{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.3726364846870839,
  "eval_steps": 500,
  "global_step": 5597,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006657789613848203,
      "grad_norm": 235.8969256890802,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 1.7364160537719726,
      "step": 10
    },
    {
      "epoch": 0.0013315579227696406,
      "grad_norm": 454.87323539275144,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.213594627380371,
      "step": 20
    },
    {
      "epoch": 0.0019973368841544607,
      "grad_norm": 56.07729488657429,
      "learning_rate": 3.342134831460675e-06,
      "loss": 0.7420229911804199,
      "step": 30
    },
    {
      "epoch": 0.002663115845539281,
      "grad_norm": 150.0094098857343,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.5900703430175781,
      "step": 40
    },
    {
      "epoch": 0.003328894806924101,
      "grad_norm": 10.492010241968945,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.6019559383392334,
      "step": 50
    },
    {
      "epoch": 0.0039946737683089215,
      "grad_norm": 58.9158199879992,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.5498143672943115,
      "step": 60
    },
    {
      "epoch": 0.004660452729693742,
      "grad_norm": 38.7751489526521,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.5471227645874024,
      "step": 70
    },
    {
      "epoch": 0.005326231691078562,
      "grad_norm": 44.65639947244615,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.5947897911071778,
      "step": 80
    },
    {
      "epoch": 0.005992010652463382,
      "grad_norm": 42.021501768665296,
      "learning_rate": 1.005e-05,
      "loss": 0.5396684169769287,
      "step": 90
    },
    {
      "epoch": 0.006657789613848202,
      "grad_norm": 102.55835870245248,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.5591165065765381,
      "step": 100
    },
    {
      "epoch": 0.007323568575233023,
      "grad_norm": 63.550219661252996,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.5836419582366943,
      "step": 110
    },
    {
      "epoch": 0.007989347536617843,
      "grad_norm": 29.408506432576182,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.5985694885253906,
      "step": 120
    },
    {
      "epoch": 0.008655126498002663,
      "grad_norm": 8.348962303851348,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.6041747093200683,
      "step": 130
    },
    {
      "epoch": 0.009320905459387484,
      "grad_norm": 6.192722737720432,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.5608590602874756,
      "step": 140
    },
    {
      "epoch": 0.009986684420772303,
      "grad_norm": 9.660859490705201,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.5558311462402343,
      "step": 150
    },
    {
      "epoch": 0.010652463382157125,
      "grad_norm": 4.046195620530419,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.5265848636627197,
      "step": 160
    },
    {
      "epoch": 0.011318242343541944,
      "grad_norm": 7.900205661847944,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.5326771259307861,
      "step": 170
    },
    {
      "epoch": 0.011984021304926764,
      "grad_norm": 41.222719067968995,
      "learning_rate": 2e-05,
      "loss": 0.5493779182434082,
      "step": 180
    },
    {
      "epoch": 0.012649800266311585,
      "grad_norm": 11.51285962029523,
      "learning_rate": 2e-05,
      "loss": 0.5718745708465576,
      "step": 190
    },
    {
      "epoch": 0.013315579227696404,
      "grad_norm": 69.2723017393699,
      "learning_rate": 2e-05,
      "loss": 0.6055091381072998,
      "step": 200
    },
    {
      "epoch": 0.013981358189081226,
      "grad_norm": 31.148285201072877,
      "learning_rate": 2e-05,
      "loss": 0.5791210174560547,
      "step": 210
    },
    {
      "epoch": 0.014647137150466045,
      "grad_norm": 14.988316338719642,
      "learning_rate": 2e-05,
      "loss": 0.5553981781005859,
      "step": 220
    },
    {
      "epoch": 0.015312916111850865,
      "grad_norm": 5.657204329585924,
      "learning_rate": 2e-05,
      "loss": 0.5496438026428223,
      "step": 230
    },
    {
      "epoch": 0.015978695073235686,
      "grad_norm": 4.013199968748617,
      "learning_rate": 2e-05,
      "loss": 0.5337262630462647,
      "step": 240
    },
    {
      "epoch": 0.016644474034620507,
      "grad_norm": 39.164202345517396,
      "learning_rate": 2e-05,
      "loss": 0.5856839179992676,
      "step": 250
    },
    {
      "epoch": 0.017310252996005325,
      "grad_norm": 8.999791467884044,
      "learning_rate": 2e-05,
      "loss": 0.6439942836761474,
      "step": 260
    },
    {
      "epoch": 0.017976031957390146,
      "grad_norm": 12.92963394726804,
      "learning_rate": 2e-05,
      "loss": 0.6030735015869141,
      "step": 270
    },
    {
      "epoch": 0.018641810918774968,
      "grad_norm": 10.165083476794843,
      "learning_rate": 2e-05,
      "loss": 0.5539159297943115,
      "step": 280
    },
    {
      "epoch": 0.019307589880159785,
      "grad_norm": 7.211149855639102,
      "learning_rate": 2e-05,
      "loss": 0.5592308044433594,
      "step": 290
    },
    {
      "epoch": 0.019973368841544607,
      "grad_norm": 5.551546545660331,
      "learning_rate": 2e-05,
      "loss": 0.5586291790008545,
      "step": 300
    },
    {
      "epoch": 0.020639147802929428,
      "grad_norm": 14.196903064174798,
      "learning_rate": 2e-05,
      "loss": 0.5297722816467285,
      "step": 310
    },
    {
      "epoch": 0.02130492676431425,
      "grad_norm": 8.695901179045977,
      "learning_rate": 2e-05,
      "loss": 0.5313222885131836,
      "step": 320
    },
    {
      "epoch": 0.021970705725699067,
      "grad_norm": 73.2391576714464,
      "learning_rate": 2e-05,
      "loss": 0.5299333572387696,
      "step": 330
    },
    {
      "epoch": 0.022636484687083888,
      "grad_norm": 10.173641738816025,
      "learning_rate": 2e-05,
      "loss": 0.5360995292663574,
      "step": 340
    },
    {
      "epoch": 0.02330226364846871,
      "grad_norm": 6.198967892661863,
      "learning_rate": 2e-05,
      "loss": 0.5845535278320313,
      "step": 350
    },
    {
      "epoch": 0.023968042609853527,
      "grad_norm": 5.164050381832549,
      "learning_rate": 2e-05,
      "loss": 0.5637627601623535,
      "step": 360
    },
    {
      "epoch": 0.02463382157123835,
      "grad_norm": 4.33608376434225,
      "learning_rate": 2e-05,
      "loss": 0.5674751281738282,
      "step": 370
    },
    {
      "epoch": 0.02529960053262317,
      "grad_norm": 6.676449836472397,
      "learning_rate": 2e-05,
      "loss": 0.5269466400146484,
      "step": 380
    },
    {
      "epoch": 0.02596537949400799,
      "grad_norm": 34.60400985665482,
      "learning_rate": 2e-05,
      "loss": 0.5582229614257812,
      "step": 390
    },
    {
      "epoch": 0.02663115845539281,
      "grad_norm": 2.0382084950493686,
      "learning_rate": 2e-05,
      "loss": 0.533026123046875,
      "step": 400
    },
    {
      "epoch": 0.02729693741677763,
      "grad_norm": 12.318088970720998,
      "learning_rate": 2e-05,
      "loss": 0.5565634727478027,
      "step": 410
    },
    {
      "epoch": 0.02796271637816245,
      "grad_norm": 5.5824838209854395,
      "learning_rate": 2e-05,
      "loss": 0.5633725643157959,
      "step": 420
    },
    {
      "epoch": 0.02862849533954727,
      "grad_norm": 2.2430627781253163,
      "learning_rate": 2e-05,
      "loss": 0.5287482738494873,
      "step": 430
    },
    {
      "epoch": 0.02929427430093209,
      "grad_norm": 3.386109702013143,
      "learning_rate": 2e-05,
      "loss": 0.5164429187774658,
      "step": 440
    },
    {
      "epoch": 0.02996005326231691,
      "grad_norm": 4.52797743621058,
      "learning_rate": 2e-05,
      "loss": 0.49767842292785647,
      "step": 450
    },
    {
      "epoch": 0.03062583222370173,
      "grad_norm": 8.559435723184958,
      "learning_rate": 2e-05,
      "loss": 0.545814323425293,
      "step": 460
    },
    {
      "epoch": 0.031291611185086554,
      "grad_norm": 2.4213157821392652,
      "learning_rate": 2e-05,
      "loss": 0.5681285858154297,
      "step": 470
    },
    {
      "epoch": 0.03195739014647137,
      "grad_norm": 5.002642882306833,
      "learning_rate": 2e-05,
      "loss": 0.5424331188201904,
      "step": 480
    },
    {
      "epoch": 0.03262316910785619,
      "grad_norm": 11.069858537054463,
      "learning_rate": 2e-05,
      "loss": 0.5557833671569824,
      "step": 490
    },
    {
      "epoch": 0.033288948069241014,
      "grad_norm": 10.790349593013875,
      "learning_rate": 2e-05,
      "loss": 0.5815642833709717,
      "step": 500
    },
    {
      "epoch": 0.03395472703062583,
      "grad_norm": 3.3611249244779615,
      "learning_rate": 2e-05,
      "loss": 0.5631533622741699,
      "step": 510
    },
    {
      "epoch": 0.03462050599201065,
      "grad_norm": 5.726005842469412,
      "learning_rate": 2e-05,
      "loss": 0.569805908203125,
      "step": 520
    },
    {
      "epoch": 0.035286284953395475,
      "grad_norm": 14.583572065182688,
      "learning_rate": 2e-05,
      "loss": 0.5803721904754638,
      "step": 530
    },
    {
      "epoch": 0.03595206391478029,
      "grad_norm": 4.297489848956763,
      "learning_rate": 2e-05,
      "loss": 0.5814414024353027,
      "step": 540
    },
    {
      "epoch": 0.03661784287616511,
      "grad_norm": 5.944128076013417,
      "learning_rate": 2e-05,
      "loss": 0.564124059677124,
      "step": 550
    },
    {
      "epoch": 0.037283621837549935,
      "grad_norm": 3.2423104393351307,
      "learning_rate": 2e-05,
      "loss": 0.6266448974609375,
      "step": 560
    },
    {
      "epoch": 0.03794940079893475,
      "grad_norm": 43.85395650355969,
      "learning_rate": 2e-05,
      "loss": 0.572756004333496,
      "step": 570
    },
    {
      "epoch": 0.03861517976031957,
      "grad_norm": 3.9530294075080237,
      "learning_rate": 2e-05,
      "loss": 0.5774184226989746,
      "step": 580
    },
    {
      "epoch": 0.039280958721704395,
      "grad_norm": 11.062180432200206,
      "learning_rate": 2e-05,
      "loss": 0.5977572441101074,
      "step": 590
    },
    {
      "epoch": 0.03994673768308921,
      "grad_norm": 6.443399893752391,
      "learning_rate": 2e-05,
      "loss": 0.5534317016601562,
      "step": 600
    },
    {
      "epoch": 0.04061251664447404,
      "grad_norm": 18.52097082878789,
      "learning_rate": 2e-05,
      "loss": 0.554855489730835,
      "step": 610
    },
    {
      "epoch": 0.041278295605858856,
      "grad_norm": 3.3934097540968753,
      "learning_rate": 2e-05,
      "loss": 0.5600467681884765,
      "step": 620
    },
    {
      "epoch": 0.041944074567243674,
      "grad_norm": 7.706610686679725,
      "learning_rate": 2e-05,
      "loss": 0.5719536781311035,
      "step": 630
    },
    {
      "epoch": 0.0426098535286285,
      "grad_norm": 4.425720098798294,
      "learning_rate": 2e-05,
      "loss": 0.5739123344421386,
      "step": 640
    },
    {
      "epoch": 0.043275632490013316,
      "grad_norm": 1.7190184550483938,
      "learning_rate": 2e-05,
      "loss": 0.546561336517334,
      "step": 650
    },
    {
      "epoch": 0.043941411451398134,
      "grad_norm": 8.48706166477749,
      "learning_rate": 2e-05,
      "loss": 0.5902390480041504,
      "step": 660
    },
    {
      "epoch": 0.04460719041278296,
      "grad_norm": 2.034988576266325,
      "learning_rate": 2e-05,
      "loss": 0.5392104148864746,
      "step": 670
    },
    {
      "epoch": 0.045272969374167776,
      "grad_norm": 2.8908851748120794,
      "learning_rate": 2e-05,
      "loss": 0.5352708339691162,
      "step": 680
    },
    {
      "epoch": 0.045938748335552594,
      "grad_norm": 11.225417918014127,
      "learning_rate": 2e-05,
      "loss": 0.5520500183105469,
      "step": 690
    },
    {
      "epoch": 0.04660452729693742,
