{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7452063914780293,
  "eval_steps": 500,
  "global_step": 11193,
  "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": 0.775657106034942,
      "learning_rate": 2e-05,
      "loss": 0.4408143162727356,
      "step": 5090
    },
    {
      "epoch": 0.3395472703062583,
      "grad_norm": 0.7652274824001466,
      "learning_rate": 2e-05,
      "loss": 0.5473819732666015,
      "step": 5100
    },
    {
      "epoch": 0.3402130492676431,
      "grad_norm": 0.7528330171334362,
      "learning_rate": 2e-05,
      "loss": 0.5170705795288086,
      "step": 5110
    },
    {
      "epoch": 0.340878828229028,
      "grad_norm": 0.8925923311037358,
      "learning_rate": 2e-05,
      "loss": 0.492161226272583,
      "step": 5120
    },
    {
      "epoch": 0.3415446071904128,
      "grad_norm": 0.7512360139621633,
      "learning_rate": 2e-05,
      "loss": 0.4925504684448242,
      "step": 5130
    },
    {
      "epoch": 0.3422103861517976,
      "grad_norm": 0.7883107053804542,
      "learning_rate": 2e-05,
      "loss": 0.520624828338623,
      "step": 5140
    },
    {
      "epoch": 0.3428761651131824,
      "grad_norm": 1.4904571413411363,
      "learning_rate": 2e-05,
      "loss": 0.4803762435913086,
      "step": 5150
    },
    {
      "epoch": 0.34354194407456723,
      "grad_norm": 0.8496568961795381,
      "learning_rate": 2e-05,
      "loss": 0.4806715965270996,
      "step": 5160
    },
    {
      "epoch": 0.34420772303595204,
      "grad_norm": 0.7038986423374711,
      "learning_rate": 2e-05,
      "loss": 0.47636804580688474,
      "step": 5170
    },
    {
      "epoch": 0.3448735019973369,
      "grad_norm": 1.2058571551775035,
      "learning_rate": 2e-05,
      "loss": 0.5191559791564941,
      "step": 5180
    },
    {
      "epoch": 0.3455392809587217,
      "grad_norm": 1.2969797923261084,
      "learning_rate": 2e-05,
      "loss": 0.49425506591796875,
      "step": 5190
    },
    {
      "epoch": 0.34620505992010653,
      "grad_norm": 0.7345964232166122,
      "learning_rate": 2e-05,
      "loss": 0.48132872581481934,
      "step": 5200
    },
    {
      "epoch": 0.34687083888149134,
      "grad_norm": 0.929472832564713,
      "learning_rate": 2e-05,
      "loss": 0.5117568969726562,
      "step": 5210
    },
    {
      "epoch": 0.34753661784287615,
      "grad_norm": 0.8733447220411382,
      "learning_rate": 2e-05,
      "loss": 0.502954626083374,
      "step": 5220
    },
    {
      "epoch": 0.34820239680426096,
      "grad_norm": 0.7536704221678274,
      "learning_rate": 2e-05,
      "loss": 0.5149160861968994,
      "step": 5230
    },
    {
      "epoch": 0.3488681757656458,
      "grad_norm": 1.411705813475078,
      "learning_rate": 2e-05,
      "loss": 0.5036342144012451,
      "step": 5240
    },
    {
      "epoch": 0.34953395472703064,
      "grad_norm": 1.6784322827417468,
      "learning_rate": 2e-05,
      "loss": 0.5009526252746582,
      "step": 5250
    },
    {
      "epoch": 0.35019973368841545,
      "grad_norm": 1.2806958131271295,
      "learning_rate": 2e-05,
      "loss": 0.5028404235839844,
      "step": 5260
    },
    {
      "epoch": 0.35086551264980026,
      "grad_norm": 0.8441942144104595,
      "learning_rate": 2e-05,
      "loss": 0.5119084358215332,
      "step": 5270
    },
    {
      "epoch": 0.35153129161118507,
      "grad_norm": 1.1936363701855224,
      "learning_rate": 2e-05,
      "loss": 0.490773344039917,
      "step": 5280
    },
    {
      "epoch": 0.3521970705725699,
      "grad_norm": 1.7859719686594668,
      "learning_rate": 2e-05,
      "loss": 0.5252262115478515,
      "step": 5290
    },
    {
      "epoch": 0.35286284953395475,
      "grad_norm": 6.168779540319371,
      "learning_rate": 2e-05,
      "loss": 0.49935426712036135,
      "step": 5300
    },
    {
      "epoch": 0.35352862849533956,
      "grad_norm": 1.1036589113959252,
      "learning_rate": 2e-05,
      "loss": 0.5132413864135742,
      "step": 5310
    },
    {
      "epoch": 0.35419440745672437,
      "grad_norm": 0.8131745787025006,
      "learning_rate": 2e-05,
      "loss": 0.4843848705291748,
      "step": 5320
    },
    {
      "epoch": 0.3548601864181092,
      "grad_norm": 0.792591102880359,
      "learning_rate": 2e-05,
      "loss": 0.5198407173156738,
      "step": 5330
    },
    {
      "epoch": 0.355525965379494,
      "grad_norm": 1.033479956540613,
      "learning_rate": 2e-05,
      "loss": 0.5262128829956054,
      "step": 5340
    },
    {
      "epoch": 0.3561917443408788,
      "grad_norm": 1.4813314615278566,
      "learning_rate": 2e-05,
      "loss": 0.5308228492736816,
      "step": 5350
    },
    {
      "epoch": 0.35685752330226367,
      "grad_norm": 1.0987602810034713,
      "learning_rate": 2e-05,
      "loss": 0.5044429779052735,
      "step": 5360
    },
    {
      "epoch": 0.3575233022636485,
      "grad_norm": 1.3231782837912938,
      "learning_rate": 2e-05,
      "loss": 0.5401172637939453,
      "step": 5370
    },
    {
      "epoch": 0.3581890812250333,
      "grad_norm": 0.9658139111592468,
      "learning_rate": 2e-05,
      "loss": 0.5170308113098144,
      "step": 5380
    },
    {
      "epoch": 0.3588548601864181,
      "grad_norm": 1.149361551249373,
      "learning_rate": 2e-05,
      "loss": 0.49140563011169436,
      "step": 5390
    },
    {
      "epoch": 0.3595206391478029,
      "grad_norm": 1.0762368075970354,
      "learning_rate": 2e-05,
      "loss": 0.5080306053161621,
      "step": 5400
    },
    {
      "epoch": 0.3601864181091877,
      "grad_norm": 0.7558702879646912,
      "learning_rate": 2e-05,
      "loss": 0.5148478507995605,
      "step": 5410
    },
    {
      "epoch": 0.3608521970705726,
      "grad_norm": 0.8387994614324568,
      "learning_rate": 2e-05,
      "loss": 0.47871007919311526,
      "step": 5420
    },
    {
      "epoch": 0.3615179760319574,
      "grad_norm": 2.430916382718952,
      "learning_rate": 2e-05,
      "loss": 0.5179977416992188,
      "step": 5430
    },
    {
      "epoch": 0.3621837549933422,
      "grad_norm": 2.018684808859819,
      "learning_rate": 2e-05,
      "loss": 0.517619514465332,
      "step": 5440
    },
    {
      "epoch": 0.362849533954727,
      "grad_norm": 1.458895990874264,
      "learning_rate": 2e-05,
      "loss": 0.5117883682250977,
      "step": 5450
    },
    {
      "epoch": 0.36351531291611183,
      "grad_norm": 2.1197641617809704,
      "learning_rate": 2e-05,
      "loss": 0.46628618240356445,
      "step": 5460
    },
    {
      "epoch": 0.3641810918774967,
      "grad_norm": 1.6213976276148911,
      "learning_rate": 2e-05,
      "loss": 0.5075090408325196,
      "step": 5470
    },
    {
      "epoch": 0.3648468708388815,
      "grad_norm": 0.7646733554463707,
      "learning_rate": 2e-05,
      "loss": 0.4746342658996582,
      "step": 5480
    },
    {
      "epoch": 0.3655126498002663,
      "grad_norm": 2.6657957660186673,
      "learning_rate": 2e-05,
      "loss": 0.4818234443664551,
      "step": 5490
    },
    {
      "epoch": 0.36617842876165113,
      "grad_norm": 0.9121394608957039,
      "learning_rate": 2e-05,
      "loss": 0.4824518203735352,
      "step": 5500
    },
    {
      "epoch": 0.36684420772303594,
      "grad_norm": 1.6135734406211635,
      "learning_rate": 2e-05,
      "loss": 0.48235087394714354,
