579696,
      "learning_rate": 2e-05,
      "loss": 0.6031442165374756,
      "step": 1220
    },
    {
      "epoch": 0.15805705474171164,
      "grad_norm": 0.9688379604966122,
      "learning_rate": 2e-05,
      "loss": 0.6165518760681152,
      "step": 1230
    },
    {
      "epoch": 0.1593420714469288,
      "grad_norm": 0.8892837373403698,
      "learning_rate": 2e-05,
      "loss": 0.6012338161468506,
      "step": 1240
    },
    {
      "epoch": 0.160627088152146,
      "grad_norm": 1.4152027685469248,
      "learning_rate": 2e-05,
      "loss": 0.5886805534362793,
      "step": 1250
    },
    {
      "epoch": 0.16191210485736315,
      "grad_norm": 2.1457539636692267,
      "learning_rate": 2e-05,
      "loss": 0.6081838130950927,
      "step": 1260
    },
    {
      "epoch": 0.16319712156258032,
      "grad_norm": 1.7040104665899867,
      "learning_rate": 2e-05,
      "loss": 0.6133224487304687,
      "step": 1270
    },
    {
      "epoch": 0.16448213826779748,
      "grad_norm": 1.2352798817335287,
      "learning_rate": 2e-05,
      "loss": 0.6118340492248535,
      "step": 1280
    },
    {
      "epoch": 0.16576715497301464,
      "grad_norm": 1.2198439029279724,
      "learning_rate": 2e-05,
      "loss": 0.6088254451751709,
      "step": 1290
    },
    {
      "epoch": 0.1670521716782318,
      "grad_norm": 3.6356160898778853,
      "learning_rate": 2e-05,
      "loss": 0.5922161102294922,
      "step": 1300
    },
    {
      "epoch": 0.168337188383449,
      "grad_norm": 1.434121136195825,
      "learning_rate": 2e-05,
      "loss": 0.6113036155700684,
      "step": 1310
    },
    {
      "epoch": 0.16962220508866616,
      "grad_norm": 1.8215041680182131,
      "learning_rate": 2e-05,
      "loss": 0.6138492584228515,
      "step": 1320
    },
    {
      "epoch": 0.17090722179388332,
      "grad_norm": 0.9733584275000338,
      "learning_rate": 2e-05,
      "loss": 0.6006131649017334,
      "step": 1330
    },
    {
      "epoch": 0.17219223849910048,
      "grad_norm": 0.8291757298058455,
      "learning_rate": 2e-05,
      "loss": 0.6175620555877686,
      "step": 1340
    },
    {
      "epoch": 0.17347725520431764,
      "grad_norm": 1.7778569183139799,
      "learning_rate": 2e-05,
      "loss": 0.6048360347747803,
      "step": 1350
    },
    {
      "epoch": 0.17476227190953483,
      "grad_norm": 2.3344157936720307,
      "learning_rate": 2e-05,
      "loss": 0.6050002098083496,
      "step": 1360
    },
    {
      "epoch": 0.176047288614752,
      "grad_norm": 3.506671656079186,
      "learning_rate": 2e-05,
      "loss": 0.6108762741088867,
      "step": 1370
    },
    {
      "epoch": 0.17733230531996916,
      "grad_norm": 1.1491594873900661,
      "learning_rate": 2e-05,
      "loss": 0.5949950695037842,
      "step": 1380
    },
    {
      "epoch": 0.17861732202518632,
      "grad_norm": 0.9698616451139325,
      "learning_rate": 2e-05,
      "loss": 0.6035851955413818,
      "step": 1390
    },
    {
      "epoch": 0.17990233873040348,
      "grad_norm": 2.3293601794536607,
      "learning_rate": 2e-05,
      "loss": 0.6041230201721192,
      "step": 1400
    },
    {
      "epoch": 0.18118735543562067,
      "grad_norm": 0.9819292447087378,
      "learning_rate": 2e-05,
      "loss": 0.6033549308776855,
      "step": 1410
    },
    {
      "epoch": 0.18247237214083784,
      "grad_norm": 5.127718733826317,
      "learning_rate": 2e-05,
      "loss": 0.5804622650146485,
      "step": 1420