      "grad_norm": 3.063402372041763,
      "learning_rate": 2e-05,
      "loss": 0.5145657539367676,
      "step": 700
    },
    {
      "epoch": 0.04727030625832224,
      "grad_norm": 6.11767198069841,
      "learning_rate": 2e-05,
      "loss": 0.5366698265075683,
      "step": 710
    },
    {
      "epoch": 0.047936085219707054,
      "grad_norm": 2.25307348329916,
      "learning_rate": 2e-05,
      "loss": 0.551693344116211,
      "step": 720
    },
    {
      "epoch": 0.04860186418109188,
      "grad_norm": 4.55732420618067,
      "learning_rate": 2e-05,
      "loss": 0.5258326530456543,
      "step": 730
    },
    {
      "epoch": 0.0492676431424767,
      "grad_norm": 3.4702813032613635,
      "learning_rate": 2e-05,
      "loss": 0.5708011627197266,
      "step": 740
    },
    {
      "epoch": 0.049933422103861515,
      "grad_norm": 2.195262021450343,
      "learning_rate": 2e-05,
      "loss": 0.5412505149841309,
      "step": 750
    },
    {
      "epoch": 0.05059920106524634,
      "grad_norm": 4.418015821922213,
      "learning_rate": 2e-05,
      "loss": 0.5207402229309082,
      "step": 760
    },
    {
      "epoch": 0.05126498002663116,
      "grad_norm": 2.6654030789993564,
      "learning_rate": 2e-05,
      "loss": 0.5646546840667724,
      "step": 770
    },
    {
      "epoch": 0.05193075898801598,
      "grad_norm": 1.463701773082979,
      "learning_rate": 2e-05,
      "loss": 0.5652810573577881,
      "step": 780
    },
    {
      "epoch": 0.0525965379494008,
      "grad_norm": 1.1644755512746217,
      "learning_rate": 2e-05,
      "loss": 0.5259516716003418,
      "step": 790
    },
    {
      "epoch": 0.05326231691078562,
      "grad_norm": 1.4724936122706427,
      "learning_rate": 2e-05,
      "loss": 0.5653295040130615,
      "step": 800
    },
    {
      "epoch": 0.05392809587217044,
      "grad_norm": 1.451947446736679,
      "learning_rate": 2e-05,
      "loss": 0.523100757598877,
      "step": 810
    },
    {
      "epoch": 0.05459387483355526,
      "grad_norm": 1.2453818129937126,
      "learning_rate": 2e-05,
      "loss": 0.5161214351654053,
      "step": 820
    },
    {
      "epoch": 0.05525965379494008,
      "grad_norm": 8.366800471437534,
      "learning_rate": 2e-05,
      "loss": 0.5262998580932617,
      "step": 830
    },
    {
      "epoch": 0.0559254327563249,
      "grad_norm": 2.239038325903991,
      "learning_rate": 2e-05,
      "loss": 0.5443816184997559,
      "step": 840
    },
    {
      "epoch": 0.05659121171770972,
      "grad_norm": 2.1435511292161107,
      "learning_rate": 2e-05,
      "loss": 0.5617143154144287,
      "step": 850
    },
    {
      "epoch": 0.05725699067909454,
      "grad_norm": 1.6208532172828776,
      "learning_rate": 2e-05,
      "loss": 0.5223365783691406,
      "step": 860
    },
    {
      "epoch": 0.05792276964047936,
      "grad_norm": 1.552594266356934,
      "learning_rate": 2e-05,
      "loss": 0.5362037181854248,
      "step": 870
    },
    {
      "epoch": 0.05858854860186418,
      "grad_norm": 1.732568906745257,
      "learning_rate": 2e-05,
      "loss": 0.5471210479736328,
      "step": 880
    },
    {
      "epoch": 0.059254327563249,
      "grad_norm": 3.922703835213154,
      "learning_rate": 2e-05,
      "loss": 0.5224350929260254,
      "step": 890
    },
    {
      "epoch": 0.05992010652463382,
      "grad_norm": 1.8443327131294107,
      "learning_rate": 2e-05,
      "loss": 0.5521405220031739,
      "step": 900
    },
    {
      "epoch": 0.06058588548601864,
      "grad_norm": 18.041405872499528,
      "learning_rate": 2e-05,
      "loss": 0.5715356349945069,
      "step": 910
    },
    {
      "epoch": 0.06125166444740346,
      "grad_norm": 1.975134294116238,
      "learning_rate": 2e-05,
      "loss": 0.5748860836029053,
      "step": 920
    },
    {
      "epoch": 0.061917443408788284,
      "grad_norm": 1.0758923518754975,
      "learning_rate": 2e-05,
      "loss": 0.5490294456481933,
      "step": 930
    },
    {
      "epoch": 0.06258322237017311,
      "grad_norm": 3.1323788152588383,
      "learning_rate": 2e-05,
      "loss": 0.5143556118011474,
      "step": 940
    },
    {
      "epoch": 0.06324900133155792,
      "grad_norm": 2.312493834654343,
      "learning_rate": 2e-05,
      "loss": 0.4982307434082031,
      "step": 950
    },
    {
      "epoch": 0.06391478029294274,
      "grad_norm": 1.8957287997713355,
      "learning_rate": 2e-05,
      "loss": 0.5262045383453369,
      "step": 960
    },
    {
      "epoch": 0.06458055925432757,
      "grad_norm": 1.180103969220926,
      "learning_rate": 2e-05,
      "loss": 0.5141486167907715,
      "step": 970
    },
    {
      "epoch": 0.06524633821571238,
      "grad_norm": 1.5035058067710745,
      "learning_rate": 2e-05,
      "loss": 0.5460874080657959,
      "step": 980
    },
    {
      "epoch": 0.0659121171770972,
      "grad_norm": 6.3846590160328445,
      "learning_rate": 2e-05,
      "loss": 0.5823369026184082,
      "step": 990
    },
    {
      "epoch": 0.06657789613848203,
      "grad_norm": 3.231925320560546,
      "learning_rate": 2e-05,
      "loss": 0.5262383460998535,
      "step": 1000
    },
    {
      "epoch": 0.06724367509986684,
      "grad_norm": 2.315135540989524,
      "learning_rate": 2e-05,
      "loss": 0.5371553421020507,
      "step": 1010
    },
    {
      "epoch": 0.06790945406125166,
      "grad_norm": 4.496108973654741,
      "learning_rate": 2e-05,
      "loss": 0.5573309898376465,
      "step": 1020
    },
    {
      "epoch": 0.06857523302263649,
      "grad_norm": 2.166198851365719,
      "learning_rate": 2e-05,
      "loss": 0.5329337120056152,
      "step": 1030
    },
    {
      "epoch": 0.0692410119840213,
      "grad_norm": 1.0425981603986465,
      "learning_rate": 2e-05,
      "loss": 0.5511478424072266,
      "step": 1040
    },
    {
      "epoch": 0.06990679094540612,
      "grad_norm": 2.2014938720266937,
      "learning_rate": 2e-05,
      "loss": 0.5591570854187011,
      "step": 1050
    },
    {
      "epoch": 0.07057256990679095,
      "grad_norm": 4.876753934559851,
      "learning_rate": 2e-05,
      "loss": 0.5555380821228028,
      "step": 1060
    },
    {
      "epoch": 0.07123834886817576,
      "grad_norm": 2.857934925368923,
      "learning_rate": 2e-05,
      "loss": 0.5619023323059082,
      "step": 1070
    },
    {
      "epoch": 0.07190412782956059,
      "grad_norm": 2.3266365189381832,
      "learning_rate": 2e-05,
      "loss": 0.5243755340576172,
      "step": 1080
    },
    {
      "epoch": 0.07256990679094541,
      "grad_norm": 1.9277942910270909,
      "learning_rate": 2e-05,
      "loss": 0.5400153160095215,
      "step": 1090
    },
    {
      "epoch": 0.07323568575233022,
      "grad_norm": 1.1764647055530073,
      "learning_rate": 2e-05,
      "loss": 0.5243360519409179,
      "step": 1100
    },
    {
      "epoch": 0.07390146471371505,
      "grad_norm": 1.1058092372935702,
      "learning_rate": 2e-05,
      "loss": 0.544226598739624,
      "step": 1110
    },
    {
      "epoch": 0.07456724367509987,
      "grad_norm": 2.841445800892787,
      "learning_rate": 2e-05,
      "loss": 0.550119161605835,
      "step": 1120
    },
    {
      "epoch": 0.07523302263648468,
      "grad_norm": 1.4562183729193323,
      "learning_rate": 2e-05,
      "loss": 0.5335268020629883,
      "step": 1130
    },
    {
      "epoch": 0.0758988015978695,
      "grad_norm": 2.5489855904207768,
      "learning_rate": 2e-05,
      "loss": 0.5286331176757812,
      "step": 1140
    },
    {
      "epoch": 0.07656458055925433,
      "grad_norm": 4.365595677900958,
      "learning_rate": 2e-05,
      "loss": 0.5628453254699707,
      "step": 1150
    },
    {
      "epoch": 0.07723035952063914,
      "grad_norm": 1.8349543210000938,
      "learning_rate": 2e-05,
      "loss": 0.5462607860565185,
      "step": 1160
    },
    {
      "epoch": 0.07789613848202397,
      "grad_norm": 2.5076678969731416,
      "learning_rate": 2e-05,
      "loss": 0.5364606380462646,
      "step": 1170
    },
    {
      "epoch": 0.07856191744340879,
      "grad_norm": 8.268429157439165,
      "learning_rate": 2e-05,
      "loss": 0.5362617492675781,
      "step": 1180
    },
    {
      "epoch": 0.0792276964047936,
      "grad_norm": 2.4583852803700648,
      "learning_rate": 2e-05,
      "loss": 0.5119704246520996,
      "step": 1190
    },
    {
      "epoch": 0.07989347536617843,
      "grad_norm": 2.1784687391986437,
      "learning_rate": 2e-05,
      "loss": 0.5308552742004394,
      "step": 1200
    },
    {
      "epoch": 0.08055925432756325,
      "grad_norm": 1.2303988845257494,
      "learning_rate": 2e-05,
      "loss": 0.539813232421875,
      "step": 1210
    },
    {
      "epoch": 0.08122503328894808,
      "grad_norm": 8.727296596355615,
      "learning_rate": 2e-05,
      "loss": 0.5063223838806152,
      "step": 1220
    },
    {
      "epoch": 0.08189081225033289,
      "grad_norm": 1.6512049866144545,
      "learning_rate": 2e-05,
      "loss": 0.5685141563415528,
      "step": 1230
    },
    {
      "epoch": 0.08255659121171771,
      "grad_norm": 1.8297074024323168,
      "learning_rate": 2e-05,
      "loss": 0.5602605819702149,
      "step": 1240
    },
    {
      "epoch": 0.08322237017310254,
      "grad_norm": 1.050479235459917,
      "learning_rate": 2e-05,
      "loss": 0.5343076229095459,
      "step": 1250
    },
    {
      "epoch": 0.08388814913448735,
      "grad_norm": 0.9525187920106077,
      "learning_rate": 2e-05,
      "loss": 0.5467718124389649,
      "step": 1260
    },
    {
      "epoch": 0.08455392809587217,
      "grad_norm": 5.619490448508517,
      "learning_rate": 2e-05,
      "loss": 0.5230856895446777,
      "step": 1270
    },
    {
      "epoch": 0.085219707057257,
      "grad_norm": 31.785133935281873,
      "learning_rate": 2e-05,
      "loss": 0.5371414184570312,
      "step": 1280
    },
    {
      "epoch": 0.08588548601864181,
      "grad_norm": 1.6841285905458097,
      "learning_rate": 2e-05,
      "loss": 0.5382813930511474,
      "step": 1290
    },
    {
      "epoch": 0.08655126498002663,
      "grad_norm": 3.074425583552512,
      "learning_rate": 2e-05,
      "loss": 0.5496634483337403,
      "step": 1300
    },
    {
      "epoch": 0.08721704394141146,
      "grad_norm": 1.1493550672820814,
      "learning_rate": 2e-05,
      "loss": 0.515593433380127,
      "step": 1310
    },
    {
      "epoch": 0.08788282290279627,
      "grad_norm": 5.47637808249577,
      "learning_rate": 2e-05,
      "loss": 0.5221123695373535,
      "step": 1320
    },
    {
      "epoch": 0.08854860186418109,
      "grad_norm": 2.043477169717248,
      "learning_rate": 2e-05,
      "loss": 0.5497914314270019,
      "step": 1330
    },
    {
      "epoch": 0.08921438082556592,
      "grad_norm": 2.055175898286977,
      "learning_rate": 2e-05,
      "loss": 0.5801389694213868,
      "step": 1340
    },
    {
      "epoch": 0.08988015978695073,
      "grad_norm": 2.1668021346153115,
      "learning_rate": 2e-05,
      "loss": 0.517332124710083,
      "step": 1350
    },
    {
      "epoch": 0.09054593874833555,
      "grad_norm": 1.5460307704723255,
      "learning_rate": 2e-05,
      "loss": 0.5286785125732422,
      "step": 1360
    },
    {
      "epoch": 0.09121171770972038,
      "grad_norm": 7.339714501931053,
      "learning_rate": 2e-05,
      "loss": 0.5441049098968506,
      "step": 1370
    },
    {
      "epoch": 0.09187749667110519,
      "grad_norm": 2.9991980124421977,
      "learning_rate": 2e-05,
      "loss": 0.5296726226806641,
      "step": 1380
    },
    {
      "epoch": 0.09254327563249001,
      "grad_norm": 1.9704344709203743,
      "learning_rate": 2e-05,
      "loss": 0.5755397796630859,
      "step": 1390
    },
    {
      "epoch": 0.09320905459387484,
      "grad_norm": 3.1974312777990006,
      "learning_rate": 2e-05,
      "loss": 0.5350383758544922,
      "step": 1400