      "step": 5510
    },
    {
      "epoch": 0.36750998668442075,
      "grad_norm": 1.3960325356226222,
      "learning_rate": 2e-05,
      "loss": 0.5139723300933838,
      "step": 5520
    },
    {
      "epoch": 0.3681757656458056,
      "grad_norm": 1.17005063345042,
      "learning_rate": 2e-05,
      "loss": 0.49424166679382325,
      "step": 5530
    },
    {
      "epoch": 0.36884154460719043,
      "grad_norm": 1.0011370822085015,
      "learning_rate": 2e-05,
      "loss": 0.49950056076049804,
      "step": 5540
    },
    {
      "epoch": 0.36950732356857524,
      "grad_norm": 2.1395692439580847,
      "learning_rate": 2e-05,
      "loss": 0.5043370246887207,
      "step": 5550
    },
    {
      "epoch": 0.37017310252996005,
      "grad_norm": 1.0607498428333721,
      "learning_rate": 2e-05,
      "loss": 0.5101522922515869,
      "step": 5560
    },
    {
      "epoch": 0.37083888149134486,
      "grad_norm": 2.1530841301596415,
      "learning_rate": 2e-05,
      "loss": 0.5189852714538574,
      "step": 5570
    },
    {
      "epoch": 0.3715046604527297,
      "grad_norm": 0.9344499625237342,
      "learning_rate": 2e-05,
      "loss": 0.4525765419006348,
      "step": 5580
    },
    {
      "epoch": 0.37217043941411454,
      "grad_norm": 2.2973121109417947,
      "learning_rate": 2e-05,
      "loss": 0.4925910472869873,
      "step": 5590
    },
    {
      "epoch": 0.37283621837549935,
      "grad_norm": 0.8085612497191603,
      "learning_rate": 2e-05,
      "loss": 0.5140914916992188,
      "step": 5600
    },
    {
      "epoch": 0.37350199733688416,
      "grad_norm": 2.6831933580381047,
      "learning_rate": 2e-05,
      "loss": 0.49075927734375,
      "step": 5610
    },
    {
      "epoch": 0.374167776298269,
      "grad_norm": 1.1974083378969524,
      "learning_rate": 2e-05,
      "loss": 0.5222440242767334,
      "step": 5620
    },
    {
      "epoch": 0.3748335552596538,
      "grad_norm": 1.4717276840617313,
      "learning_rate": 2e-05,
      "loss": 0.47299890518188475,
      "step": 5630
    },
    {
      "epoch": 0.3754993342210386,
      "grad_norm": 1.4963554967642803,
      "learning_rate": 2e-05,
      "loss": 0.5043756484985351,
      "step": 5640
    },
    {
      "epoch": 0.37616511318242346,
      "grad_norm": 1.538181281384448,
      "learning_rate": 2e-05,
      "loss": 0.46834144592285154,
      "step": 5650
    },
    {
      "epoch": 0.37683089214380827,
      "grad_norm": 0.9359024975186113,
      "learning_rate": 2e-05,
      "loss": 0.495298957824707,
      "step": 5660
    },
    {
      "epoch": 0.3774966711051931,
      "grad_norm": 0.9498016237961042,
      "learning_rate": 2e-05,
      "loss": 0.48943500518798827,
      "step": 5670
    },
    {
      "epoch": 0.3781624500665779,
      "grad_norm": 2.0225447433344366,
      "learning_rate": 2e-05,
      "loss": 0.5029576301574707,
      "step": 5680
    },
    {
      "epoch": 0.3788282290279627,
      "grad_norm": 10.441604383829482,
      "learning_rate": 2e-05,
      "loss": 0.49347338676452634,
      "step": 5690
    },
    {
      "epoch": 0.3794940079893475,
      "grad_norm": 2.903999643899844,
      "learning_rate": 2e-05,
      "loss": 0.46612205505371096,
      "step": 5700
    },
    {
      "epoch": 0.3801597869507324,
      "grad_norm": 2.2316599399628227,
      "learning_rate": 2e-05,
      "loss": 0.469527530670166,
      "step": 5710
    },
    {
      "epoch": 0.3808255659121172,
      "grad_norm": 1.255029678929026,
      "learning_rate": 2e-05,
      "loss": 0.48831701278686523,
      "step": 5720
    },
    {
      "epoch": 0.381491344873502,
      "grad_norm": 0.875781251283338,
      "learning_rate": 2e-05,
      "loss": 0.5176331520080566,
      "step": 5730
    },
    {
      "epoch": 0.3821571238348868,
      "grad_norm": 0.8971159741066369,
      "learning_rate": 2e-05,
      "loss": 0.4906322479248047,
      "step": 5740
    },
    {
      "epoch": 0.3828229027962716,
      "grad_norm": 0.9952298678667112,
      "learning_rate": 2e-05,
      "loss": 0.5012186527252197,
      "step": 5750
    },
    {
      "epoch": 0.38348868175765644,
      "grad_norm": 0.796070031818567,
      "learning_rate": 2e-05,
      "loss": 0.5225784301757812,
      "step": 5760
    },
    {
      "epoch": 0.3841544607190413,
      "grad_norm": 1.4032363884857786,
      "learning_rate": 2e-05,
      "loss": 0.46617441177368163,
      "step": 5770
    },
    {
      "epoch": 0.3848202396804261,
      "grad_norm": 0.884411405191963,
      "learning_rate": 2e-05,
      "loss": 0.4763692855834961,
      "step": 5780
    },
    {
      "epoch": 0.3854860186418109,
      "grad_norm": 1.060347800280748,
      "learning_rate": 2e-05,
      "loss": 0.4906328201293945,
      "step": 5790
    },
    {
      "epoch": 0.38615179760319573,
      "grad_norm": 0.9419471999470568,
      "learning_rate": 2e-05,
      "loss": 0.48093838691711427,
      "step": 5800
    },
    {
      "epoch": 0.38681757656458055,
      "grad_norm": 0.7113685274186327,
      "learning_rate": 2e-05,
      "loss": 0.5210418701171875,
      "step": 5810
    },
    {
      "epoch": 0.38748335552596536,
      "grad_norm": 1.3679241157840796,
      "learning_rate": 2e-05,
      "loss": 0.49266958236694336,
      "step": 5820
    },
    {
      "epoch": 0.3881491344873502,
      "grad_norm": 0.8280282208394262,
      "learning_rate": 2e-05,
      "loss": 0.5148045539855957,
      "step": 5830
    },
    {
      "epoch": 0.38881491344873503,
      "grad_norm": 2.0427765525764023,
      "learning_rate": 2e-05,
      "loss": 0.5092354774475097,
      "step": 5840
    },
    {
      "epoch": 0.38948069241011984,
      "grad_norm": 0.8286489305640794,
      "learning_rate": 2e-05,
      "loss": 0.49468584060668946,
      "step": 5850
    },
    {
      "epoch": 0.39014647137150466,
      "grad_norm": 0.8383561561280071,
      "learning_rate": 2e-05,
      "loss": 0.5192167282104492,
      "step": 5860
    },
    {
      "epoch": 0.39081225033288947,
      "grad_norm": 1.5206655738216717,
      "learning_rate": 2e-05,
      "loss": 0.4842246055603027,
      "step": 5870
    },
    {
      "epoch": 0.3914780292942743,
      "grad_norm": 0.7438996136637234,
      "learning_rate": 2e-05,
      "loss": 0.46892547607421875,
      "step": 5880
    },
    {
      "epoch": 0.39214380825565914,
      "grad_norm": 1.0964162484222373,
      "learning_rate": 2e-05,
      "loss": 0.48102149963378904,
      "step": 5890
    },
    {
      "epoch": 0.39280958721704395,
      "grad_norm": 1.1776247264973239,
      "learning_rate": 2e-05,
      "loss": 0.49990339279174806,
      "step": 5900
    },
    {
      "epoch": 0.39347536617842876,
      "grad_norm": 1.2952984894591315,
      "learning_rate": 2e-05,
      "loss": 0.5113763332366943,
      "step": 5910
    },
    {
      "epoch": 0.3941411451398136,
      "grad_norm": 0.9461383859629898,
      "learning_rate": 2e-05,
      "loss": 0.4822257995605469,
      "step": 5920
    },
    {
      "epoch": 0.3948069241011984,
      "grad_norm": 0.8312475148424021,
      "learning_rate": 2e-05,
      "loss": 0.4912898063659668,
      "step": 5930
    },
    {
      "epoch": 0.3954727030625832,
      "grad_norm": 1.1414400061604992,
      "learning_rate": 2e-05,
      "loss": 0.5029790878295899,
      "step": 5940
    },
    {
      "epoch": 0.39613848202396806,