    },
    {
      "epoch": 0.183757388846055,
      "grad_norm": 1.3831078732533943,
      "learning_rate": 2e-05,
      "loss": 0.5929896354675293,
      "step": 1430
    },
    {
      "epoch": 0.18504240555127216,
      "grad_norm": 1.409835795434802,
      "learning_rate": 2e-05,
      "loss": 0.620157527923584,
      "step": 1440
    },
    {
      "epoch": 0.18632742225648932,
      "grad_norm": 0.9787989385784489,
      "learning_rate": 2e-05,
      "loss": 0.6010499477386475,
      "step": 1450
    },
    {
      "epoch": 0.1876124389617065,
      "grad_norm": 1.125282063521845,
      "learning_rate": 2e-05,
      "loss": 0.6063499450683594,
      "step": 1460
    },
    {
      "epoch": 0.18889745566692367,
      "grad_norm": 1.1878340415798876,
      "learning_rate": 2e-05,
      "loss": 0.6011593818664551,
      "step": 1470
    },
    {
      "epoch": 0.19018247237214084,
      "grad_norm": 1.0586030412435297,
      "learning_rate": 2e-05,
      "loss": 0.5826192378997803,
      "step": 1480
    },
    {
      "epoch": 0.191467489077358,
      "grad_norm": 1.4030416349158283,
      "learning_rate": 2e-05,
      "loss": 0.5979415893554687,
      "step": 1490
    },
    {
      "epoch": 0.19275250578257516,
      "grad_norm": 0.9882974034594119,
      "learning_rate": 2e-05,
      "loss": 0.5784321784973144,
      "step": 1500
    },
    {
      "epoch": 0.19403752248779235,
      "grad_norm": 1.5827378342447296,
      "learning_rate": 2e-05,
      "loss": 0.5972190380096436,
      "step": 1510
    },
    {
      "epoch": 0.19532253919300951,
      "grad_norm": 1.772835353002146,
      "learning_rate": 2e-05,
      "loss": 0.5883693218231201,
      "step": 1520
    },
    {
      "epoch": 0.19660755589822668,
      "grad_norm": 0.9504309854461958,
      "learning_rate": 2e-05,
      "loss": 0.5878360748291016,
      "step": 1530
    },
    {
      "epoch": 0.19789257260344384,
      "grad_norm": 3.4573479282773754,
      "learning_rate": 2e-05,
      "loss": 0.5961189270019531,
      "step": 1540
    },
    {
      "epoch": 0.199177589308661,
      "grad_norm": 1.068378062020343,
      "learning_rate": 2e-05,
      "loss": 0.6108866691589355,
      "step": 1550
    },
    {
      "epoch": 0.2004626060138782,
      "grad_norm": 0.9475492216550739,
      "learning_rate": 2e-05,
      "loss": 0.5879310607910156,
      "step": 1560
    },
    {
      "epoch": 0.20174762271909535,
      "grad_norm": 2.6996846803236796,
      "learning_rate": 2e-05,
      "loss": 0.6067344665527343,
      "step": 1570
    },
    {
      "epoch": 0.20303263942431252,
      "grad_norm": 1.503312484489778,
      "learning_rate": 2e-05,
      "loss": 0.5966094970703125,
      "step": 1580
    },
    {
      "epoch": 0.20431765612952968,
      "grad_norm": 2.1144705955722105,
      "learning_rate": 2e-05,
      "loss": 0.601826810836792,
      "step": 1590
    },
    {
      "epoch": 0.20560267283474684,
      "grad_norm": 2.0298939231075686,
      "learning_rate": 2e-05,
      "loss": 0.613339900970459,
      "step": 1600
    },
    {
      "epoch": 0.20688768953996403,
      "grad_norm": 0.9838666275922071,
      "learning_rate": 2e-05,
      "loss": 0.5876136779785156,
      "step": 1610
    },
    {
      "epoch": 0.2081727062451812,
      "grad_norm": 1.3776226412627304,
      "learning_rate": 2e-05,
      "loss": 0.5954419612884522,
      "step": 1620
    },
    {
      "epoch": 0.20945772295039836,
      "grad_norm": 0.8085782872052595,