    },
    {
      "epoch": 0.09387483355525965,
      "grad_norm": 4.835254285595909,
      "learning_rate": 2e-05,
      "loss": 0.5583718299865723,
      "step": 1410
    },
    {
      "epoch": 0.09454061251664447,
      "grad_norm": 4.086077808375664,
      "learning_rate": 2e-05,
      "loss": 0.5181577682495118,
      "step": 1420
    },
    {
      "epoch": 0.0952063914780293,
      "grad_norm": 1.3712201211608543,
      "learning_rate": 2e-05,
      "loss": 0.5456793785095215,
      "step": 1430
    },
    {
      "epoch": 0.09587217043941411,
      "grad_norm": 2.3710137180040323,
      "learning_rate": 2e-05,
      "loss": 0.6178105354309082,
      "step": 1440
    },
    {
      "epoch": 0.09653794940079893,
      "grad_norm": 1.6819343647716947,
      "learning_rate": 2e-05,
      "loss": 0.5533603191375732,
      "step": 1450
    },
    {
      "epoch": 0.09720372836218376,
      "grad_norm": 1.2365582424751604,
      "learning_rate": 2e-05,
      "loss": 0.5364840984344482,
      "step": 1460
    },
    {
      "epoch": 0.09786950732356857,
      "grad_norm": 2.296453401394828,
      "learning_rate": 2e-05,
      "loss": 0.5314805030822753,
      "step": 1470
    },
    {
      "epoch": 0.0985352862849534,
      "grad_norm": 1.24283562111146,
      "learning_rate": 2e-05,
      "loss": 0.5186245918273926,
      "step": 1480
    },
    {
      "epoch": 0.09920106524633822,
      "grad_norm": 1.3611004524534043,
      "learning_rate": 2e-05,
      "loss": 0.524735975265503,
      "step": 1490
    },
    {
      "epoch": 0.09986684420772303,
      "grad_norm": 3.9363806116623707,
      "learning_rate": 2e-05,
      "loss": 0.550562858581543,
      "step": 1500
    },
    {
      "epoch": 0.10053262316910785,
      "grad_norm": 1.972796502107073,
      "learning_rate": 2e-05,
      "loss": 0.528165340423584,
      "step": 1510
    },
    {
      "epoch": 0.10119840213049268,
      "grad_norm": 1.1667369026127106,
      "learning_rate": 2e-05,
      "loss": 0.5359246730804443,
      "step": 1520
    },
    {
      "epoch": 0.10186418109187749,
      "grad_norm": 1.8610540585135436,
      "learning_rate": 2e-05,
      "loss": 0.52705078125,
      "step": 1530
    },
    {
      "epoch": 0.10252996005326231,
      "grad_norm": 1.324687718260285,
      "learning_rate": 2e-05,
      "loss": 0.5163713455200195,
      "step": 1540
    },
    {
      "epoch": 0.10319573901464714,
      "grad_norm": 2.232586255903306,
      "learning_rate": 2e-05,
      "loss": 0.5380361557006836,
      "step": 1550
    },
    {
      "epoch": 0.10386151797603196,
      "grad_norm": 2.8377026825388247,
      "learning_rate": 2e-05,
      "loss": 0.5271590232849122,
      "step": 1560
    },
    {
      "epoch": 0.10452729693741677,
      "grad_norm": 1.0232432789770685,
      "learning_rate": 2e-05,
      "loss": 0.5482112884521484,
      "step": 1570
    },
    {
      "epoch": 0.1051930758988016,
      "grad_norm": 3.0682327172834913,
      "learning_rate": 2e-05,
      "loss": 0.5437918663024902,
      "step": 1580
    },
    {
      "epoch": 0.10585885486018642,
      "grad_norm": 1.3849474396156232,
      "learning_rate": 2e-05,
      "loss": 0.49279470443725587,
      "step": 1590
    },
    {
      "epoch": 0.10652463382157124,
      "grad_norm": 2.0299726086559993,
      "learning_rate": 2e-05,
      "loss": 0.5641605377197265,
      "step": 1600
    },
    {
      "epoch": 0.10719041278295606,
      "grad_norm": 1.817550197200932,
      "learning_rate": 2e-05,
      "loss": 0.5258097648620605,
      "step": 1610
    },
    {
      "epoch": 0.10785619174434088,
      "grad_norm": 1.1651824761649607,
      "learning_rate": 2e-05,
      "loss": 0.5535667419433594,
      "step": 1620
    },
    {
      "epoch": 0.1085219707057257,
      "grad_norm": 1.5674773668466062,
      "learning_rate": 2e-05,
      "loss": 0.5567864894866943,
      "step": 1630
    },
    {
      "epoch": 0.10918774966711052,
      "grad_norm": 1.1591412095258105,
      "learning_rate": 2e-05,
      "loss": 0.5203337669372559,
      "step": 1640
    },
    {
      "epoch": 0.10985352862849534,
      "grad_norm": 1.1505534469163556,
      "learning_rate": 2e-05,
      "loss": 0.5513369560241699,
      "step": 1650
    },
    {
      "epoch": 0.11051930758988016,
      "grad_norm": 1.0506019078549658,
      "learning_rate": 2e-05,
      "loss": 0.5552059173583984,
      "step": 1660
    },
    {
      "epoch": 0.11118508655126498,
      "grad_norm": 1.2459158335355898,
      "learning_rate": 2e-05,
      "loss": 0.5355972290039063,
      "step": 1670
    },
    {
      "epoch": 0.1118508655126498,
      "grad_norm": 1.4295425078712645,
      "learning_rate": 2e-05,
      "loss": 0.54208984375,
      "step": 1680
    },
    {
      "epoch": 0.11251664447403462,
      "grad_norm": 1.005803323010581,
      "learning_rate": 2e-05,
      "loss": 0.5317649841308594,
      "step": 1690
    },
    {
      "epoch": 0.11318242343541944,
      "grad_norm": 1.2565701592686467,
      "learning_rate": 2e-05,
      "loss": 0.5282378196716309,
      "step": 1700
    },
    {
      "epoch": 0.11384820239680427,
      "grad_norm": 0.9550909473074799,
      "learning_rate": 2e-05,
      "loss": 0.5165553092956543,
      "step": 1710
    },
    {
      "epoch": 0.11451398135818908,
      "grad_norm": 1.3278097810505283,
      "learning_rate": 2e-05,
      "loss": 0.5109647750854492,
      "step": 1720
    },
    {
      "epoch": 0.1151797603195739,
      "grad_norm": 0.9808078433806386,
      "learning_rate": 2e-05,
      "loss": 0.5323241233825684,
      "step": 1730
    },
    {
      "epoch": 0.11584553928095873,
      "grad_norm": 1.0288764189211852,
      "learning_rate": 2e-05,
      "loss": 0.5097640037536622,
      "step": 1740
    },
    {
      "epoch": 0.11651131824234354,
      "grad_norm": 1.0455186608218316,
      "learning_rate": 2e-05,
      "loss": 0.5255751609802246,
      "step": 1750
    },
    {
      "epoch": 0.11717709720372836,
      "grad_norm": 0.9659796277901321,
      "learning_rate": 2e-05,
      "loss": 0.5367124080657959,
      "step": 1760
    },
    {
      "epoch": 0.11784287616511319,
      "grad_norm": 2.3356253200269204,
      "learning_rate": 2e-05,
      "loss": 0.5488951683044434,
      "step": 1770
    },
    {
      "epoch": 0.118508655126498,
      "grad_norm": 1.1896876383401835,
      "learning_rate": 2e-05,
      "loss": 0.5306941509246826,
      "step": 1780
    },
    {
      "epoch": 0.11917443408788282,
      "grad_norm": 1.2306037178037514,
      "learning_rate": 2e-05,
      "loss": 0.5404845237731933,
      "step": 1790
    },
    {
      "epoch": 0.11984021304926765,
      "grad_norm": 1.6449345630034806,
      "learning_rate": 2e-05,
      "loss": 0.5113681316375732,
      "step": 1800
    },
    {
      "epoch": 0.12050599201065246,
      "grad_norm": 13.948346139257183,
      "learning_rate": 2e-05,
      "loss": 0.5406979560852051,
      "step": 1810
    },
    {
      "epoch": 0.12117177097203728,
      "grad_norm": 0.8897567907562994,
      "learning_rate": 2e-05,
      "loss": 0.545926856994629,
      "step": 1820
    },
    {
      "epoch": 0.1218375499334221,
      "grad_norm": 1.2112299469668055,
      "learning_rate": 2e-05,
      "loss": 0.5686315059661865,
      "step": 1830
    },
    {
      "epoch": 0.12250332889480692,
      "grad_norm": 1.2807189268351256,
      "learning_rate": 2e-05,
      "loss": 0.53879714012146,
      "step": 1840
    },
    {
      "epoch": 0.12316910785619174,
      "grad_norm": 1.227450535129884,
      "learning_rate": 2e-05,
      "loss": 0.5089795112609863,
      "step": 1850
    },
    {
      "epoch": 0.12383488681757657,
      "grad_norm": 1.2772771188177927,
      "learning_rate": 2e-05,
      "loss": 0.5080512523651123,
      "step": 1860
    },
    {
      "epoch": 0.12450066577896138,
      "grad_norm": 0.9281113939706479,
      "learning_rate": 2e-05,
      "loss": 0.5448731899261474,
      "step": 1870
    },
    {
      "epoch": 0.12516644474034622,
      "grad_norm": 1.6719732819004125,
      "learning_rate": 2e-05,
      "loss": 0.5150301933288575,
      "step": 1880
    },
    {
      "epoch": 0.12583222370173103,
      "grad_norm": 1.133321175275307,
      "learning_rate": 2e-05,
      "loss": 0.5294013977050781,
      "step": 1890
    },
    {
      "epoch": 0.12649800266311584,
      "grad_norm": 1.8063664874074499,
      "learning_rate": 2e-05,
      "loss": 0.5121081352233887,
      "step": 1900
    },
    {
      "epoch": 0.12716378162450068,
      "grad_norm": 1.7628751698806506,
      "learning_rate": 2e-05,
      "loss": 0.5407076835632324,
      "step": 1910
    },
    {
      "epoch": 0.1278295605858855,
      "grad_norm": 1.7994852581180745,
      "learning_rate": 2e-05,
      "loss": 0.5252192974090576,
      "step": 1920
    },
    {
      "epoch": 0.1284953395472703,
      "grad_norm": 1.0915149354977984,
      "learning_rate": 2e-05,
      "loss": 0.525769853591919,
      "step": 1930
    },
    {
      "epoch": 0.12916111850865514,
      "grad_norm": 2.151838757501649,
      "learning_rate": 2e-05,
      "loss": 0.51082444190979,
      "step": 1940
    },
    {
      "epoch": 0.12982689747003995,
      "grad_norm": 1.1402975756842395,
      "learning_rate": 2e-05,
      "loss": 0.5432845115661621,
      "step": 1950
    },
    {
      "epoch": 0.13049267643142476,
      "grad_norm": 2.2552774036378183,
      "learning_rate": 2e-05,
      "loss": 0.5111236572265625,
      "step": 1960
    },
    {
      "epoch": 0.1311584553928096,
      "grad_norm": 2.3875590430821876,
      "learning_rate": 2e-05,
      "loss": 0.509479284286499,
      "step": 1970
    },
    {
      "epoch": 0.1318242343541944,
      "grad_norm": 1.1594488373781606,
      "learning_rate": 2e-05,
      "loss": 0.5075606822967529,
      "step": 1980
    },
    {
      "epoch": 0.13249001331557922,
      "grad_norm": 1.1364466542686973,
      "learning_rate": 2e-05,
      "loss": 0.5161935329437256,
      "step": 1990
    },
    {
      "epoch": 0.13315579227696406,
      "grad_norm": 33.5151639217467,
      "learning_rate": 2e-05,
      "loss": 0.5208932876586914,
      "step": 2000
    },
    {
      "epoch": 0.13382157123834887,
      "grad_norm": 3.718913798025406,
      "learning_rate": 2e-05,
      "loss": 0.533729362487793,
      "step": 2010
    },
    {
      "epoch": 0.13448735019973368,
      "grad_norm": 1.1708432227005658,
      "learning_rate": 2e-05,
      "loss": 0.4863370418548584,
      "step": 2020
    },
    {
      "epoch": 0.13515312916111852,
      "grad_norm": 1.5381854574243472,
      "learning_rate": 2e-05,
      "loss": 0.5670979499816895,
      "step": 2030
    },
    {
      "epoch": 0.13581890812250333,
      "grad_norm": 1.5413767152488844,
      "learning_rate": 2e-05,
      "loss": 0.5272467613220215,
      "step": 2040
    },
    {
      "epoch": 0.13648468708388814,
      "grad_norm": 2.4149932246072727,
      "learning_rate": 2e-05,
      "loss": 0.5259864330291748,
      "step": 2050
    },
    {
      "epoch": 0.13715046604527298,
      "grad_norm": 0.8576063037502878,
      "learning_rate": 2e-05,
      "loss": 0.5189313888549805,
      "step": 2060
    },
    {
      "epoch": 0.1378162450066578,
      "grad_norm": 2.0944398013918635,
      "learning_rate": 2e-05,
      "loss": 0.5383773803710937,
      "step": 2070
    },
    {
      "epoch": 0.1384820239680426,
      "grad_norm": 2.202299198114209,
      "learning_rate": 2e-05,
      "loss": 0.5349685668945312,
      "step": 2080
    },
    {
      "epoch": 0.13914780292942744,
      "grad_norm": 0.8804253733660655,
      "learning_rate": 2e-05,
      "loss": 0.5294415473937988,
      "step": 2090
    },
    {
      "epoch": 0.13981358189081225,
      "grad_norm": 1.1171759751806474,
      "learning_rate": 2e-05,
      "loss": 0.5218812942504882,
      "step": 2100
    },
    {
      "epoch": 0.14047936085219706,
      "grad_norm": 1.6634271061653607,