      "grad_norm": 7.429931318969115,
      "learning_rate": 2e-05,
      "loss": 0.5471612930297851,
      "step": 5950
    },
    {
      "epoch": 0.3968042609853529,
      "grad_norm": 0.8844615183128913,
      "learning_rate": 2e-05,
      "loss": 0.48610363006591795,
      "step": 5960
    },
    {
      "epoch": 0.3974700399467377,
      "grad_norm": 1.0522176175740205,
      "learning_rate": 2e-05,
      "loss": 0.5192839622497558,
      "step": 5970
    },
    {
      "epoch": 0.3981358189081225,
      "grad_norm": 2.290919601033034,
      "learning_rate": 2e-05,
      "loss": 0.48499088287353515,
      "step": 5980
    },
    {
      "epoch": 0.3988015978695073,
      "grad_norm": 1.064333217359438,
      "learning_rate": 2e-05,
      "loss": 0.4868661403656006,
      "step": 5990
    },
    {
      "epoch": 0.3994673768308921,
      "grad_norm": 1.4924922755088486,
      "learning_rate": 2e-05,
      "loss": 0.505479383468628,
      "step": 6000
    },
    {
      "epoch": 0.400133155792277,
      "grad_norm": 1.5603926261992598,
      "learning_rate": 2e-05,
      "loss": 0.49382658004760743,
      "step": 6010
    },
    {
      "epoch": 0.4007989347536618,
      "grad_norm": 0.8194875928933423,
      "learning_rate": 2e-05,
      "loss": 0.46831388473510743,
      "step": 6020
    },
    {
      "epoch": 0.4014647137150466,
      "grad_norm": 1.3362742936981198,
      "learning_rate": 2e-05,
      "loss": 0.5071053981781006,
      "step": 6030
    },
    {
      "epoch": 0.4021304926764314,
      "grad_norm": 0.924972068348562,
      "learning_rate": 2e-05,
      "loss": 0.48670034408569335,
      "step": 6040
    },
    {
      "epoch": 0.40279627163781623,
      "grad_norm": 1.1878355647972068,
      "learning_rate": 2e-05,
      "loss": 0.4847214698791504,
      "step": 6050
    },
    {
      "epoch": 0.40346205059920104,
      "grad_norm": 1.7822396859925567,
      "learning_rate": 2e-05,
      "loss": 0.5144247531890869,
      "step": 6060
    },
    {
      "epoch": 0.4041278295605859,
      "grad_norm": 4.167255936700916,
      "learning_rate": 2e-05,
      "loss": 0.490191650390625,
      "step": 6070
    },
    {
      "epoch": 0.4047936085219707,
      "grad_norm": 0.9426668915483598,
      "learning_rate": 2e-05,
      "loss": 0.4681194305419922,
      "step": 6080
    },
    {
      "epoch": 0.4054593874833555,
      "grad_norm": 1.19432855799138,
      "learning_rate": 2e-05,
      "loss": 0.5135694980621338,
      "step": 6090
    },
    {
      "epoch": 0.40612516644474034,
      "grad_norm": 1.016706160297129,
      "learning_rate": 2e-05,
      "loss": 0.48879384994506836,
      "step": 6100
    },
    {
      "epoch": 0.40679094540612515,
      "grad_norm": 0.9499011325416631,
      "learning_rate": 2e-05,
      "loss": 0.49025468826293944,
      "step": 6110
    },
    {
      "epoch": 0.40745672436750996,
      "grad_norm": 0.9126588516266851,
      "learning_rate": 2e-05,
      "loss": 0.4749858856201172,
      "step": 6120
    },
    {
      "epoch": 0.4081225033288948,
      "grad_norm": 0.9355171418016612,
      "learning_rate": 2e-05,
      "loss": 0.500188159942627,
      "step": 6130
    },
    {
      "epoch": 0.40878828229027964,
      "grad_norm": 0.8967521419859817,
      "learning_rate": 2e-05,
      "loss": 0.47826151847839354,
      "step": 6140
    },
    {
      "epoch": 0.40945406125166445,
      "grad_norm": 3.3986913744430454,
      "learning_rate": 2e-05,
      "loss": 0.48213891983032225,
      "step": 6150
    },
    {
      "epoch": 0.41011984021304926,
      "grad_norm": 0.9589301814097364,
      "learning_rate": 2e-05,
      "loss": 0.5077547073364258,
      "step": 6160
    },
    {
      "epoch": 0.41078561917443407,
      "grad_norm": 0.7075132959203675,
      "learning_rate": 2e-05,
      "loss": 0.5007363796234131,
      "step": 6170
    },
    {
      "epoch": 0.41145139813581894,
      "grad_norm": 1.3973356247761861,
      "learning_rate": 2e-05,
      "loss": 0.48351297378540037,
      "step": 6180
    },
    {
      "epoch": 0.41211717709720375,
      "grad_norm": 0.9990822682389805,
      "learning_rate": 2e-05,
      "loss": 0.4992527961730957,
      "step": 6190
    },
    {
      "epoch": 0.41278295605858856,
      "grad_norm": 0.751040454402971,
      "learning_rate": 2e-05,
      "loss": 0.48354105949401854,
      "step": 6200
    },
    {
      "epoch": 0.41344873501997337,
      "grad_norm": 0.9423544369082731,
      "learning_rate": 2e-05,
      "loss": 0.4822075843811035,
      "step": 6210
    },
    {
      "epoch": 0.4141145139813582,
      "grad_norm": 1.1799571947317014,
      "learning_rate": 2e-05,
      "loss": 0.4575021743774414,
      "step": 6220
    },
    {
      "epoch": 0.414780292942743,
      "grad_norm": 1.1310985813046948,
      "learning_rate": 2e-05,
      "loss": 0.4705653667449951,
      "step": 6230
    },
    {
      "epoch": 0.41544607190412786,
      "grad_norm": 3.5848890165959792,
      "learning_rate": 2e-05,
      "loss": 0.5174999237060547,
      "step": 6240
    },
    {
      "epoch": 0.41611185086551267,
      "grad_norm": 1.1859685758512668,
      "learning_rate": 2e-05,
      "loss": 0.48163366317749023,
      "step": 6250
    },
    {
      "epoch": 0.4167776298268975,
      "grad_norm": 0.8338719538134148,
      "learning_rate": 2e-05,
      "loss": 0.4845992088317871,
      "step": 6260
    },
    {
      "epoch": 0.4174434087882823,
      "grad_norm": 0.6847959121922226,
      "learning_rate": 2e-05,
      "loss": 0.5286206245422364,
      "step": 6270
    },
    {
      "epoch": 0.4181091877496671,
      "grad_norm": 2.8215900344330516,
      "learning_rate": 2e-05,
      "loss": 0.47819175720214846,
      "step": 6280
    },
    {
      "epoch": 0.4187749667110519,
      "grad_norm": 2.19186478981424,
      "learning_rate": 2e-05,
      "loss": 0.512361478805542,
      "step": 6290
    },
    {
      "epoch": 0.4194407456724368,
      "grad_norm": 1.507534589919609,
      "learning_rate": 2e-05,
      "loss": 0.4772976875305176,
      "step": 6300
    },
    {
      "epoch": 0.4201065246338216,
      "grad_norm": 1.0225274702970308,
      "learning_rate": 2e-05,
      "loss": 0.483499813079834,
      "step": 6310
    },
    {
      "epoch": 0.4207723035952064,
      "grad_norm": 1.1643971555473187,
      "learning_rate": 2e-05,
      "loss": 0.5064920425415039,
      "step": 6320
    },
    {
      "epoch": 0.4214380825565912,
      "grad_norm": 1.0017422252834902,
      "learning_rate": 2e-05,
      "loss": 0.5019732475280761,
      "step": 6330
    },
    {
      "epoch": 0.422103861517976,
      "grad_norm": 0.9506466642669978,
      "learning_rate": 2e-05,
      "loss": 0.4763456344604492,
      "step": 6340
    },
    {
      "epoch": 0.42276964047936083,
      "grad_norm": 3.7782232908769133,
      "learning_rate": 2e-05,
      "loss": 0.4850940704345703,
      "step": 6350
    },
    {
      "epoch": 0.4234354194407457,
      "grad_norm": 1.9304936012316058,
      "learning_rate": 2e-05,
      "loss": 0.5048925876617432,
      "step": 6360
    },
    {
      "epoch": 0.4241011984021305,
      "grad_norm": 9.816309643512046,
      "learning_rate": 2e-05,
      "loss": 0.4963711738586426,
      "step": 6370
    },
    {
      "epoch": 0.4247669773635153,
      "grad_norm": 0.7401904876258701,
      "learning_rate": 2e-05,
      "loss": 0.4862386703491211,
      "step": 6380