      "learning_rate": 2e-05,
      "loss": 0.595486831665039,
      "step": 1630
    },
    {
      "epoch": 0.21074273965561552,
      "grad_norm": 0.8662216027678745,
      "learning_rate": 2e-05,
      "loss": 0.5841521263122559,
      "step": 1640
    },
    {
      "epoch": 0.21202775636083268,
      "grad_norm": 1.2239187193929486,
      "learning_rate": 2e-05,
      "loss": 0.5863484859466552,
      "step": 1650
    },
    {
      "epoch": 0.21331277306604987,
      "grad_norm": 1.0494852911396721,
      "learning_rate": 2e-05,
      "loss": 0.5841124534606934,
      "step": 1660
    },
    {
      "epoch": 0.21459778977126703,
      "grad_norm": 2.981594663490346,
      "learning_rate": 2e-05,
      "loss": 0.5967587947845459,
      "step": 1670
    },
    {
      "epoch": 0.2158828064764842,
      "grad_norm": 1.296867754625597,
      "learning_rate": 2e-05,
      "loss": 0.6040875434875488,
      "step": 1680
    },
    {
      "epoch": 0.21716782318170136,
      "grad_norm": 1.6322116014400065,
      "learning_rate": 2e-05,
      "loss": 0.6083320617675781,
      "step": 1690
    },
    {
      "epoch": 0.21845283988691852,
      "grad_norm": 1.1441615552670734,
      "learning_rate": 2e-05,
      "loss": 0.6085511684417725,
      "step": 1700
    },
    {
      "epoch": 0.2197378565921357,
      "grad_norm": 0.9746942510805159,
      "learning_rate": 2e-05,
      "loss": 0.5959947586059571,
      "step": 1710
    },
    {
      "epoch": 0.22102287329735287,
      "grad_norm": 1.847667452156604,
      "learning_rate": 2e-05,
      "loss": 0.5838113307952881,
      "step": 1720
    },
    {
      "epoch": 0.22230789000257004,
      "grad_norm": 1.3014028403342264,
      "learning_rate": 2e-05,
      "loss": 0.5898797988891602,
      "step": 1730
    },
    {
      "epoch": 0.2235929067077872,
      "grad_norm": 1.794520627282578,
      "learning_rate": 2e-05,
      "loss": 0.5770421504974366,
      "step": 1740
    },
    {
      "epoch": 0.22487792341300436,
      "grad_norm": 1.181651271225444,
      "learning_rate": 2e-05,
      "loss": 0.5804369926452637,
      "step": 1750
    },
    {
      "epoch": 0.22616294011822155,
      "grad_norm": 1.4302258126941458,
      "learning_rate": 2e-05,
      "loss": 0.5999673843383789,
      "step": 1760
    },
    {
      "epoch": 0.2274479568234387,
      "grad_norm": 1.5184617174292843,
      "learning_rate": 2e-05,
      "loss": 0.601544713973999,
      "step": 1770
    },
    {
      "epoch": 0.22873297352865588,
      "grad_norm": 1.3090615344632057,
      "learning_rate": 2e-05,
      "loss": 0.5916304588317871,
      "step": 1780
    },
    {
      "epoch": 0.23001799023387304,
      "grad_norm": 1.084083761154048,
      "learning_rate": 2e-05,
      "loss": 0.599252462387085,
      "step": 1790
    },
    {
      "epoch": 0.2313030069390902,
      "grad_norm": 1.06362134663052,
      "learning_rate": 2e-05,
      "loss": 0.5892449855804444,
      "step": 1800
    },
    {
      "epoch": 0.23258802364430736,
      "grad_norm": 1.3218270041853413,
      "learning_rate": 2e-05,
      "loss": 0.5810562133789062,
      "step": 1810
    },
    {
      "epoch": 0.23387304034952455,
      "grad_norm": 1.8735021692385667,
      "learning_rate": 2e-05,
      "loss": 0.563934326171875,
      "step": 1820
    },
    {
      "epoch": 0.23515805705474171,
      "grad_norm": 1.1029961980816403,
      "learning_rate": 2e-05,
      "loss": 0.6061427116394043,
      "step": 1830
    },
    {