      "learning_rate": 2e-05,
      "loss": 0.5050858974456787,
      "step": 2110
    },
    {
      "epoch": 0.1411451398135819,
      "grad_norm": 1.1835409358711206,
      "learning_rate": 2e-05,
      "loss": 0.5161876678466797,
      "step": 2120
    },
    {
      "epoch": 0.1418109187749667,
      "grad_norm": 2.101176098222384,
      "learning_rate": 2e-05,
      "loss": 0.5150922775268555,
      "step": 2130
    },
    {
      "epoch": 0.14247669773635152,
      "grad_norm": 1.0781328275401703,
      "learning_rate": 2e-05,
      "loss": 0.5243738174438477,
      "step": 2140
    },
    {
      "epoch": 0.14314247669773636,
      "grad_norm": 2.6622883554564107,
      "learning_rate": 2e-05,
      "loss": 0.5393617630004883,
      "step": 2150
    },
    {
      "epoch": 0.14380825565912117,
      "grad_norm": 3.249905568049039,
      "learning_rate": 2e-05,
      "loss": 0.5379310131072998,
      "step": 2160
    },
    {
      "epoch": 0.14447403462050598,
      "grad_norm": 1.1442728323685893,
      "learning_rate": 2e-05,
      "loss": 0.5616377830505371,
      "step": 2170
    },
    {
      "epoch": 0.14513981358189082,
      "grad_norm": 1.126585913850443,
      "learning_rate": 2e-05,
      "loss": 0.49822397232055665,
      "step": 2180
    },
    {
      "epoch": 0.14580559254327563,
      "grad_norm": 1.3884430019967056,
      "learning_rate": 2e-05,
      "loss": 0.5449071407318116,
      "step": 2190
    },
    {
      "epoch": 0.14647137150466044,
      "grad_norm": 1.146409501415657,
      "learning_rate": 2e-05,
      "loss": 0.5238983154296875,
      "step": 2200
    },
    {
      "epoch": 0.14713715046604528,
      "grad_norm": 1.4908765281795415,
      "learning_rate": 2e-05,
      "loss": 0.5086498737335206,
      "step": 2210
    },
    {
      "epoch": 0.1478029294274301,
      "grad_norm": 1.4052241706860102,
      "learning_rate": 2e-05,
      "loss": 0.4959442138671875,
      "step": 2220
    },
    {
      "epoch": 0.1484687083888149,
      "grad_norm": 0.9232977271364534,
      "learning_rate": 2e-05,
      "loss": 0.49895877838134767,
      "step": 2230
    },
    {
      "epoch": 0.14913448735019974,
      "grad_norm": 1.251500253657333,
      "learning_rate": 2e-05,
      "loss": 0.5154998779296875,
      "step": 2240
    },
    {
      "epoch": 0.14980026631158455,
      "grad_norm": 0.9654701095863362,
      "learning_rate": 2e-05,
      "loss": 0.5073034286499023,
      "step": 2250
    },
    {
      "epoch": 0.15046604527296936,
      "grad_norm": 0.9584873628963919,
      "learning_rate": 2e-05,
      "loss": 0.5323206901550293,
      "step": 2260
    },
    {
      "epoch": 0.1511318242343542,
      "grad_norm": 1.115220481547035,
      "learning_rate": 2e-05,
      "loss": 0.5065470695495605,
      "step": 2270
    },
    {
      "epoch": 0.151797603195739,
      "grad_norm": 4.662097485396481,
      "learning_rate": 2e-05,
      "loss": 0.4943705081939697,
      "step": 2280
    },
    {
      "epoch": 0.15246338215712382,
      "grad_norm": 11.103524498016363,
      "learning_rate": 2e-05,
      "loss": 0.5245217323303223,
      "step": 2290
    },
    {
      "epoch": 0.15312916111850866,
      "grad_norm": 1.4598282653889072,
      "learning_rate": 2e-05,
      "loss": 0.5193574905395508,
      "step": 2300
    },
    {
      "epoch": 0.15379494007989347,
      "grad_norm": 1.7782516999908409,
      "learning_rate": 2e-05,
      "loss": 0.48776745796203613,
      "step": 2310
    },
    {
      "epoch": 0.15446071904127828,
      "grad_norm": 3.2706999654913154,
      "learning_rate": 2e-05,
      "loss": 0.5390383720397949,
      "step": 2320
    },
    {
      "epoch": 0.15512649800266312,
      "grad_norm": 45.974392184868876,
      "learning_rate": 2e-05,
      "loss": 0.5336198329925537,
      "step": 2330
    },
    {
      "epoch": 0.15579227696404793,
      "grad_norm": 2.7586767382582815,
      "learning_rate": 2e-05,
      "loss": 0.5713801383972168,
      "step": 2340
    },
    {
      "epoch": 0.15645805592543274,
      "grad_norm": 2.9092759262358094,
      "learning_rate": 2e-05,
      "loss": 0.5405255794525147,
      "step": 2350
    },
    {
      "epoch": 0.15712383488681758,
      "grad_norm": 2.1258359511475198,
      "learning_rate": 2e-05,
      "loss": 0.5386243343353272,
      "step": 2360
    },
    {
      "epoch": 0.1577896138482024,
      "grad_norm": 1.2381595039295767,
      "learning_rate": 2e-05,
      "loss": 0.5679190635681153,
      "step": 2370
    },
    {
      "epoch": 0.1584553928095872,
      "grad_norm": 17.18583576572219,
      "learning_rate": 2e-05,
      "loss": 0.5169041633605957,
      "step": 2380
    },
    {
      "epoch": 0.15912117177097204,
      "grad_norm": 5.3082457989157215,
      "learning_rate": 2e-05,
      "loss": 0.5436855316162109,
      "step": 2390
    },
    {
      "epoch": 0.15978695073235685,
      "grad_norm": 1.227956735412768,
      "learning_rate": 2e-05,
      "loss": 0.555968189239502,
      "step": 2400
    },
    {
      "epoch": 0.1604527296937417,
      "grad_norm": 1.4804230692740696,
      "learning_rate": 2e-05,
      "loss": 0.5193206787109375,
      "step": 2410
    },
    {
      "epoch": 0.1611185086551265,
      "grad_norm": 4.436062371833058,
      "learning_rate": 2e-05,
      "loss": 0.5349526882171631,
      "step": 2420
    },
    {
      "epoch": 0.1617842876165113,
      "grad_norm": 2.3524883548432056,
      "learning_rate": 2e-05,
      "loss": 0.5464076519012451,
      "step": 2430
    },
    {
      "epoch": 0.16245006657789615,
      "grad_norm": 2.4186439262098625,
      "learning_rate": 2e-05,
      "loss": 0.5067841053009033,
      "step": 2440
    },
    {
      "epoch": 0.16311584553928096,
      "grad_norm": 1.9343779365891003,
      "learning_rate": 2e-05,
      "loss": 0.5334929943084716,
      "step": 2450
    },
    {
      "epoch": 0.16378162450066577,
      "grad_norm": 1.1199215401132792,
      "learning_rate": 2e-05,
      "loss": 0.5157291412353515,
      "step": 2460
    },
    {
      "epoch": 0.1644474034620506,
      "grad_norm": 2.8193397360794696,
      "learning_rate": 2e-05,
      "loss": 0.5031748294830323,
      "step": 2470
    },
    {
      "epoch": 0.16511318242343542,
      "grad_norm": 1.5928757383633196,
      "learning_rate": 2e-05,
      "loss": 0.5512802124023437,
      "step": 2480
    },
    {
      "epoch": 0.16577896138482023,
      "grad_norm": 1.9522306023391771,
      "learning_rate": 2e-05,
      "loss": 0.547333812713623,
      "step": 2490
    },
    {
      "epoch": 0.16644474034620507,
      "grad_norm": 2.172031447099043,
      "learning_rate": 2e-05,
      "loss": 0.5480151176452637,
      "step": 2500
    },
    {
      "epoch": 0.16711051930758988,
      "grad_norm": 1.0415649470614736,
      "learning_rate": 2e-05,
      "loss": 0.5497314453125,
      "step": 2510
    },
    {
      "epoch": 0.1677762982689747,
      "grad_norm": 1.2241977445131844,
      "learning_rate": 2e-05,
      "loss": 0.5218514919281005,
      "step": 2520
    },
    {
      "epoch": 0.16844207723035953,
      "grad_norm": 1.46895120578611,
      "learning_rate": 2e-05,
      "loss": 0.5450241088867187,
      "step": 2530
    },
    {
      "epoch": 0.16910785619174434,
      "grad_norm": 1.2668093618694478,
      "learning_rate": 2e-05,
      "loss": 0.5406510353088378,
      "step": 2540
    },
    {
      "epoch": 0.16977363515312915,
      "grad_norm": 1.5516773761861413,
      "learning_rate": 2e-05,
      "loss": 0.5550095694405692,
      "step": 2550
    },
    {
      "epoch": 0.170439414114514,
      "grad_norm": 1.6338388499883492,
      "learning_rate": 2e-05,
      "loss": 0.48791279792785647,
      "step": 2560
    },
    {
      "epoch": 0.1711051930758988,
      "grad_norm": 1.3168457071634436,
      "learning_rate": 2e-05,
      "loss": 0.5387988090515137,
      "step": 2570
    },
    {
      "epoch": 0.17177097203728361,
      "grad_norm": 1.135511982994294,
      "learning_rate": 2e-05,
      "loss": 0.49612579345703123,
      "step": 2580
    },
    {
      "epoch": 0.17243675099866845,
      "grad_norm": 1.6232625191093284,
      "learning_rate": 2e-05,
      "loss": 0.524751091003418,
      "step": 2590
    },
    {
      "epoch": 0.17310252996005326,
      "grad_norm": 6.817203169230776,
      "learning_rate": 2e-05,
      "loss": 0.5172528266906739,
      "step": 2600
    },
    {
      "epoch": 0.17376830892143807,
      "grad_norm": 1.2517730043085542,
      "learning_rate": 2e-05,
      "loss": 0.531795883178711,
      "step": 2610
    },
    {
      "epoch": 0.1744340878828229,
      "grad_norm": 0.9940009074523652,
      "learning_rate": 2e-05,
      "loss": 0.5489285945892334,
      "step": 2620
    },
    {
      "epoch": 0.17509986684420772,
      "grad_norm": 1.2664514784905307,
      "learning_rate": 2e-05,
      "loss": 0.5306877136230469,
      "step": 2630
    },
    {
      "epoch": 0.17576564580559254,
      "grad_norm": 2.1567398254309413,
      "learning_rate": 2e-05,
      "loss": 0.4839469432830811,
      "step": 2640
    },
    {
      "epoch": 0.17643142476697737,
      "grad_norm": 1.89754947983118,
      "learning_rate": 2e-05,
      "loss": 0.5594792842864991,
      "step": 2650
    },
    {
      "epoch": 0.17709720372836218,
      "grad_norm": 2.061818945173113,
      "learning_rate": 2e-05,
      "loss": 0.5012704849243164,
      "step": 2660
    },
    {
      "epoch": 0.177762982689747,
      "grad_norm": 1.4158798457286839,
      "learning_rate": 2e-05,
      "loss": 0.48920383453369143,
      "step": 2670
    },
    {
      "epoch": 0.17842876165113183,
      "grad_norm": 2.8354559213733035,
      "learning_rate": 2e-05,
      "loss": 0.5250689506530761,
      "step": 2680
    },
    {
      "epoch": 0.17909454061251665,
      "grad_norm": 1.2031026360019217,
      "learning_rate": 2e-05,
      "loss": 0.5020029544830322,
      "step": 2690
    },
    {
      "epoch": 0.17976031957390146,
      "grad_norm": 4.189125613372769,
      "learning_rate": 2e-05,
      "loss": 0.5361767768859863,
      "step": 2700
    },
    {
      "epoch": 0.1804260985352863,
      "grad_norm": 1.9153682219774628,
      "learning_rate": 2e-05,
      "loss": 0.5162026882171631,
      "step": 2710
    },
    {
      "epoch": 0.1810918774966711,
      "grad_norm": 10.386726102309732,
      "learning_rate": 2e-05,
      "loss": 0.5353858470916748,
      "step": 2720
    },
    {
      "epoch": 0.18175765645805592,
      "grad_norm": 0.7568090688854552,
      "learning_rate": 2e-05,
      "loss": 0.5098324775695801,
      "step": 2730
    },
    {
      "epoch": 0.18242343541944075,
      "grad_norm": 1.1070324816026587,
      "learning_rate": 2e-05,
      "loss": 0.4892563819885254,
      "step": 2740
    },
    {
      "epoch": 0.18308921438082557,
      "grad_norm": 1.2144582057092466,
      "learning_rate": 2e-05,
      "loss": 0.5089690208435058,
      "step": 2750
    },
    {
      "epoch": 0.18375499334221038,
      "grad_norm": 0.7897294228899566,
      "learning_rate": 2e-05,
      "loss": 0.5053329944610596,
      "step": 2760
    },
    {
      "epoch": 0.18442077230359522,
      "grad_norm": 0.9118833896549625,
      "learning_rate": 2e-05,
      "loss": 0.5430496215820313,
      "step": 2770
    },
    {
      "epoch": 0.18508655126498003,
      "grad_norm": 0.7860101270620378,
      "learning_rate": 2e-05,
      "loss": 0.4970438003540039,
      "step": 2780
    },
    {
      "epoch": 0.18575233022636484,
      "grad_norm": 1.1781917504023258,
      "learning_rate": 2e-05,
      "loss": 0.51654052734375,
      "step": 2790
    },
    {
      "epoch": 0.18641810918774968,
      "grad_norm": 4.121689618158715,
      "learning_rate": 2e-05,
      "loss": 0.48903255462646483,
      "step": 2800
    },
    {
      "epoch": 0.1870838881491345,
      "grad_norm": 1.2970443183023916,
      "learning_rate": 2e-05,
      "loss": 0.5320328712463379,
      "step": 2810
    },
    {