    },
    {
      "epoch": 0.42543275632490013,
      "grad_norm": 0.8792518074573596,
      "learning_rate": 2e-05,
      "loss": 0.48553714752197263,
      "step": 6390
    },
    {
      "epoch": 0.42609853528628494,
      "grad_norm": 0.8789546849702128,
      "learning_rate": 2e-05,
      "loss": 0.4846966743469238,
      "step": 6400
    },
    {
      "epoch": 0.42676431424766975,
      "grad_norm": 0.943268034011324,
      "learning_rate": 2e-05,
      "loss": 0.4870206356048584,
      "step": 6410
    },
    {
      "epoch": 0.4274300932090546,
      "grad_norm": 1.5319811601143154,
      "learning_rate": 2e-05,
      "loss": 0.5020936489105224,
      "step": 6420
    },
    {
      "epoch": 0.42809587217043943,
      "grad_norm": 0.7188970431243824,
      "learning_rate": 2e-05,
      "loss": 0.5026087284088134,
      "step": 6430
    },
    {
      "epoch": 0.42876165113182424,
      "grad_norm": 0.7917568688490801,
      "learning_rate": 2e-05,
      "loss": 0.5055661201477051,
      "step": 6440
    },
    {
      "epoch": 0.42942743009320905,
      "grad_norm": 3.971556453210136,
      "learning_rate": 2e-05,
      "loss": 0.4908576488494873,
      "step": 6450
    },
    {
      "epoch": 0.43009320905459386,
      "grad_norm": 0.7417684608876657,
      "learning_rate": 2e-05,
      "loss": 0.49938573837280276,
      "step": 6460
    },
    {
      "epoch": 0.43075898801597867,
      "grad_norm": 0.866749986672127,
      "learning_rate": 2e-05,
      "loss": 0.4738125801086426,
      "step": 6470
    },
    {
      "epoch": 0.43142476697736354,
      "grad_norm": 0.9073505047396723,
      "learning_rate": 2e-05,
      "loss": 0.4906943321228027,
      "step": 6480
    },
    {
      "epoch": 0.43209054593874835,
      "grad_norm": 0.7208107157625016,
      "learning_rate": 2e-05,
      "loss": 0.4763502597808838,
      "step": 6490
    },
    {
      "epoch": 0.43275632490013316,
      "grad_norm": 1.1416440139269408,
      "learning_rate": 2e-05,
      "loss": 0.5031285762786866,
      "step": 6500
    },
    {
      "epoch": 0.43342210386151797,
      "grad_norm": 0.7276079298786019,
      "learning_rate": 2e-05,
      "loss": 0.48192667961120605,
      "step": 6510
    },
    {
      "epoch": 0.4340878828229028,
      "grad_norm": 0.6483938174858527,
      "learning_rate": 2e-05,
      "loss": 0.4865872383117676,
      "step": 6520
    },
    {
      "epoch": 0.4347536617842876,
      "grad_norm": 1.2485547364964256,
      "learning_rate": 2e-05,
      "loss": 0.5085481166839599,
      "step": 6530
    },
    {
      "epoch": 0.43541944074567246,
      "grad_norm": 1.0447058136843723,
      "learning_rate": 2e-05,
      "loss": 0.5132888793945313,
      "step": 6540
    },
    {
      "epoch": 0.43608521970705727,
      "grad_norm": 0.8632279779402634,
      "learning_rate": 2e-05,
      "loss": 0.5005044937133789,
      "step": 6550
    },
    {
      "epoch": 0.4367509986684421,
      "grad_norm": 1.07274041462215,
      "learning_rate": 2e-05,
      "loss": 0.5193679332733154,
      "step": 6560
    },
    {
      "epoch": 0.4374167776298269,
      "grad_norm": 0.9008196239675039,
      "learning_rate": 2e-05,
      "loss": 0.4842292308807373,
      "step": 6570
    },
    {
      "epoch": 0.4380825565912117,
      "grad_norm": 0.9823378743681465,
      "learning_rate": 2e-05,
      "loss": 0.4746143341064453,
      "step": 6580
    },
    {
      "epoch": 0.4387483355525965,
      "grad_norm": 0.7652625759648644,
      "learning_rate": 2e-05,
      "loss": 0.4891983985900879,
      "step": 6590
    },
    {
      "epoch": 0.4394141145139814,
      "grad_norm": 1.6110113829842359,
      "learning_rate": 2e-05,
      "loss": 0.4958930015563965,
      "step": 6600
    },
    {
      "epoch": 0.4400798934753662,
      "grad_norm": 0.971384685590848,
      "learning_rate": 2e-05,
      "loss": 0.5086921691894531,
      "step": 6610
    },
    {
      "epoch": 0.440745672436751,
      "grad_norm": 1.0900481974739966,
      "learning_rate": 2e-05,
      "loss": 0.47820425033569336,
      "step": 6620
    },
    {
      "epoch": 0.4414114513981358,
      "grad_norm": 1.308446078611129,
      "learning_rate": 2e-05,
      "loss": 0.4876715660095215,
      "step": 6630
    },
    {
      "epoch": 0.4420772303595206,
      "grad_norm": 1.264251997343659,
      "learning_rate": 2e-05,
      "loss": 0.5069456100463867,
      "step": 6640
    },
    {
      "epoch": 0.44274300932090543,
      "grad_norm": 0.8113547655313536,
      "learning_rate": 2e-05,
      "loss": 0.4778906345367432,
      "step": 6650
    },
    {
      "epoch": 0.4434087882822903,
      "grad_norm": 1.7476835151613705,
      "learning_rate": 2e-05,
      "loss": 0.5135702610015869,
      "step": 6660
    },
    {
      "epoch": 0.4440745672436751,
      "grad_norm": 1.7993006577859965,
      "learning_rate": 2e-05,
      "loss": 0.49834189414978025,
      "step": 6670
    },
    {
      "epoch": 0.4447403462050599,
      "grad_norm": 1.572357621901521,
      "learning_rate": 2e-05,
      "loss": 0.5126683712005615,
      "step": 6680
    },
    {
      "epoch": 0.44540612516644473,
      "grad_norm": 1.0646233472708153,
      "learning_rate": 2e-05,
      "loss": 0.5019433498382568,
      "step": 6690
    },
    {
      "epoch": 0.44607190412782954,
      "grad_norm": 0.8508349077200343,
      "learning_rate": 2e-05,
      "loss": 0.5023555755615234,
      "step": 6700
    },
    {
      "epoch": 0.44673768308921435,
      "grad_norm": 1.1707454232132926,
      "learning_rate": 2e-05,
      "loss": 0.5022927284240722,
      "step": 6710
    },
    {
      "epoch": 0.4474034620505992,
      "grad_norm": 2.2855637641948903,
      "learning_rate": 2e-05,
      "loss": 0.5250548839569091,
      "step": 6720
    },
    {
      "epoch": 0.44806924101198403,
      "grad_norm": 1.2933835877837678,
      "learning_rate": 2e-05,
      "loss": 0.4873981475830078,
      "step": 6730
    },
    {
      "epoch": 0.44873501997336884,
      "grad_norm": 1.1963092919771603,
      "learning_rate": 2e-05,
      "loss": 0.4907349109649658,
      "step": 6740
    },
    {
      "epoch": 0.44940079893475365,
      "grad_norm": 1.4802368091899663,
      "learning_rate": 2e-05,
      "loss": 0.4964576721191406,
      "step": 6750
    },
    {
      "epoch": 0.45006657789613846,
      "grad_norm": 0.9403949032792729,
      "learning_rate": 2e-05,
      "loss": 0.46328134536743165,
      "step": 6760
    },
    {
      "epoch": 0.4507323568575233,
      "grad_norm": 2.973426782575252,
      "learning_rate": 2e-05,
      "loss": 0.46786975860595703,
      "step": 6770
    },
    {
      "epoch": 0.45139813581890814,
      "grad_norm": 1.2305429165156052,
      "learning_rate": 2e-05,
      "loss": 0.5148143291473388,
      "step": 6780
    },
    {
      "epoch": 0.45206391478029295,
      "grad_norm": 0.9671064165894856,
      "learning_rate": 2e-05,
      "loss": 0.5022674560546875,
      "step": 6790
    },
    {
      "epoch": 0.45272969374167776,
      "grad_norm": 1.148595706718197,
      "learning_rate": 2e-05,
      "loss": 0.45687170028686525,
      "step": 6800
    },
    {
      "epoch": 0.4533954727030626,
      "grad_norm": 1.833519327249137,
      "learning_rate": 2e-05,
      "loss": 0.5314641952514648,
      "step": 6810
    },
    {
      "epoch": 0.4540612516644474,
      "grad_norm": 1.3450121212396629,