      "epoch": 0.23644307375995888,
      "grad_norm": 1.0956401424009607,
      "learning_rate": 2e-05,
      "loss": 0.580710506439209,
      "step": 1840
    },
    {
      "epoch": 0.23772809046517604,
      "grad_norm": 0.9022890666227324,
      "learning_rate": 2e-05,
      "loss": 0.5831015586853028,
      "step": 1850
    },
    {
      "epoch": 0.2390131071703932,
      "grad_norm": 0.9739413408546281,
      "learning_rate": 2e-05,
      "loss": 0.5803080558776855,
      "step": 1860
    },
    {
      "epoch": 0.2402981238756104,
      "grad_norm": 0.8887115233132612,
      "learning_rate": 2e-05,
      "loss": 0.5808173656463623,
      "step": 1870
    },
    {
      "epoch": 0.24158314058082755,
      "grad_norm": 1.2281390991341632,
      "learning_rate": 2e-05,
      "loss": 0.584747314453125,
      "step": 1880
    },
    {
      "epoch": 0.24286815728604472,
      "grad_norm": 1.0293295686123922,
      "learning_rate": 2e-05,
      "loss": 0.5973689079284668,
      "step": 1890
    },
    {
      "epoch": 0.24415317399126188,
      "grad_norm": 2.240993859946262,
      "learning_rate": 2e-05,
      "loss": 0.5756341934204101,
      "step": 1900
    },
    {
      "epoch": 0.24543819069647904,
      "grad_norm": 2.542763529237098,
      "learning_rate": 2e-05,
      "loss": 0.568706464767456,
      "step": 1910
    },
    {
      "epoch": 0.24672320740169623,
      "grad_norm": 1.8658118445309126,
      "learning_rate": 2e-05,
      "loss": 0.5815502166748047,
      "step": 1920
    },
    {
      "epoch": 0.2480082241069134,
      "grad_norm": 1.3680858039002373,
      "learning_rate": 2e-05,
      "loss": 0.5713666915893555,
      "step": 1930
    },
    {
      "epoch": 0.24929324081213056,
      "grad_norm": 1.4545423045311263,
      "learning_rate": 2e-05,
      "loss": 0.5888118267059326,
      "step": 1940
    },
    {
      "epoch": 0.25057825751734775,
      "grad_norm": 1.0001999301868607,
      "learning_rate": 2e-05,
      "loss": 0.581107234954834,
      "step": 1950
    },
    {
      "epoch": 0.2518632742225649,
      "grad_norm": 3.804939278531223,
      "learning_rate": 2e-05,
      "loss": 0.5979499816894531,
      "step": 1960
    },
    {
      "epoch": 0.25314829092778207,
      "grad_norm": 1.6052651400683768,
      "learning_rate": 2e-05,
      "loss": 0.5852889060974121,
      "step": 1970
    },
    {
      "epoch": 0.2544333076329992,
      "grad_norm": 1.465455399606115,
      "learning_rate": 2e-05,
      "loss": 0.5984594345092773,
      "step": 1980
    },
    {
      "epoch": 0.2557183243382164,
      "grad_norm": 0.9532879108802271,
      "learning_rate": 2e-05,
      "loss": 0.5933744430541992,
      "step": 1990
    },
    {
      "epoch": 0.2570033410434336,
      "grad_norm": 0.9971702008546999,
      "learning_rate": 2e-05,
      "loss": 0.5816676616668701,
      "step": 2000
    },
    {
      "epoch": 0.2582883577486507,
      "grad_norm": 0.9234236488287841,
      "learning_rate": 2e-05,
      "loss": 0.5809596538543701,
      "step": 2010
    },
    {
      "epoch": 0.2595733744538679,
      "grad_norm": 1.2689341107385237,
      "learning_rate": 2e-05,
      "loss": 0.5777791023254395,
      "step": 2020
    },
    {
      "epoch": 0.26085839115908505,
      "grad_norm": 2.127441593187859,
      "learning_rate": 2e-05,
      "loss": 0.5782732963562012,
      "step": 2030
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 963613493821440.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}