      "epoch": 0.1877496671105193,
      "grad_norm": 0.8586607287337011,
      "learning_rate": 2e-05,
      "loss": 0.5191061973571778,
      "step": 2820
    },
    {
      "epoch": 0.18841544607190414,
      "grad_norm": 2.8985801095464225,
      "learning_rate": 2e-05,
      "loss": 0.5013152122497558,
      "step": 2830
    },
    {
      "epoch": 0.18908122503328895,
      "grad_norm": 21.51395609278927,
      "learning_rate": 2e-05,
      "loss": 0.5246681213378906,
      "step": 2840
    },
    {
      "epoch": 0.18974700399467376,
      "grad_norm": 2.4165860458208006,
      "learning_rate": 2e-05,
      "loss": 0.47137789726257323,
      "step": 2850
    },
    {
      "epoch": 0.1904127829560586,
      "grad_norm": 4.844783021166387,
      "learning_rate": 2e-05,
      "loss": 0.49054689407348634,
      "step": 2860
    },
    {
      "epoch": 0.1910785619174434,
      "grad_norm": 2.405356829367001,
      "learning_rate": 2e-05,
      "loss": 0.5773238182067871,
      "step": 2870
    },
    {
      "epoch": 0.19174434087882822,
      "grad_norm": 1.7990765839687963,
      "learning_rate": 2e-05,
      "loss": 0.505347204208374,
      "step": 2880
    },
    {
      "epoch": 0.19241011984021306,
      "grad_norm": 1.2044957779548762,
      "learning_rate": 2e-05,
      "loss": 0.5127789497375488,
      "step": 2890
    },
    {
      "epoch": 0.19307589880159787,
      "grad_norm": 1.063490565543371,
      "learning_rate": 2e-05,
      "loss": 0.5302878379821777,
      "step": 2900
    },
    {
      "epoch": 0.19374167776298268,
      "grad_norm": 3.700073608303432,
      "learning_rate": 2e-05,
      "loss": 0.5352324485778809,
      "step": 2910
    },
    {
      "epoch": 0.19440745672436752,
      "grad_norm": 8.590259035339203,
      "learning_rate": 2e-05,
      "loss": 0.4822662353515625,
      "step": 2920
    },
    {
      "epoch": 0.19507323568575233,
      "grad_norm": 2.4879825281591836,
      "learning_rate": 2e-05,
      "loss": 0.5392212867736816,
      "step": 2930
    },
    {
      "epoch": 0.19573901464713714,
      "grad_norm": 1.3373370366078645,
      "learning_rate": 2e-05,
      "loss": 0.5486340999603272,
      "step": 2940
    },
    {
      "epoch": 0.19640479360852198,
      "grad_norm": 5.556279108086433,
      "learning_rate": 2e-05,
      "loss": 0.4851529121398926,
      "step": 2950
    },
    {
      "epoch": 0.1970705725699068,
      "grad_norm": 1.068529976987903,
      "learning_rate": 2e-05,
      "loss": 0.5052083969116211,
      "step": 2960
    },
    {
      "epoch": 0.1977363515312916,
      "grad_norm": 1.3753811052907894,
      "learning_rate": 2e-05,
      "loss": 0.517616605758667,
      "step": 2970
    },
    {
      "epoch": 0.19840213049267644,
      "grad_norm": 2.996220415550952,
      "learning_rate": 2e-05,
      "loss": 0.5059660911560059,
      "step": 2980
    },
    {
      "epoch": 0.19906790945406125,
      "grad_norm": 2.6449892829111463,
      "learning_rate": 2e-05,
      "loss": 0.5084340572357178,
      "step": 2990
    },
    {
      "epoch": 0.19973368841544606,
      "grad_norm": 0.9882899156507222,
      "learning_rate": 2e-05,
      "loss": 0.49653186798095705,
      "step": 3000
    },
    {
      "epoch": 0.2003994673768309,
      "grad_norm": 1.7161469338073023,
      "learning_rate": 2e-05,
      "loss": 0.5185492515563965,
      "step": 3010
    },
    {
      "epoch": 0.2010652463382157,
      "grad_norm": 0.9041592586805045,
      "learning_rate": 2e-05,
      "loss": 0.5098721981048584,
      "step": 3020
    },
    {
      "epoch": 0.20173102529960052,
      "grad_norm": 0.9783353566281064,
      "learning_rate": 2e-05,
      "loss": 0.517940616607666,
      "step": 3030
    },
    {
      "epoch": 0.20239680426098536,
      "grad_norm": 0.9653780061443151,
      "learning_rate": 2e-05,
      "loss": 0.511451244354248,
      "step": 3040
    },
    {
      "epoch": 0.20306258322237017,
      "grad_norm": 0.9668779676489377,
      "learning_rate": 2e-05,
      "loss": 0.5162418365478516,
      "step": 3050
    },
    {
      "epoch": 0.20372836218375498,
      "grad_norm": 0.8707336125028304,
      "learning_rate": 2e-05,
      "loss": 0.4893776893615723,
      "step": 3060
    },
    {
      "epoch": 0.20439414114513982,
      "grad_norm": 1.3065293499742094,
      "learning_rate": 2e-05,
      "loss": 0.5248083591461181,
      "step": 3070
    },
    {
      "epoch": 0.20505992010652463,
      "grad_norm": 0.7943073650335887,
      "learning_rate": 2e-05,
      "loss": 0.5420484066009521,
      "step": 3080
    },
    {
      "epoch": 0.20572569906790947,
      "grad_norm": 1.3026857378391827,
      "learning_rate": 2e-05,
      "loss": 0.5097403049468994,
      "step": 3090
    },
    {
      "epoch": 0.20639147802929428,
      "grad_norm": 1.03444621102349,
      "learning_rate": 2e-05,
      "loss": 0.5248050689697266,
      "step": 3100
    },
    {
      "epoch": 0.2070572569906791,
      "grad_norm": 1.2449936959717085,
      "learning_rate": 2e-05,
      "loss": 0.4905829906463623,
      "step": 3110
    },
    {
      "epoch": 0.20772303595206393,
      "grad_norm": 0.8608665495034602,
      "learning_rate": 2e-05,
      "loss": 0.5451303482055664,
      "step": 3120
    },
    {
      "epoch": 0.20838881491344874,
      "grad_norm": 1.010354384246961,
      "learning_rate": 2e-05,
      "loss": 0.5173676490783692,
      "step": 3130
    },
    {
      "epoch": 0.20905459387483355,
      "grad_norm": 0.7896367681665551,
      "learning_rate": 2e-05,
      "loss": 0.5063128471374512,
      "step": 3140
    },
    {
      "epoch": 0.2097203728362184,
      "grad_norm": 0.9715035025543058,
      "learning_rate": 2e-05,
      "loss": 0.5053380012512207,
      "step": 3150
    },
    {
      "epoch": 0.2103861517976032,
      "grad_norm": 0.9431339490574885,
      "learning_rate": 2e-05,
      "loss": 0.4823415279388428,
      "step": 3160
    },
    {
      "epoch": 0.211051930758988,
      "grad_norm": 0.8167532613672482,
      "learning_rate": 2e-05,
      "loss": 0.48851833343505857,
      "step": 3170
    },
    {
      "epoch": 0.21171770972037285,
      "grad_norm": 2.4114847981103806,
      "learning_rate": 2e-05,
      "loss": 0.5435774803161622,
      "step": 3180
    },
    {
      "epoch": 0.21238348868175766,
      "grad_norm": 1.2140393465557255,
      "learning_rate": 2e-05,
      "loss": 0.5360036373138428,
      "step": 3190
    },
    {
      "epoch": 0.21304926764314247,
      "grad_norm": 1.0407004570983678,
      "learning_rate": 2e-05,
      "loss": 0.5048234939575196,
      "step": 3200
    },
    {
      "epoch": 0.2137150466045273,
      "grad_norm": 0.863933940833467,
      "learning_rate": 2e-05,
      "loss": 0.535888671875,
      "step": 3210
    },
    {
      "epoch": 0.21438082556591212,
      "grad_norm": 1.2058245379505919,
      "learning_rate": 2e-05,
      "loss": 0.5083333015441894,
      "step": 3220
    },
    {
      "epoch": 0.21504660452729693,
      "grad_norm": 1.3664091585305587,
      "learning_rate": 2e-05,
      "loss": 0.5413249492645263,
      "step": 3230
    },
    {
      "epoch": 0.21571238348868177,
      "grad_norm": 0.6962904207652797,
      "learning_rate": 2e-05,
      "loss": 0.4808966636657715,
      "step": 3240
    },
    {
      "epoch": 0.21637816245006658,
      "grad_norm": 0.740128221358628,
      "learning_rate": 2e-05,
      "loss": 0.5061465263366699,
      "step": 3250
    },
    {
      "epoch": 0.2170439414114514,
      "grad_norm": 0.8419040189705957,
      "learning_rate": 2e-05,
      "loss": 0.5213862895965576,
      "step": 3260
    },
    {
      "epoch": 0.21770972037283623,
      "grad_norm": 0.7510835253307209,
      "learning_rate": 2e-05,
      "loss": 0.5062170028686523,
      "step": 3270
    },
    {
      "epoch": 0.21837549933422104,
      "grad_norm": 0.7304388693513041,
      "learning_rate": 2e-05,
      "loss": 0.4619168281555176,
      "step": 3280
    },
    {
      "epoch": 0.21904127829560585,
      "grad_norm": 1.243281900475937,
      "learning_rate": 2e-05,
      "loss": 0.5489372253417969,
      "step": 3290
    },
    {
      "epoch": 0.2197070572569907,
      "grad_norm": 0.7459328822238472,
      "learning_rate": 2e-05,
      "loss": 0.5225613594055176,
      "step": 3300
    },
    {
      "epoch": 0.2203728362183755,
      "grad_norm": 1.3082922679684965,
      "learning_rate": 2e-05,
      "loss": 0.5122794151306153,
      "step": 3310
    },
    {
      "epoch": 0.2210386151797603,
      "grad_norm": 3.2464211942631294,
      "learning_rate": 2e-05,
      "loss": 0.5047653198242188,
      "step": 3320
    },
    {
      "epoch": 0.22170439414114515,
      "grad_norm": 1.8841051765690815,
      "learning_rate": 2e-05,
      "loss": 0.5151290893554688,
      "step": 3330
    },
    {
      "epoch": 0.22237017310252996,
      "grad_norm": 1.6308524594443232,
      "learning_rate": 2e-05,
      "loss": 0.5321723937988281,
      "step": 3340
    },
    {
      "epoch": 0.22303595206391477,
      "grad_norm": 2.537879937622207,
      "learning_rate": 2e-05,
      "loss": 0.508517074584961,
      "step": 3350
    },
    {
      "epoch": 0.2237017310252996,
      "grad_norm": 1.562704915907372,
      "learning_rate": 2e-05,
      "loss": 0.5185061454772949,
      "step": 3360
    },
    {
      "epoch": 0.22436750998668442,
      "grad_norm": 1.0423447657477265,
      "learning_rate": 2e-05,
      "loss": 0.4994522571563721,
      "step": 3370
    },
    {
      "epoch": 0.22503328894806923,
      "grad_norm": 1.7212869298621174,
      "learning_rate": 2e-05,
      "loss": 0.5194188117980957,
      "step": 3380
    },
    {
      "epoch": 0.22569906790945407,
      "grad_norm": 1.6005087546982406,
      "learning_rate": 2e-05,
      "loss": 0.5205344200134278,
      "step": 3390
    },
    {
      "epoch": 0.22636484687083888,
      "grad_norm": 4.077302759588426,
      "learning_rate": 2e-05,
      "loss": 0.5079793930053711,
      "step": 3400
    },
    {
      "epoch": 0.2270306258322237,
      "grad_norm": 1.4912606628966467,
      "learning_rate": 2e-05,
      "loss": 0.5286508560180664,
      "step": 3410
    },
    {
      "epoch": 0.22769640479360853,
      "grad_norm": 1.0678516412840953,
      "learning_rate": 2e-05,
      "loss": 0.521702766418457,
      "step": 3420
    },
    {
      "epoch": 0.22836218375499334,
      "grad_norm": 39.28302184225686,
      "learning_rate": 2e-05,
      "loss": 0.5519749641418457,
      "step": 3430
    },
    {
      "epoch": 0.22902796271637815,
      "grad_norm": 2.1882514647003606,
      "learning_rate": 2e-05,
      "loss": 0.552126407623291,
      "step": 3440
    },
    {
      "epoch": 0.229693741677763,
      "grad_norm": 1.1609148379054735,
      "learning_rate": 2e-05,
      "loss": 0.5376743316650391,
      "step": 3450
    },
    {
      "epoch": 0.2303595206391478,
      "grad_norm": 1.7774911076235682,
      "learning_rate": 2e-05,
      "loss": 0.48469066619873047,
      "step": 3460
    },
    {
      "epoch": 0.2310252996005326,
      "grad_norm": 1.2287032049079654,
      "learning_rate": 2e-05,
      "loss": 0.5088033199310302,
      "step": 3470
    },
    {
      "epoch": 0.23169107856191745,
      "grad_norm": 1.4533443303562856,
      "learning_rate": 2e-05,
      "loss": 0.5274369239807128,
      "step": 3480
    },
    {
      "epoch": 0.23235685752330226,
      "grad_norm": 0.8066672307480074,
      "learning_rate": 2e-05,
      "loss": 0.511710262298584,
      "step": 3490
    },
    {
      "epoch": 0.23302263648468707,
      "grad_norm": 1.2334305624936501,
      "learning_rate": 2e-05,
      "loss": 0.5370091915130615,
      "step": 3500
    },
    {
      "epoch": 0.2336884154460719,
      "grad_norm": 0.8122647621032251,
      "learning_rate": 2e-05,
      "loss": 0.4930243492126465,
      "step": 3510
    },
    {
      "epoch": 0.23435419440745672,
      "grad_norm": 1.0293597130346235,
      "learning_rate": 2e-05,