      "learning_rate": 2e-05,
      "loss": 0.4872748374938965,
      "step": 6820
    },
    {
      "epoch": 0.45472703062583225,
      "grad_norm": 0.69687319315159,
      "learning_rate": 2e-05,
      "loss": 0.46487088203430177,
      "step": 6830
    },
    {
      "epoch": 0.45539280958721706,
      "grad_norm": 0.9262064215023846,
      "learning_rate": 2e-05,
      "loss": 0.4719380855560303,
      "step": 6840
    },
    {
      "epoch": 0.4560585885486019,
      "grad_norm": 0.9573763177943553,
      "learning_rate": 2e-05,
      "loss": 0.4422274589538574,
      "step": 6850
    },
    {
      "epoch": 0.4567243675099867,
      "grad_norm": 0.8792811545634763,
      "learning_rate": 2e-05,
      "loss": 0.48426046371459963,
      "step": 6860
    },
    {
      "epoch": 0.4573901464713715,
      "grad_norm": 0.997440004218314,
      "learning_rate": 2e-05,
      "loss": 0.47877278327941897,
      "step": 6870
    },
    {
      "epoch": 0.4580559254327563,
      "grad_norm": 1.3272129674895463,
      "learning_rate": 2e-05,
      "loss": 0.4713019371032715,
      "step": 6880
    },
    {
      "epoch": 0.45872170439414117,
      "grad_norm": 5.9222181112288474,
      "learning_rate": 2e-05,
      "loss": 0.4813715934753418,
      "step": 6890
    },
    {
      "epoch": 0.459387483355526,
      "grad_norm": 0.991153860635775,
      "learning_rate": 2e-05,
      "loss": 0.48161749839782714,
      "step": 6900
    },
    {
      "epoch": 0.4600532623169108,
      "grad_norm": 0.9980684155585694,
      "learning_rate": 2e-05,
      "loss": 0.49303412437438965,
      "step": 6910
    },
    {
      "epoch": 0.4607190412782956,
      "grad_norm": 1.7730551677003341,
      "learning_rate": 2e-05,
      "loss": 0.5213896751403808,
      "step": 6920
    },
    {
      "epoch": 0.4613848202396804,
      "grad_norm": 0.7457624635546,
      "learning_rate": 2e-05,
      "loss": 0.5071815967559814,
      "step": 6930
    },
    {
      "epoch": 0.4620505992010652,
      "grad_norm": 0.85323315267625,
      "learning_rate": 2e-05,
      "loss": 0.4767635345458984,
      "step": 6940
    },
    {
      "epoch": 0.4627163781624501,
      "grad_norm": 1.026885698897434,
      "learning_rate": 2e-05,
      "loss": 0.47399234771728516,
      "step": 6950
    },
    {
      "epoch": 0.4633821571238349,
      "grad_norm": 1.7325015032779578,
      "learning_rate": 2e-05,
      "loss": 0.502871322631836,
      "step": 6960
    },
    {
      "epoch": 0.4640479360852197,
      "grad_norm": 0.8967769915712436,
      "learning_rate": 2e-05,
      "loss": 0.4948737621307373,
      "step": 6970
    },
    {
      "epoch": 0.4647137150466045,
      "grad_norm": 0.7160053302657097,
      "learning_rate": 2e-05,
      "loss": 0.4808650016784668,
      "step": 6980
    },
    {
      "epoch": 0.46537949400798934,
      "grad_norm": 1.684702552530139,
      "learning_rate": 2e-05,
      "loss": 0.5181941509246826,
      "step": 6990
    },
    {
      "epoch": 0.46604527296937415,
      "grad_norm": 1.665989816821541,
      "learning_rate": 2e-05,
      "loss": 0.49725189208984377,
      "step": 7000
    },
    {
      "epoch": 0.466711051930759,
      "grad_norm": 0.9257279140501771,
      "learning_rate": 2e-05,
      "loss": 0.5164238929748535,
      "step": 7010
    },
    {
      "epoch": 0.4673768308921438,
      "grad_norm": 0.9757770827971751,
      "learning_rate": 2e-05,
      "loss": 0.48355464935302733,
      "step": 7020
    },
    {
      "epoch": 0.46804260985352863,
      "grad_norm": 0.995831768775156,
      "learning_rate": 2e-05,
      "loss": 0.5016801834106446,
      "step": 7030
    },
    {
      "epoch": 0.46870838881491345,
      "grad_norm": 1.0909467582826373,
      "learning_rate": 2e-05,
      "loss": 0.4890376091003418,
      "step": 7040
    },
    {
      "epoch": 0.46937416777629826,
      "grad_norm": 1.345799174623364,
      "learning_rate": 2e-05,
      "loss": 0.47906742095947263,
      "step": 7050
    },
    {
      "epoch": 0.47003994673768307,
      "grad_norm": 1.070818441502087,
      "learning_rate": 2e-05,
      "loss": 0.4930426597595215,
      "step": 7060
    },
    {
      "epoch": 0.47070572569906793,
      "grad_norm": 0.8888707479320619,
      "learning_rate": 2e-05,
      "loss": 0.4882223606109619,
      "step": 7070
    },
    {
      "epoch": 0.47137150466045274,
      "grad_norm": 0.9059281047844211,
      "learning_rate": 2e-05,
      "loss": 0.46319828033447263,
      "step": 7080
    },
    {
      "epoch": 0.47203728362183756,
      "grad_norm": 2.3477193102434817,
      "learning_rate": 2e-05,
      "loss": 0.4885292053222656,
      "step": 7090
    },
    {
      "epoch": 0.47270306258322237,
      "grad_norm": 0.8478617374496548,
      "learning_rate": 2e-05,
      "loss": 0.46761417388916016,
      "step": 7100
    },
    {
      "epoch": 0.4733688415446072,
      "grad_norm": 4.683256322734641,
      "learning_rate": 2e-05,
      "loss": 0.49107751846313474,
      "step": 7110
    },
    {
      "epoch": 0.474034620505992,
      "grad_norm": 1.898445258918608,
      "learning_rate": 2e-05,
      "loss": 0.4632240295410156,
      "step": 7120
    },
    {
      "epoch": 0.47470039946737685,
      "grad_norm": 1.4108055303045577,
      "learning_rate": 2e-05,
      "loss": 0.48217363357543946,
      "step": 7130
    },
    {
      "epoch": 0.47536617842876167,
      "grad_norm": 3.0439952185895107,
      "learning_rate": 2e-05,
      "loss": 0.5219317436218261,
      "step": 7140
    },
    {
      "epoch": 0.4760319573901465,
      "grad_norm": 1.2094525247407826,
      "learning_rate": 2e-05,
      "loss": 0.5016321182250977,
      "step": 7150
    },
    {
      "epoch": 0.4766977363515313,
      "grad_norm": 0.8186616533507598,
      "learning_rate": 2e-05,
      "loss": 0.4769547462463379,
      "step": 7160
    },
    {
      "epoch": 0.4773635153129161,
      "grad_norm": 0.7746036623550615,
      "learning_rate": 2e-05,
      "loss": 0.5178488731384278,
      "step": 7170
    },
    {
      "epoch": 0.4780292942743009,
      "grad_norm": 1.600161004221614,
      "learning_rate": 2e-05,
      "loss": 0.524369192123413,
      "step": 7180
    },
    {
      "epoch": 0.4786950732356858,
      "grad_norm": 3.742027135396859,
      "learning_rate": 2e-05,
      "loss": 0.4885438919067383,
      "step": 7190
    },
    {
      "epoch": 0.4793608521970706,
      "grad_norm": 1.1736638618640247,
      "learning_rate": 2e-05,
      "loss": 0.5123039722442627,
      "step": 7200
    },
    {
      "epoch": 0.4800266311584554,
      "grad_norm": 0.8545666029078597,
      "learning_rate": 2e-05,
      "loss": 0.5153939247131347,
      "step": 7210
    },
    {
      "epoch": 0.4806924101198402,
      "grad_norm": 1.142023625191342,
      "learning_rate": 2e-05,
      "loss": 0.4551522731781006,
      "step": 7220
    },
    {
      "epoch": 0.481358189081225,
      "grad_norm": 0.8504413214190477,
      "learning_rate": 2e-05,
      "loss": 0.47111029624938966,
      "step": 7230
    },
    {
      "epoch": 0.48202396804260983,
      "grad_norm": 1.2173490679373802,
      "learning_rate": 2e-05,
      "loss": 0.46646413803100584,
      "step": 7240
    },
    {
      "epoch": 0.4826897470039947,
      "grad_norm": 1.4146266250713242,
      "learning_rate": 2e-05,
      "loss": 0.5095558166503906,
      "step": 7250