      "loss": 0.5016162395477295,
      "step": 3520
    },
    {
      "epoch": 0.23501997336884153,
      "grad_norm": 2.388665934276866,
      "learning_rate": 2e-05,
      "loss": 0.4976178169250488,
      "step": 3530
    },
    {
      "epoch": 0.23568575233022637,
      "grad_norm": 1.3775743686938855,
      "learning_rate": 2e-05,
      "loss": 0.5403234481811523,
      "step": 3540
    },
    {
      "epoch": 0.23635153129161118,
      "grad_norm": 1.6577020500755977,
      "learning_rate": 2e-05,
      "loss": 0.5317672729492188,
      "step": 3550
    },
    {
      "epoch": 0.237017310252996,
      "grad_norm": 2.3908115540897628,
      "learning_rate": 2e-05,
      "loss": 0.4974231719970703,
      "step": 3560
    },
    {
      "epoch": 0.23768308921438083,
      "grad_norm": 1.031045357626065,
      "learning_rate": 2e-05,
      "loss": 0.498930549621582,
      "step": 3570
    },
    {
      "epoch": 0.23834886817576564,
      "grad_norm": 0.8395074217298867,
      "learning_rate": 2e-05,
      "loss": 0.4856985092163086,
      "step": 3580
    },
    {
      "epoch": 0.23901464713715045,
      "grad_norm": 1.0230205289625802,
      "learning_rate": 2e-05,
      "loss": 0.5000131607055665,
      "step": 3590
    },
    {
      "epoch": 0.2396804260985353,
      "grad_norm": 4.053357613949253,
      "learning_rate": 2e-05,
      "loss": 0.5177973747253418,
      "step": 3600
    },
    {
      "epoch": 0.2403462050599201,
      "grad_norm": 0.8832927400741434,
      "learning_rate": 2e-05,
      "loss": 0.5114580154418945,
      "step": 3610
    },
    {
      "epoch": 0.24101198402130491,
      "grad_norm": 3.7843579891587065,
      "learning_rate": 2e-05,
      "loss": 0.5087918281555176,
      "step": 3620
    },
    {
      "epoch": 0.24167776298268975,
      "grad_norm": 1.3304023244169265,
      "learning_rate": 2e-05,
      "loss": 0.5211354255676269,
      "step": 3630
    },
    {
      "epoch": 0.24234354194407456,
      "grad_norm": 0.7865724563550792,
      "learning_rate": 2e-05,
      "loss": 0.5050969600677491,
      "step": 3640
    },
    {
      "epoch": 0.24300932090545938,
      "grad_norm": 2.141400484408736,
      "learning_rate": 2e-05,
      "loss": 0.4845928192138672,
      "step": 3650
    },
    {
      "epoch": 0.2436750998668442,
      "grad_norm": 1.0184615796409235,
      "learning_rate": 2e-05,
      "loss": 0.5146331310272216,
      "step": 3660
    },
    {
      "epoch": 0.24434087882822902,
      "grad_norm": 0.7709479525492031,
      "learning_rate": 2e-05,
      "loss": 0.5078842163085937,
      "step": 3670
    },
    {
      "epoch": 0.24500665778961384,
      "grad_norm": 1.4548121742794884,
      "learning_rate": 2e-05,
      "loss": 0.5010724544525147,
      "step": 3680
    },
    {
      "epoch": 0.24567243675099867,
      "grad_norm": 0.8876189778676887,
      "learning_rate": 2e-05,
      "loss": 0.518458080291748,
      "step": 3690
    },
    {
      "epoch": 0.24633821571238348,
      "grad_norm": 0.9281219280256224,
      "learning_rate": 2e-05,
      "loss": 0.537723445892334,
      "step": 3700
    },
    {
      "epoch": 0.2470039946737683,
      "grad_norm": 2.7037001062871298,
      "learning_rate": 2e-05,
      "loss": 0.5055455684661865,
      "step": 3710
    },
    {
      "epoch": 0.24766977363515313,
      "grad_norm": 1.1363622276928462,
      "learning_rate": 2e-05,
      "loss": 0.496932315826416,
      "step": 3720
    },
    {
      "epoch": 0.24833555259653795,
      "grad_norm": 1.1127196215613906,
      "learning_rate": 2e-05,
      "loss": 0.5194563865661621,
      "step": 3730
    },
    {
      "epoch": 0.24900133155792276,
      "grad_norm": 2.0844253751189323,
      "learning_rate": 2e-05,
      "loss": 0.4858367919921875,
      "step": 3740
    },
    {
      "epoch": 0.2496671105193076,
      "grad_norm": 2.0006263935914723,
      "learning_rate": 2e-05,
      "loss": 0.5048156261444092,
      "step": 3750
    },
    {
      "epoch": 0.25033288948069243,
      "grad_norm": 1.1675911612161902,
      "learning_rate": 2e-05,
      "loss": 0.5153700828552246,
      "step": 3760
    },
    {
      "epoch": 0.25099866844207724,
      "grad_norm": 0.9486803213046088,
      "learning_rate": 2e-05,
      "loss": 0.48859167098999023,
      "step": 3770
    },
    {
      "epoch": 0.25166444740346205,
      "grad_norm": 1.3881243224140245,
      "learning_rate": 2e-05,
      "loss": 0.524126386642456,
      "step": 3780
    },
    {
      "epoch": 0.25233022636484687,
      "grad_norm": 1.0486330683655587,
      "learning_rate": 2e-05,
      "loss": 0.5430418014526367,
      "step": 3790
    },
    {
      "epoch": 0.2529960053262317,
      "grad_norm": 11.56118736609529,
      "learning_rate": 2e-05,
      "loss": 0.5077038764953613,
      "step": 3800
    },
    {
      "epoch": 0.2536617842876165,
      "grad_norm": 0.9741627842709685,
      "learning_rate": 2e-05,
      "loss": 0.5419612884521484,
      "step": 3810
    },
    {
      "epoch": 0.25432756324900135,
      "grad_norm": 0.9290843492148043,
      "learning_rate": 2e-05,
      "loss": 0.49710342288017273,
      "step": 3820
    },
    {
      "epoch": 0.25499334221038616,
      "grad_norm": 2.6308822411738007,
      "learning_rate": 2e-05,
      "loss": 0.5124654769897461,
      "step": 3830
    },
    {
      "epoch": 0.255659121171771,
      "grad_norm": 1.1185091131663327,
      "learning_rate": 2e-05,
      "loss": 0.5031078338623047,
      "step": 3840
    },
    {
      "epoch": 0.2563249001331558,
      "grad_norm": 1.3690006279190134,
      "learning_rate": 2e-05,
      "loss": 0.5067041397094727,
      "step": 3850
    },
    {
      "epoch": 0.2569906790945406,
      "grad_norm": 0.799484780042948,
      "learning_rate": 2e-05,
      "loss": 0.5096898078918457,
      "step": 3860
    },
    {
      "epoch": 0.2576564580559254,
      "grad_norm": 0.7258939087757793,
      "learning_rate": 2e-05,
      "loss": 0.4661886215209961,
      "step": 3870
    },
    {
      "epoch": 0.2583222370173103,
      "grad_norm": 1.61369770480611,
      "learning_rate": 2e-05,
      "loss": 0.5107232093811035,
      "step": 3880
    },
    {
      "epoch": 0.2589880159786951,
      "grad_norm": 1.0039177676313962,
      "learning_rate": 2e-05,
      "loss": 0.5029183864593506,
      "step": 3890
    },
    {
      "epoch": 0.2596537949400799,
      "grad_norm": 0.8160001268448994,
      "learning_rate": 2e-05,
      "loss": 0.48589162826538085,
      "step": 3900
    },
    {
      "epoch": 0.2603195739014647,
      "grad_norm": 1.4046833072762457,
      "learning_rate": 2e-05,
      "loss": 0.5333718299865723,
      "step": 3910
    },
    {
      "epoch": 0.2609853528628495,
      "grad_norm": 0.9965069420996371,
      "learning_rate": 2e-05,
      "loss": 0.5455557823181152,
      "step": 3920
    },
    {
      "epoch": 0.26165113182423433,
      "grad_norm": 1.0250597199067253,
      "learning_rate": 2e-05,
      "loss": 0.5289182662963867,
      "step": 3930
    },
    {
      "epoch": 0.2623169107856192,
      "grad_norm": 0.9677078897417387,
      "learning_rate": 2e-05,
      "loss": 0.493173885345459,
      "step": 3940
    },
    {
      "epoch": 0.262982689747004,
      "grad_norm": 0.9621641204139285,
      "learning_rate": 2e-05,
      "loss": 0.5003304004669189,
      "step": 3950
    },
    {
      "epoch": 0.2636484687083888,
      "grad_norm": 1.1353405900331912,
      "learning_rate": 2e-05,
      "loss": 0.5109395027160645,
      "step": 3960
    },
    {
      "epoch": 0.2643142476697736,
      "grad_norm": 0.805433929339191,
      "learning_rate": 2e-05,
      "loss": 0.531881856918335,
      "step": 3970
    },
    {
      "epoch": 0.26498002663115844,
      "grad_norm": 0.8096583140865243,
      "learning_rate": 2e-05,
      "loss": 0.4772097110748291,
      "step": 3980
    },
    {
      "epoch": 0.26564580559254325,
      "grad_norm": 1.1736570304975338,
      "learning_rate": 2e-05,
      "loss": 0.5015225410461426,
      "step": 3990
    },
    {
      "epoch": 0.2663115845539281,
      "grad_norm": 0.9645373670602955,
      "learning_rate": 2e-05,
      "loss": 0.48263139724731446,
      "step": 4000
    },
    {
      "epoch": 0.2669773635153129,
      "grad_norm": 0.7497348059956147,
      "learning_rate": 2e-05,
      "loss": 0.5048856735229492,
      "step": 4010
    },
    {
      "epoch": 0.26764314247669774,
      "grad_norm": 0.7570627707881178,
      "learning_rate": 2e-05,
      "loss": 0.4915769577026367,
      "step": 4020
    },
    {
      "epoch": 0.26830892143808255,
      "grad_norm": 2.877033197727491,
      "learning_rate": 2e-05,
      "loss": 0.5209522724151612,
      "step": 4030
    },
    {
      "epoch": 0.26897470039946736,
      "grad_norm": 1.0046283386539243,
      "learning_rate": 2e-05,
      "loss": 0.4895641326904297,
      "step": 4040
    },
    {
      "epoch": 0.26964047936085217,
      "grad_norm": 0.7603710351929752,
      "learning_rate": 2e-05,
      "loss": 0.4662502288818359,
      "step": 4050
    },
    {
      "epoch": 0.27030625832223704,
      "grad_norm": 1.1310404665097906,
      "learning_rate": 2e-05,
      "loss": 0.5316073894500732,
      "step": 4060
    },
    {
      "epoch": 0.27097203728362185,
      "grad_norm": 1.0486738635794561,
      "learning_rate": 2e-05,
      "loss": 0.5163720607757568,
      "step": 4070
    },
    {
      "epoch": 0.27163781624500666,
      "grad_norm": 0.7108975856397438,
      "learning_rate": 2e-05,
      "loss": 0.48084630966186526,
      "step": 4080
    },
    {
      "epoch": 0.27230359520639147,
      "grad_norm": 0.864796162660025,
      "learning_rate": 2e-05,
      "loss": 0.49396839141845705,
      "step": 4090
    },
    {
      "epoch": 0.2729693741677763,
      "grad_norm": 0.8208284050024944,
      "learning_rate": 2e-05,
      "loss": 0.5256934642791748,
      "step": 4100
    },
    {
      "epoch": 0.27363515312916115,
      "grad_norm": 2.630467853288637,
      "learning_rate": 2e-05,
      "loss": 0.5484153747558593,
      "step": 4110
    },
    {
      "epoch": 0.27430093209054596,
      "grad_norm": 1.0056742151613567,
      "learning_rate": 2e-05,
      "loss": 0.48935537338256835,
      "step": 4120
    },
    {
      "epoch": 0.27496671105193077,
      "grad_norm": 1.5002976887819175,
      "learning_rate": 2e-05,
      "loss": 0.5222125053405762,
      "step": 4130
    },
    {
      "epoch": 0.2756324900133156,
      "grad_norm": 1.0133928246077137,
      "learning_rate": 2e-05,
      "loss": 0.4955153465270996,
      "step": 4140
    },
    {
      "epoch": 0.2762982689747004,
      "grad_norm": 0.6960742778515082,
      "learning_rate": 2e-05,
      "loss": 0.5115083694458008,
      "step": 4150
    },
    {
      "epoch": 0.2769640479360852,
      "grad_norm": 0.7783860551179537,
      "learning_rate": 2e-05,
      "loss": 0.5342399597167968,
      "step": 4160
    },
    {
      "epoch": 0.27762982689747007,
      "grad_norm": 1.974116263465,
      "learning_rate": 2e-05,
      "loss": 0.4964879035949707,
      "step": 4170
    },
    {
      "epoch": 0.2782956058588549,
      "grad_norm": 0.6937371616954415,
      "learning_rate": 2e-05,
      "loss": 0.5123511791229248,
      "step": 4180
    },
    {
      "epoch": 0.2789613848202397,
      "grad_norm": 0.9217533215837527,
      "learning_rate": 2e-05,
      "loss": 0.48314218521118163,
      "step": 4190
    },
    {
      "epoch": 0.2796271637816245,
      "grad_norm": 0.9458648845255355,
      "learning_rate": 2e-05,
      "loss": 0.4914853096008301,
      "step": 4200
    },
    {
      "epoch": 0.2802929427430093,
      "grad_norm": 1.046716472300594,
      "learning_rate": 2e-05,
      "loss": 0.5107494354248047,
      "step": 4210
    },
    {
      "epoch": 0.2809587217043941,
      "grad_norm": 0.844331303877975,
      "learning_rate": 2e-05,
      "loss": 0.5139216423034668,
      "step": 4220
    },
    {
      "epoch": 0.281624500665779,