    },
    {
      "epoch": 0.4833555259653795,
      "grad_norm": 0.8495313800141681,
      "learning_rate": 2e-05,
      "loss": 0.4952106952667236,
      "step": 7260
    },
    {
      "epoch": 0.4840213049267643,
      "grad_norm": 0.7685426807575293,
      "learning_rate": 2e-05,
      "loss": 0.49155611991882325,
      "step": 7270
    },
    {
      "epoch": 0.48468708388814913,
      "grad_norm": 1.0252483767981677,
      "learning_rate": 2e-05,
      "loss": 0.4980724334716797,
      "step": 7280
    },
    {
      "epoch": 0.48535286284953394,
      "grad_norm": 0.9126233315007255,
      "learning_rate": 2e-05,
      "loss": 0.4920331001281738,
      "step": 7290
    },
    {
      "epoch": 0.48601864181091875,
      "grad_norm": 0.8326190943955303,
      "learning_rate": 2e-05,
      "loss": 0.48849968910217284,
      "step": 7300
    },
    {
      "epoch": 0.4866844207723036,
      "grad_norm": 0.751392312893913,
      "learning_rate": 2e-05,
      "loss": 0.48441600799560547,
      "step": 7310
    },
    {
      "epoch": 0.4873501997336884,
      "grad_norm": 0.8301261718217364,
      "learning_rate": 2e-05,
      "loss": 0.4601811408996582,
      "step": 7320
    },
    {
      "epoch": 0.48801597869507324,
      "grad_norm": 0.8599689223442075,
      "learning_rate": 2e-05,
      "loss": 0.48480677604675293,
      "step": 7330
    },
    {
      "epoch": 0.48868175765645805,
      "grad_norm": 0.8623082734645444,
      "learning_rate": 2e-05,
      "loss": 0.4812324047088623,
      "step": 7340
    },
    {
      "epoch": 0.48934753661784286,
      "grad_norm": 0.7405895953498836,
      "learning_rate": 2e-05,
      "loss": 0.4653136730194092,
      "step": 7350
    },
    {
      "epoch": 0.49001331557922767,
      "grad_norm": 2.9300284101298333,
      "learning_rate": 2e-05,
      "loss": 0.48673086166381835,
      "step": 7360
    },
    {
      "epoch": 0.49067909454061254,
      "grad_norm": 0.8811970866664997,
      "learning_rate": 2e-05,
      "loss": 0.4796687126159668,
      "step": 7370
    },
    {
      "epoch": 0.49134487350199735,
      "grad_norm": 0.7740284456527395,
      "learning_rate": 2e-05,
      "loss": 0.4577164173126221,
      "step": 7380
    },
    {
      "epoch": 0.49201065246338216,
      "grad_norm": 0.8337139239662512,
      "learning_rate": 2e-05,
      "loss": 0.5007697582244873,
      "step": 7390
    },
    {
      "epoch": 0.49267643142476697,
      "grad_norm": 0.8168313409903787,
      "learning_rate": 2e-05,
      "loss": 0.48616838455200195,
      "step": 7400
    },
    {
      "epoch": 0.4933422103861518,
      "grad_norm": 1.284076854231399,
      "learning_rate": 2e-05,
      "loss": 0.47765727043151857,
      "step": 7410
    },
    {
      "epoch": 0.4940079893475366,
      "grad_norm": 0.7792258964208438,
      "learning_rate": 2e-05,
      "loss": 0.48416986465454104,
      "step": 7420
    },
    {
      "epoch": 0.49467376830892146,
      "grad_norm": 0.8707394735638327,
      "learning_rate": 2e-05,
      "loss": 0.468843412399292,
      "step": 7430
    },
    {
      "epoch": 0.49533954727030627,
      "grad_norm": 1.0551702129262766,
      "learning_rate": 2e-05,
      "loss": 0.5051675796508789,
      "step": 7440
    },
    {
      "epoch": 0.4960053262316911,
      "grad_norm": 0.8693985219627198,
      "learning_rate": 2e-05,
      "loss": 0.49992761611938474,
      "step": 7450
    },
    {
      "epoch": 0.4966711051930759,
      "grad_norm": 0.7695804914558463,
      "learning_rate": 2e-05,
      "loss": 0.4934567928314209,
      "step": 7460
    },
    {
      "epoch": 0.4973368841544607,
      "grad_norm": 0.8446496292176876,
      "learning_rate": 2e-05,
      "loss": 0.4823023796081543,
      "step": 7470
    },
    {
      "epoch": 0.4980026631158455,
      "grad_norm": 0.7892622578576828,
      "learning_rate": 2e-05,
      "loss": 0.4480029582977295,
      "step": 7480
    },
    {
      "epoch": 0.4986684420772304,
      "grad_norm": 2.5784279718589076,
      "learning_rate": 2e-05,
      "loss": 0.45827531814575195,
      "step": 7490
    },
    {
      "epoch": 0.4993342210386152,
      "grad_norm": 1.1307679281141958,
      "learning_rate": 2e-05,
      "loss": 0.4958056449890137,
      "step": 7500
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.2634270766312885,
      "learning_rate": 2e-05,
      "loss": 0.5069467067718506,
      "step": 7510
    },
    {
      "epoch": 0.5006657789613849,
      "grad_norm": 1.3834648895825863,
      "learning_rate": 2e-05,
      "loss": 0.4881535530090332,
      "step": 7520
    },
    {
      "epoch": 0.5013315579227696,
      "grad_norm": 0.727328447412831,
      "learning_rate": 2e-05,
      "loss": 0.46088576316833496,
      "step": 7530
    },
    {
      "epoch": 0.5019973368841545,
      "grad_norm": 0.8387770216702528,
      "learning_rate": 2e-05,
      "loss": 0.4878988265991211,
      "step": 7540
    },
    {
      "epoch": 0.5026631158455392,
      "grad_norm": 1.0446789904617073,
      "learning_rate": 2e-05,
      "loss": 0.46657586097717285,
      "step": 7550
    },
    {
      "epoch": 0.5033288948069241,
      "grad_norm": 0.7029058219700196,
      "learning_rate": 2e-05,
      "loss": 0.46748805046081543,
      "step": 7560
    },
    {
      "epoch": 0.503994673768309,
      "grad_norm": 0.7755155873705012,
      "learning_rate": 2e-05,
      "loss": 0.44119882583618164,
      "step": 7570
    },
    {
      "epoch": 0.5046604527296937,
      "grad_norm": 0.7713923304256444,
      "learning_rate": 2e-05,
      "loss": 0.49743380546569826,
      "step": 7580
    },
    {
      "epoch": 0.5053262316910786,
      "grad_norm": 0.6984427040476636,
      "learning_rate": 2e-05,
      "loss": 0.47551746368408204,
      "step": 7590
    },
    {
      "epoch": 0.5059920106524634,
      "grad_norm": 2.937034271133466,
      "learning_rate": 2e-05,
      "loss": 0.4885881423950195,
      "step": 7600
    },
    {
      "epoch": 0.5066577896138482,
      "grad_norm": 0.6580762860133796,
      "learning_rate": 2e-05,
      "loss": 0.48879756927490237,
      "step": 7610
    },
    {
      "epoch": 0.507323568575233,
      "grad_norm": 2.7591600468448343,
      "learning_rate": 2e-05,
      "loss": 0.4890413284301758,
      "step": 7620
    },
    {
      "epoch": 0.5079893475366178,
      "grad_norm": 0.8591981752469723,
      "learning_rate": 2e-05,
      "loss": 0.48189892768859866,
      "step": 7630
    },
    {
      "epoch": 0.5086551264980027,
      "grad_norm": 0.5712836240866142,
      "learning_rate": 2e-05,
      "loss": 0.4783182144165039,
      "step": 7640
    },
    {
      "epoch": 0.5093209054593875,
      "grad_norm": 3.602547046931937,
      "learning_rate": 2e-05,
      "loss": 0.49904637336730956,
      "step": 7650
    },
    {
      "epoch": 0.5099866844207723,
      "grad_norm": 1.5521273749430926,
      "learning_rate": 2e-05,
      "loss": 0.4541231632232666,
      "step": 7660
    },
    {
      "epoch": 0.5106524633821571,
      "grad_norm": 0.9989231404516987,
      "learning_rate": 2e-05,
      "loss": 0.49024014472961425,
      "step": 7670
    },
    {
      "epoch": 0.511318242343542,
      "grad_norm": 0.7972036903158172,
      "learning_rate": 2e-05,
      "loss": 0.49155654907226565,
      "step": 7680
    },
    {
      "epoch": 0.5119840213049268,
      "grad_norm": 0.8616423308412363,