      "grad_norm": 0.7305184025822071,
      "learning_rate": 2e-05,
      "loss": 0.5103124618530274,
      "step": 4230
    },
    {
      "epoch": 0.2822902796271638,
      "grad_norm": 0.7769183635919626,
      "learning_rate": 2e-05,
      "loss": 0.5173744201660156,
      "step": 4240
    },
    {
      "epoch": 0.2829560585885486,
      "grad_norm": 0.8671411607302318,
      "learning_rate": 2e-05,
      "loss": 0.5169537544250489,
      "step": 4250
    },
    {
      "epoch": 0.2836218375499334,
      "grad_norm": 0.8184060581459874,
      "learning_rate": 2e-05,
      "loss": 0.495294189453125,
      "step": 4260
    },
    {
      "epoch": 0.28428761651131823,
      "grad_norm": 4.962472530007058,
      "learning_rate": 2e-05,
      "loss": 0.5087120532989502,
      "step": 4270
    },
    {
      "epoch": 0.28495339547270304,
      "grad_norm": 1.1049434027737588,
      "learning_rate": 2e-05,
      "loss": 0.5020315170288085,
      "step": 4280
    },
    {
      "epoch": 0.2856191744340879,
      "grad_norm": 0.7357278217772826,
      "learning_rate": 2e-05,
      "loss": 0.5621419429779053,
      "step": 4290
    },
    {
      "epoch": 0.2862849533954727,
      "grad_norm": 1.0296164853322647,
      "learning_rate": 2e-05,
      "loss": 0.5030396461486817,
      "step": 4300
    },
    {
      "epoch": 0.28695073235685753,
      "grad_norm": 0.7110530702001324,
      "learning_rate": 2e-05,
      "loss": 0.5017417907714844,
      "step": 4310
    },
    {
      "epoch": 0.28761651131824234,
      "grad_norm": 0.8211045606030541,
      "learning_rate": 2e-05,
      "loss": 0.4632246017456055,
      "step": 4320
    },
    {
      "epoch": 0.28828229027962715,
      "grad_norm": 1.5753881882549339,
      "learning_rate": 2e-05,
      "loss": 0.5080633640289307,
      "step": 4330
    },
    {
      "epoch": 0.28894806924101196,
      "grad_norm": 0.8909410480041041,
      "learning_rate": 2e-05,
      "loss": 0.5428739070892334,
      "step": 4340
    },
    {
      "epoch": 0.28961384820239683,
      "grad_norm": 1.5617369322397225,
      "learning_rate": 2e-05,
      "loss": 0.4732872486114502,
      "step": 4350
    },
    {
      "epoch": 0.29027962716378164,
      "grad_norm": 4.252614778221922,
      "learning_rate": 2e-05,
      "loss": 0.49360170364379885,
      "step": 4360
    },
    {
      "epoch": 0.29094540612516645,
      "grad_norm": 2.1795469558459484,
      "learning_rate": 2e-05,
      "loss": 0.5400909423828125,
      "step": 4370
    },
    {
      "epoch": 0.29161118508655126,
      "grad_norm": 46.737493908115894,
      "learning_rate": 2e-05,
      "loss": 0.5070481300354004,
      "step": 4380
    },
    {
      "epoch": 0.29227696404793607,
      "grad_norm": 0.9798099015876958,
      "learning_rate": 2e-05,
      "loss": 0.5294137954711914,
      "step": 4390
    },
    {
      "epoch": 0.2929427430093209,
      "grad_norm": 1.1447685883224843,
      "learning_rate": 2e-05,
      "loss": 0.5158234596252441,
      "step": 4400
    },
    {
      "epoch": 0.29360852197070575,
      "grad_norm": 3.0514208327110746,
      "learning_rate": 2e-05,
      "loss": 0.5154179096221924,
      "step": 4410
    },
    {
      "epoch": 0.29427430093209056,
      "grad_norm": 1.2711387879827873,
      "learning_rate": 2e-05,
      "loss": 0.5166564464569092,
      "step": 4420
    },
    {
      "epoch": 0.29494007989347537,
      "grad_norm": 0.9336151617677494,
      "learning_rate": 2e-05,
      "loss": 0.5096632957458496,
      "step": 4430
    },
    {
      "epoch": 0.2956058588548602,
      "grad_norm": 1.0960951819180906,
      "learning_rate": 2e-05,
      "loss": 0.5217493057250977,
      "step": 4440
    },
    {
      "epoch": 0.296271637816245,
      "grad_norm": 1.2246520392850682,
      "learning_rate": 2e-05,
      "loss": 0.5428920745849609,
      "step": 4450
    },
    {
      "epoch": 0.2969374167776298,
      "grad_norm": 1.5407104050534581,
      "learning_rate": 2e-05,
      "loss": 0.4850168228149414,
      "step": 4460
    },
    {
      "epoch": 0.29760319573901467,
      "grad_norm": 2.653733671013883,
      "learning_rate": 2e-05,
      "loss": 0.5228505611419678,
      "step": 4470
    },
    {
      "epoch": 0.2982689747003995,
      "grad_norm": 0.8770413062868089,
      "learning_rate": 2e-05,
      "loss": 0.47765750885009767,
      "step": 4480
    },
    {
      "epoch": 0.2989347536617843,
      "grad_norm": 1.8454004830012494,
      "learning_rate": 2e-05,
      "loss": 0.5130658149719238,
      "step": 4490
    },
    {
      "epoch": 0.2996005326231691,
      "grad_norm": 0.6934268742630287,
      "learning_rate": 2e-05,
      "loss": 0.5282979011535645,
      "step": 4500
    },
    {
      "epoch": 0.3002663115845539,
      "grad_norm": 2.8215164931223082,
      "learning_rate": 2e-05,
      "loss": 0.5136002540588379,
      "step": 4510
    },
    {
      "epoch": 0.3009320905459387,
      "grad_norm": 8.383895619888145,
      "learning_rate": 2e-05,
      "loss": 0.4982264995574951,
      "step": 4520
    },
    {
      "epoch": 0.3015978695073236,
      "grad_norm": 1.0776614889728835,
      "learning_rate": 2e-05,
      "loss": 0.5203305721282959,
      "step": 4530
    },
    {
      "epoch": 0.3022636484687084,
      "grad_norm": 2.529712905981004,
      "learning_rate": 2e-05,
      "loss": 0.4988111019134521,
      "step": 4540
    },
    {
      "epoch": 0.3029294274300932,
      "grad_norm": 1.017873710409615,
      "learning_rate": 2e-05,
      "loss": 0.4757256507873535,
      "step": 4550
    },
    {
      "epoch": 0.303595206391478,
      "grad_norm": 0.862823034925489,
      "learning_rate": 2e-05,
      "loss": 0.4983024597167969,
      "step": 4560
    },
    {
      "epoch": 0.30426098535286283,
      "grad_norm": 1.0533850936853175,
      "learning_rate": 2e-05,
      "loss": 0.5166812896728515,
      "step": 4570
    },
    {
      "epoch": 0.30492676431424764,
      "grad_norm": 0.7007000807917756,
      "learning_rate": 2e-05,
      "loss": 0.5053935527801514,
      "step": 4580
    },
    {
      "epoch": 0.3055925432756325,
      "grad_norm": 0.6764936087038914,
      "learning_rate": 2e-05,
      "loss": 0.5249120235443115,
      "step": 4590
    },
    {
      "epoch": 0.3062583222370173,
      "grad_norm": 438.7111503935389,
      "learning_rate": 2e-05,
      "loss": 0.5005148887634278,
      "step": 4600
    },
    {
      "epoch": 0.30692410119840213,
      "grad_norm": 0.9627087478851661,
      "learning_rate": 2e-05,
      "loss": 0.49178237915039064,
      "step": 4610
    },
    {
      "epoch": 0.30758988015978694,
      "grad_norm": 0.6471933696681087,
      "learning_rate": 2e-05,
      "loss": 0.510036039352417,
      "step": 4620
    },
    {
      "epoch": 0.30825565912117175,
      "grad_norm": 0.767068149385643,
      "learning_rate": 2e-05,
      "loss": 0.5009476184844971,
      "step": 4630
    },
    {
      "epoch": 0.30892143808255657,
      "grad_norm": 0.9288596418358097,
      "learning_rate": 2e-05,
      "loss": 0.4755398750305176,
      "step": 4640
    },
    {
      "epoch": 0.30958721704394143,
      "grad_norm": 0.8546226633973462,
      "learning_rate": 2e-05,
      "loss": 0.49333486557006834,
      "step": 4650
    },
    {
      "epoch": 0.31025299600532624,
      "grad_norm": 2.34400020867446,
      "learning_rate": 2e-05,
      "loss": 0.5015472412109375,
      "step": 4660
    },
    {
      "epoch": 0.31091877496671105,
      "grad_norm": 1.0664487262898827,
      "learning_rate": 2e-05,
      "loss": 0.5411473751068115,
      "step": 4670
    },
    {
      "epoch": 0.31158455392809586,
      "grad_norm": 1.724676337750479,
      "learning_rate": 2e-05,
      "loss": 0.523719596862793,
      "step": 4680
    },
    {
      "epoch": 0.3122503328894807,
      "grad_norm": 0.8996838164333144,
      "learning_rate": 2e-05,
      "loss": 0.5242976188659668,
      "step": 4690
    },
    {
      "epoch": 0.3129161118508655,
      "grad_norm": 0.9204555639643998,
      "learning_rate": 2e-05,
      "loss": 0.5159234046936035,
      "step": 4700
    },
    {
      "epoch": 0.31358189081225035,
      "grad_norm": 1.1892068717869835,
      "learning_rate": 2e-05,
      "loss": 0.4924888610839844,
      "step": 4710
    },
    {
      "epoch": 0.31424766977363516,
      "grad_norm": 1.325027915656967,
      "learning_rate": 2e-05,
      "loss": 0.5512666225433349,
      "step": 4720
    },
    {
      "epoch": 0.31491344873502,
      "grad_norm": 1.4734875405775765,
      "learning_rate": 2e-05,
      "loss": 0.49477415084838866,
      "step": 4730
    },
    {
      "epoch": 0.3155792276964048,
      "grad_norm": 0.6161735403162022,
      "learning_rate": 2e-05,
      "loss": 0.48506755828857423,
      "step": 4740
    },
    {
      "epoch": 0.3162450066577896,
      "grad_norm": 0.7142288807535281,
      "learning_rate": 2e-05,
      "loss": 0.4864082336425781,
      "step": 4750
    },
    {
      "epoch": 0.3169107856191744,
      "grad_norm": 2.0333361583638565,
      "learning_rate": 2e-05,
      "loss": 0.48851466178894043,
      "step": 4760
    },
    {
      "epoch": 0.3175765645805593,
      "grad_norm": 58.427645145966714,
      "learning_rate": 2e-05,
      "loss": 0.48818154335021974,
      "step": 4770
    },
    {
      "epoch": 0.3182423435419441,
      "grad_norm": 1.088496232981399,
      "learning_rate": 2e-05,
      "loss": 0.48693408966064455,
      "step": 4780
    },
    {
      "epoch": 0.3189081225033289,
      "grad_norm": 1.2095167136075542,
      "learning_rate": 2e-05,
      "loss": 0.48954501152038576,
      "step": 4790
    },
    {
      "epoch": 0.3195739014647137,
      "grad_norm": 1.3071847560213787,
      "learning_rate": 2e-05,
      "loss": 0.5208199501037598,
      "step": 4800
    },
    {
      "epoch": 0.3202396804260985,
      "grad_norm": 1.8313314268520675,
      "learning_rate": 2e-05,
      "loss": 0.5170340061187744,
      "step": 4810
    },
    {
      "epoch": 0.3209054593874834,
      "grad_norm": 1.0008270470946354,
      "learning_rate": 2e-05,
      "loss": 0.5212603569030761,
      "step": 4820
    },
    {
      "epoch": 0.3215712383488682,
      "grad_norm": 1.6362424448696546,
      "learning_rate": 2e-05,
      "loss": 0.514595365524292,
      "step": 4830
    },
    {
      "epoch": 0.322237017310253,
      "grad_norm": 0.9151714529687154,
      "learning_rate": 2e-05,
      "loss": 0.5061282157897949,
      "step": 4840
    },
    {
      "epoch": 0.3229027962716378,
      "grad_norm": 1.5011305525232022,
      "learning_rate": 2e-05,
      "loss": 0.49297170639038085,
      "step": 4850
    },
    {
      "epoch": 0.3235685752330226,
      "grad_norm": 1.3101773214730332,
      "learning_rate": 2e-05,
      "loss": 0.48419342041015623,
      "step": 4860
    },
    {
      "epoch": 0.32423435419440744,
      "grad_norm": 1.5467320135060838,
      "learning_rate": 2e-05,
      "loss": 0.4971457481384277,
      "step": 4870
    },
    {
      "epoch": 0.3249001331557923,
      "grad_norm": 4.155059692309868,
      "learning_rate": 2e-05,
      "loss": 0.4962268829345703,
      "step": 4880
    },
    {
      "epoch": 0.3255659121171771,
      "grad_norm": 6.451517523644276,
      "learning_rate": 2e-05,
      "loss": 0.5301436901092529,
      "step": 4890
    },
    {
      "epoch": 0.3262316910785619,
      "grad_norm": 2.8254680138513075,
      "learning_rate": 2e-05,
      "loss": 0.494490909576416,
      "step": 4900
    },
    {
      "epoch": 0.32689747003994674,
      "grad_norm": 1.160515203569708,
      "learning_rate": 2e-05,
      "loss": 0.5313853263854981,
      "step": 4910
    },
    {
      "epoch": 0.32756324900133155,
      "grad_norm": 0.8096846986841756,
      "learning_rate": 2e-05,
      "loss": 0.49776277542114256,
      "step": 4920
    },
    {
      "epoch": 0.32822902796271636,
      "grad_norm": 1.0688419576054293,
      "learning_rate": 2e-05,
      "loss": 0.5043133735656739,
      "step": 4930
    },
    {
      "epoch": 0.3288948069241012,
      "grad_norm": 2.002858784166947,
      "learning_rate": 2e-05,
      "loss": 0.5301778316497803,
      "step": 4940
    },
    {
      "epoch": 0.32956058588548603,
      "grad_norm": 1.898504116831972,
      "learning_rate": 2e-05,
      "loss": 0.5347577571868897,
      "step": 4950
    },
    {
      "epoch": 0.33022636484687085,
      "grad_norm": 0.9301687850623508,
      "learning_rate": 2e-05,
      "loss": 0.5165724754333496,
      "step": 4960
    },
    {
      "epoch": 0.33089214380825566,
      "grad_norm": 0.9796284630301676,
      "learning_rate": 2e-05,
      "loss": 0.5051754474639892,
      "step": 4970
    },
    {
      "epoch": 0.33155792276964047,
      "grad_norm": 0.9581790947566668,
      "learning_rate": 2e-05,
      "loss": 0.4980586051940918,
      "step": 4980
    },
    {
      "epoch": 0.3322237017310253,
      "grad_norm": 0.8199502824104467,
      "learning_rate": 2e-05,
      "loss": 0.4854344367980957,
      "step": 4990
    },
    {
      "epoch": 0.33288948069241014,
      "grad_norm": 0.6208195319841088,
      "learning_rate": 2e-05,
      "loss": 0.49005565643310545,
      "step": 5000
    },
    {
      "epoch": 0.33355525965379496,
      "grad_norm": 1.056911984111506,
      "learning_rate": 2e-05,
      "loss": 0.5171819210052491,
      "step": 5010
    },
    {
      "epoch": 0.33422103861517977,
      "grad_norm": 0.8460983418852028,
      "learning_rate": 2e-05,
      "loss": 0.4846088409423828,
      "step": 5020
    },
    {
      "epoch": 0.3348868175765646,
      "grad_norm": 0.7136039523236217,
      "learning_rate": 2e-05,
      "loss": 0.5104721069335938,
      "step": 5030
    },
    {
      "epoch": 0.3355525965379494,
      "grad_norm": 0.6636636308357498,
      "learning_rate": 2e-05,
      "loss": 0.4761074066162109,
      "step": 5040
    },
    {
      "epoch": 0.3362183754993342,
      "grad_norm": 0.7933897502798387,
      "learning_rate": 2e-05,
      "loss": 0.4880530834197998,
      "step": 5050
    },
    {
      "epoch": 0.33688415446071907,
      "grad_norm": 1.242647788413393,
      "learning_rate": 2e-05,
      "loss": 0.532662296295166,
      "step": 5060
    },
    {
      "epoch": 0.3375499334221039,
      "grad_norm": 0.7985076935275303,
      "learning_rate": 2e-05,
      "loss": 0.5067514419555664,
      "step": 5070
    },
    {
      "epoch": 0.3382157123834887,
      "grad_norm": 3.3131476459833813,
      "learning_rate": 2e-05,
      "loss": 0.5129423141479492,
      "step": 5080
    },
    {
      "epoch": 0.3388814913448735,
      "grad_norm": 0.7756571060344593,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.4408143162727356,
      "step": 5090
    },
    {
      "epoch": 0.3395472703062583,
      "grad_norm": 2.3985605866210262,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.5461221694946289,
      "step": 5100
    },
    {
      "epoch": 0.3402130492676431,
      "grad_norm": 0.6857914319174409,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.5149670600891113,
      "step": 5110
    },
    {
      "epoch": 0.340878828229028,
      "grad_norm": 0.8572112097655114,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.4913291931152344,
      "step": 5120
    },
    {
      "epoch": 0.3415446071904128,
      "grad_norm": 0.8218280452027896,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.4919825553894043,
      "step": 5130
    },
    {
      "epoch": 0.3422103861517976,
      "grad_norm": 1.0767972589940784,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.5190771102905274,
      "step": 5140
    },
    {
      "epoch": 0.3428761651131824,
      "grad_norm": 3.149366897283882,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.48006448745727537,
      "step": 5150
    },
    {
      "epoch": 0.34354194407456723,
      "grad_norm": 1.019813332886354,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.4808450698852539,
      "step": 5160
    },
    {
      "epoch": 0.34420772303595204,
      "grad_norm": 6.174985300797463,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.47633891105651854,
      "step": 5170
    },
    {
      "epoch": 0.3448735019973369,
      "grad_norm": 1.14853541763018,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.5200161457061767,
      "step": 5180
    },
    {
      "epoch": 0.3455392809587217,
      "grad_norm": 1.069176312256067,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.4952213764190674,
      "step": 5190
    },
    {
      "epoch": 0.34620505992010653,
      "grad_norm": 10.768653307613452,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.4824930191040039,
      "step": 5200
    },
    {
      "epoch": 0.34687083888149134,
      "grad_norm": 1.2869596202379514,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.5099661827087403,
      "step": 5210
    },
    {
      "epoch": 0.34753661784287615,
      "grad_norm": 1.2183433330377245,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.5001060485839843,
      "step": 5220
    },
    {
      "epoch": 0.34820239680426096,
      "grad_norm": 1.2995100747610588,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.5110831260681152,
      "step": 5230
    },
    {
      "epoch": 0.3488681757656458,
      "grad_norm": 1.365753555399226,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.5009112358093262,
      "step": 5240
    },
    {
      "epoch": 0.34953395472703064,
      "grad_norm": 1.8979390966638185,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.49884347915649413,
      "step": 5250
    },
    {
      "epoch": 0.35019973368841545,
      "grad_norm": 0.8128821806768782,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.4992386817932129,
      "step": 5260
    },
    {
      "epoch": 0.35086551264980026,
      "grad_norm": 0.9654544643343766,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.5062578201293946,
      "step": 5270
    },
    {
      "epoch": 0.35153129161118507,
      "grad_norm": 0.8915461134319917,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.48395280838012694,
      "step": 5280
    },
    {
      "epoch": 0.3521970705725699,
      "grad_norm": 0.9170445495096412,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.5183405399322509,
      "step": 5290
    },
    {
      "epoch": 0.35286284953395475,
      "grad_norm": 0.9732064925985549,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.4931805610656738,
      "step": 5300
    },
    {
      "epoch": 0.35352862849533956,
      "grad_norm": 0.9237746226939504,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.505335521697998,
      "step": 5310
    },
    {
      "epoch": 0.35419440745672437,
      "grad_norm": 0.7826548096591597,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.4760289192199707,
      "step": 5320
    },
    {
      "epoch": 0.3548601864181092,
      "grad_norm": 0.8231290346024404,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.5108865261077881,
      "step": 5330
    },
    {
      "epoch": 0.355525965379494,
      "grad_norm": 1.047614718919616,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.5170565605163574,
      "step": 5340
    },
    {
      "epoch": 0.3561917443408788,
      "grad_norm": 0.9280220138304359,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.5181658267974854,
      "step": 5350
    },
    {
      "epoch": 0.35685752330226367,
      "grad_norm": 1.4142726313406448,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.49065260887145995,
      "step": 5360
    },
    {
      "epoch": 0.3575233022636485,
      "grad_norm": 1.0408148298242503,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.5251599788665772,
      "step": 5370
    },
    {
      "epoch": 0.3581890812250333,
      "grad_norm": 0.9804396976124479,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.5034067153930664,
      "step": 5380
    },
    {
      "epoch": 0.3588548601864181,
      "grad_norm": 1.1621560729437315,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.4789381980895996,
      "step": 5390
    },
    {
      "epoch": 0.3595206391478029,
      "grad_norm": 0.8794863782560206,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.49565725326538085,
      "step": 5400
    },
    {
      "epoch": 0.3601864181091877,
      "grad_norm": 0.880541598219319,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.5029505729675293,
      "step": 5410
    },
    {
      "epoch": 0.3608521970705726,
      "grad_norm": 0.9995281848538486,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.46464385986328127,
      "step": 5420
    },
    {
      "epoch": 0.3615179760319574,
      "grad_norm": 1.6441023813290943,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.5014979839324951,
      "step": 5430
    },
    {
      "epoch": 0.3621837549933422,
      "grad_norm": 1.0020518057607677,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.5016984939575195,
      "step": 5440
    },
    {
      "epoch": 0.362849533954727,
      "grad_norm": 2.933462774521928,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.49635066986083987,
      "step": 5450
    },
    {
      "epoch": 0.36351531291611183,
      "grad_norm": 0.8982613488197462,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.4519174575805664,
      "step": 5460
    },
    {
      "epoch": 0.3641810918774967,
      "grad_norm": 1.281010704247116,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.4889498710632324,
      "step": 5470
    },
    {
      "epoch": 0.3648468708388815,
      "grad_norm": 0.8395462864445591,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.4599448680877686,
      "step": 5480
    },
    {
      "epoch": 0.3655126498002663,
      "grad_norm": 0.8487947318988419,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.4669328212738037,
      "step": 5490
    },
    {
      "epoch": 0.36617842876165113,
      "grad_norm": 0.7931036105736808,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.4652420997619629,
      "step": 5500
    },
    {
      "epoch": 0.36684420772303594,
      "grad_norm": 0.7156601763127977,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.4658137321472168,
      "step": 5510
    },
    {
      "epoch": 0.36750998668442075,
      "grad_norm": 1.0942331235164056,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.49527854919433595,
      "step": 5520
    },
    {
      "epoch": 0.3681757656458056,
      "grad_norm": 0.8810841075476273,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.4762737274169922,
      "step": 5530
    },
    {
      "epoch": 0.36884154460719043,
      "grad_norm": 0.8379244806514067,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.4807307720184326,
      "step": 5540
    },
    {
      "epoch": 0.36950732356857524,
      "grad_norm": 0.749850424151331,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.4831793785095215,
      "step": 5550
    },
    {
      "epoch": 0.37017310252996005,
      "grad_norm": 1.6629595522787208,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.4890940189361572,
      "step": 5560
    },
    {
      "epoch": 0.37083888149134486,
      "grad_norm": 0.7847860943751195,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.4997108459472656,
      "step": 5570
    },
    {
      "epoch": 0.3715046604527297,
      "grad_norm": 0.7802688684775558,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.4323744773864746,
      "step": 5580
    },
    {
      "epoch": 0.37217043941411454,
      "grad_norm": 1.2264247948511795,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.4737849235534668,
      "step": 5590
    },
    {
      "epoch": 0.3726364846870839,
      "step": 5597,
      "total_flos": 2650292248117248.0,
      "train_loss": 0.0447272484017543,
      "train_runtime": 7385.9476,
      "train_samples_per_second": 9.093,
      "train_steps_per_second": 0.758
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2650292248117248.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}