      "learning_rate": 2e-05,
      "loss": 0.47920618057250974,
      "step": 7690
    },
    {
      "epoch": 0.5126498002663116,
      "grad_norm": 0.9000956726559709,
      "learning_rate": 2e-05,
      "loss": 0.47533907890319826,
      "step": 7700
    },
    {
      "epoch": 0.5133155792276964,
      "grad_norm": 0.7225877745582483,
      "learning_rate": 2e-05,
      "loss": 0.4796298980712891,
      "step": 7710
    },
    {
      "epoch": 0.5139813581890812,
      "grad_norm": 0.7858092324097452,
      "learning_rate": 2e-05,
      "loss": 0.49879207611083987,
      "step": 7720
    },
    {
      "epoch": 0.5146471371504661,
      "grad_norm": 2.0872344887206116,
      "learning_rate": 2e-05,
      "loss": 0.47209758758544923,
      "step": 7730
    },
    {
      "epoch": 0.5153129161118508,
      "grad_norm": 0.8494975184749811,
      "learning_rate": 2e-05,
      "loss": 0.5321951866149902,
      "step": 7740
    },
    {
      "epoch": 0.5159786950732357,
      "grad_norm": 0.7473862267204129,
      "learning_rate": 2e-05,
      "loss": 0.4625349998474121,
      "step": 7750
    },
    {
      "epoch": 0.5166444740346205,
      "grad_norm": 0.6706512655599621,
      "learning_rate": 2e-05,
      "loss": 0.4736970901489258,
      "step": 7760
    },
    {
      "epoch": 0.5173102529960053,
      "grad_norm": 2.368556755459123,
      "learning_rate": 2e-05,
      "loss": 0.49304471015930174,
      "step": 7770
    },
    {
      "epoch": 0.5179760319573902,
      "grad_norm": 0.7725327974145457,
      "learning_rate": 2e-05,
      "loss": 0.47539687156677246,
      "step": 7780
    },
    {
      "epoch": 0.5186418109187749,
      "grad_norm": 0.7804407496733627,
      "learning_rate": 2e-05,
      "loss": 0.4737649917602539,
      "step": 7790
    },
    {
      "epoch": 0.5193075898801598,
      "grad_norm": 0.703032523251896,
      "learning_rate": 2e-05,
      "loss": 0.49651422500610354,
      "step": 7800
    },
    {
      "epoch": 0.5199733688415447,
      "grad_norm": 1.3808499023037744,
      "learning_rate": 2e-05,
      "loss": 0.4630948543548584,
      "step": 7810
    },
    {
      "epoch": 0.5206391478029294,
      "grad_norm": 0.8019191464170694,
      "learning_rate": 2e-05,
      "loss": 0.48961501121520995,
      "step": 7820
    },
    {
      "epoch": 0.5213049267643143,
      "grad_norm": 0.8867084512913029,
      "learning_rate": 2e-05,
      "loss": 0.4979098796844482,
      "step": 7830
    },
    {
      "epoch": 0.521970705725699,
      "grad_norm": 0.9276859325573332,
      "learning_rate": 2e-05,
      "loss": 0.48037543296813967,
      "step": 7840
    },
    {
      "epoch": 0.5226364846870839,
      "grad_norm": 0.7261974124356179,
      "learning_rate": 2e-05,
      "loss": 0.48922138214111327,
      "step": 7850
    },
    {
      "epoch": 0.5233022636484687,
      "grad_norm": 0.7695790131228593,
      "learning_rate": 2e-05,
      "loss": 0.5080029487609863,
      "step": 7860
    },
    {
      "epoch": 0.5239680426098535,
      "grad_norm": 1.0228370392382693,
      "learning_rate": 2e-05,
      "loss": 0.4830935955047607,
      "step": 7870
    },
    {
      "epoch": 0.5246338215712384,
      "grad_norm": 0.7851859158286717,
      "learning_rate": 2e-05,
      "loss": 0.49563446044921877,
      "step": 7880
    },
    {
      "epoch": 0.5252996005326231,
      "grad_norm": 0.7890166592696046,
      "learning_rate": 2e-05,
      "loss": 0.4983807563781738,
      "step": 7890
    },
    {
      "epoch": 0.525965379494008,
      "grad_norm": 0.8195601483571354,
      "learning_rate": 2e-05,
      "loss": 0.4891640663146973,
      "step": 7900
    },
    {
      "epoch": 0.5266311584553928,
      "grad_norm": 1.2008687813221497,
      "learning_rate": 2e-05,
      "loss": 0.47505483627319334,
      "step": 7910
    },
    {
      "epoch": 0.5272969374167776,
      "grad_norm": 0.8428572428115703,
      "learning_rate": 2e-05,
      "loss": 0.46642284393310546,
      "step": 7920
    },
    {
      "epoch": 0.5279627163781625,
      "grad_norm": 0.7670599661915523,
      "learning_rate": 2e-05,
      "loss": 0.49324731826782225,
      "step": 7930
    },
    {
      "epoch": 0.5286284953395473,
      "grad_norm": 3.805063239175068,
      "learning_rate": 2e-05,
      "loss": 0.4862421989440918,
      "step": 7940
    },
    {
      "epoch": 0.5292942743009321,
      "grad_norm": 1.106871473075721,
      "learning_rate": 2e-05,
      "loss": 0.48893165588378906,
      "step": 7950
    },
    {
      "epoch": 0.5299600532623169,
      "grad_norm": 1.039294003361161,
      "learning_rate": 2e-05,
      "loss": 0.521898603439331,
      "step": 7960
    },
    {
      "epoch": 0.5306258322237017,
      "grad_norm": 0.7446251193275415,
      "learning_rate": 2e-05,
      "loss": 0.4319633483886719,
      "step": 7970
    },
    {
      "epoch": 0.5312916111850865,
      "grad_norm": 0.7346237491688671,
      "learning_rate": 2e-05,
      "loss": 0.5004458904266358,
      "step": 7980
    },
    {
      "epoch": 0.5319573901464714,
      "grad_norm": 0.6859554806351251,
      "learning_rate": 2e-05,
      "loss": 0.48511176109313964,
      "step": 7990
    },
    {
      "epoch": 0.5326231691078562,
      "grad_norm": 1.1296686663056346,
      "learning_rate": 2e-05,
      "loss": 0.465463924407959,
      "step": 8000
    },
    {
      "epoch": 0.533288948069241,
      "grad_norm": 0.8514322988186486,
      "learning_rate": 2e-05,
      "loss": 0.492492151260376,
      "step": 8010
    },
    {
      "epoch": 0.5339547270306259,
      "grad_norm": 0.8419984978314533,
      "learning_rate": 2e-05,
      "loss": 0.48567972183227537,
      "step": 8020
    },
    {
      "epoch": 0.5346205059920106,
      "grad_norm": 1.272464764825967,
      "learning_rate": 2e-05,
      "loss": 0.48035612106323244,
      "step": 8030
    },
    {
      "epoch": 0.5352862849533955,
      "grad_norm": 0.6882165903267149,
      "learning_rate": 2e-05,
      "loss": 0.47874083518981936,
      "step": 8040
    },
    {
      "epoch": 0.5359520639147803,
      "grad_norm": 0.7958586345172132,
      "learning_rate": 2e-05,
      "loss": 0.4805158138275146,
      "step": 8050
    },
    {
      "epoch": 0.5366178428761651,
      "grad_norm": 1.0843343503753258,
      "learning_rate": 2e-05,
      "loss": 0.4629159927368164,
      "step": 8060
    },
    {
      "epoch": 0.53728362183755,
      "grad_norm": 0.8222389169451243,
      "learning_rate": 2e-05,
      "loss": 0.47052717208862305,
      "step": 8070
    },
    {
      "epoch": 0.5379494007989347,
      "grad_norm": 0.8454200674736059,
      "learning_rate": 2e-05,
      "loss": 0.475443172454834,
      "step": 8080
    },
    {
      "epoch": 0.5386151797603196,
      "grad_norm": 0.8131877207571115,
      "learning_rate": 2e-05,
      "loss": 0.47348747253417967,
      "step": 8090
    },
    {
      "epoch": 0.5392809587217043,
      "grad_norm": 0.7118923720106931,
      "learning_rate": 2e-05,
      "loss": 0.4865586280822754,
      "step": 8100
    },
    {
      "epoch": 0.5399467376830892,
      "grad_norm": 0.7197558759652296,
      "learning_rate": 2e-05,
      "loss": 0.5179664611816406,
      "step": 8110
    },
    {
      "epoch": 0.5406125166444741,
      "grad_norm": 0.8121855836954747,
      "learning_rate": 2e-05,
      "loss": 0.4964120864868164,
      "step": 8120
    },
    {
      "epoch": 0.5412782956058588,
      "grad_norm": 0.7255471606329174,
      "learning_rate": 2e-05,
      "loss": 0.49745879173278806,
      "step": 8130
    },
    {
      "epoch": 0.5419440745672437,
      "grad_norm": 0.8832731784464938,
      "learning_rate": 2e-05,
      "loss": 0.49059252738952636,
      "step": 8140
    },
    {
      "epoch": 0.5426098535286284,
      "grad_norm": 0.7339186235202404,
      "learning_rate": 2e-05,
      "loss": 0.461212158203125,
      "step": 8150
    },
    {
      "epoch": 0.5432756324900133,
      "grad_norm": 0.8110658305341921,
      "learning_rate": 2e-05,
      "loss": 0.4811814785003662,
      "step": 8160
    },
    {
      "epoch": 0.5439414114513982,
      "grad_norm": 0.6899032428802847,
      "learning_rate": 2e-05,
      "loss": 0.46393909454345705,
      "step": 8170
    },
    {
      "epoch": 0.5446071904127829,
      "grad_norm": 0.8179002404301037,
      "learning_rate": 2e-05,
      "loss": 0.4860383033752441,
      "step": 8180
    },
    {
      "epoch": 0.5452729693741678,
      "grad_norm": 0.959570190853773,
      "learning_rate": 2e-05,
      "loss": 0.49228992462158205,
      "step": 8190
    },
    {
      "epoch": 0.5459387483355526,
      "grad_norm": 0.6663402833967943,
      "learning_rate": 2e-05,
      "loss": 0.47037534713745116,
      "step": 8200
    },
    {
      "epoch": 0.5466045272969374,
      "grad_norm": 0.7414542357110316,
      "learning_rate": 2e-05,
      "loss": 0.48249106407165526,
      "step": 8210
    },
    {
      "epoch": 0.5472703062583223,
      "grad_norm": 0.706376941230626,
      "learning_rate": 2e-05,
      "loss": 0.4897404670715332,
      "step": 8220
    },
    {
      "epoch": 0.547936085219707,
      "grad_norm": 1.4315752578230279,
      "learning_rate": 2e-05,
      "loss": 0.4952141761779785,
      "step": 8230
    },
    {
      "epoch": 0.5486018641810919,
      "grad_norm": 0.8111190271036457,
      "learning_rate": 2e-05,
      "loss": 0.4929033279418945,
      "step": 8240
    },
    {
      "epoch": 0.5492676431424767,
      "grad_norm": 0.6752383937416857,
      "learning_rate": 2e-05,
      "loss": 0.4828777313232422,
      "step": 8250
    },
    {
      "epoch": 0.5499334221038615,
      "grad_norm": 0.7111521013907585,
      "learning_rate": 2e-05,
      "loss": 0.46419177055358884,
      "step": 8260
    },
    {
      "epoch": 0.5505992010652463,
      "grad_norm": 0.6043837630374758,
      "learning_rate": 2e-05,
      "loss": 0.47538127899169924,
      "step": 8270
    },
    {
      "epoch": 0.5512649800266312,
      "grad_norm": 0.7702930501407046,
      "learning_rate": 2e-05,
      "loss": 0.48886699676513673,
      "step": 8280
    },
    {
      "epoch": 0.551930758988016,
      "grad_norm": 0.7663434328587375,
      "learning_rate": 2e-05,
      "loss": 0.5050689220428467,
      "step": 8290
    },
    {
      "epoch": 0.5525965379494008,
      "grad_norm": 1.8344157948375064,
      "learning_rate": 2e-05,
      "loss": 0.4880217552185059,
      "step": 8300
    },
    {
      "epoch": 0.5532623169107856,
      "grad_norm": 0.7893749449894395,
      "learning_rate": 2e-05,
      "loss": 0.48651580810546874,
      "step": 8310
    },
    {
      "epoch": 0.5539280958721704,
      "grad_norm": 0.9015115092031301,
      "learning_rate": 2e-05,
      "loss": 0.4942117691040039,
      "step": 8320
    },
    {
      "epoch": 0.5545938748335553,
      "grad_norm": 1.2213393794623784,
      "learning_rate": 2e-05,
      "loss": 0.46471633911132815,
      "step": 8330
    },
    {
      "epoch": 0.5552596537949401,
      "grad_norm": 0.5885983178678502,
      "learning_rate": 2e-05,
      "loss": 0.502247953414917,
      "step": 8340
    },
    {
      "epoch": 0.5559254327563249,
      "grad_norm": 0.6755916796665048,
      "learning_rate": 2e-05,
      "loss": 0.4664464473724365,
      "step": 8350
    },
    {
      "epoch": 0.5565912117177098,
      "grad_norm": 1.212098467171452,
      "learning_rate": 2e-05,
      "loss": 0.5075779914855957,
      "step": 8360
    },
    {
      "epoch": 0.5572569906790945,
      "grad_norm": 1.1741950849332599,
      "learning_rate": 2e-05,
      "loss": 0.4766756534576416,
      "step": 8370
    },
    {
      "epoch": 0.5579227696404794,
      "grad_norm": 1.918912446319798,
      "learning_rate": 2e-05,
      "loss": 0.47287635803222655,
      "step": 8380
    },
    {
      "epoch": 0.5585885486018641,
      "grad_norm": 0.9406181130859921,
      "learning_rate": 2e-05,
      "loss": 0.46350693702697754,
      "step": 8390
    },
    {
      "epoch": 0.559254327563249,
      "grad_norm": 0.847791532462806,
      "learning_rate": 2e-05,
      "loss": 0.4719341278076172,
      "step": 8400
    },
    {
      "epoch": 0.5599201065246339,
      "grad_norm": 1.0278059843447318,
      "learning_rate": 2e-05,
      "loss": 0.4813251495361328,
      "step": 8410
    },
    {
      "epoch": 0.5605858854860186,
      "grad_norm": 0.8138642318245913,
      "learning_rate": 2e-05,
      "loss": 0.476318359375,
      "step": 8420
    },
    {
      "epoch": 0.5612516644474035,
      "grad_norm": 0.7921994865514762,
      "learning_rate": 2e-05,
      "loss": 0.4772101879119873,
      "step": 8430
    },
    {
      "epoch": 0.5619174434087882,
      "grad_norm": 1.1483976820898667,
      "learning_rate": 2e-05,
      "loss": 0.4917590141296387,
      "step": 8440
    },
    {
      "epoch": 0.5625832223701731,
      "grad_norm": 0.8452299966753258,
      "learning_rate": 2e-05,
      "loss": 0.48238887786865237,
      "step": 8450
    },
    {
      "epoch": 0.563249001331558,
      "grad_norm": 0.8940397634511973,
      "learning_rate": 2e-05,
      "loss": 0.45291476249694823,
      "step": 8460
    },
    {
      "epoch": 0.5639147802929427,
      "grad_norm": 0.7418972419752081,
      "learning_rate": 2e-05,
      "loss": 0.4984915733337402,
      "step": 8470
    },
    {
      "epoch": 0.5645805592543276,
      "grad_norm": 0.6339385469997258,
      "learning_rate": 2e-05,
      "loss": 0.4605108261108398,
      "step": 8480
    },
    {
      "epoch": 0.5652463382157124,
      "grad_norm": 1.0100024822585743,
      "learning_rate": 2e-05,
      "loss": 0.48880486488342284,
      "step": 8490
    },
    {
      "epoch": 0.5659121171770972,
      "grad_norm": 0.7191837876964811,
      "learning_rate": 2e-05,
      "loss": 0.5096979141235352,
      "step": 8500
    },
    {
      "epoch": 0.566577896138482,
      "grad_norm": 0.747096572168829,
      "learning_rate": 2e-05,
      "loss": 0.4727903366088867,
      "step": 8510
    },
    {
      "epoch": 0.5672436750998668,
      "grad_norm": 0.6821895520965394,
      "learning_rate": 2e-05,
      "loss": 0.4779521942138672,
      "step": 8520
    },
    {
      "epoch": 0.5679094540612517,
      "grad_norm": 2.2156225840562365,
      "learning_rate": 2e-05,
      "loss": 0.49666624069213866,
      "step": 8530
    },
    {
      "epoch": 0.5685752330226365,
      "grad_norm": 1.3252373471820276,
      "learning_rate": 2e-05,
      "loss": 0.4711931228637695,
      "step": 8540
    },
    {
      "epoch": 0.5692410119840213,
      "grad_norm": 0.8529522199183435,
      "learning_rate": 2e-05,
      "loss": 0.497772741317749,
      "step": 8550
    },
    {
      "epoch": 0.5699067909454061,
      "grad_norm": 1.7079860518255314,
      "learning_rate": 2e-05,
      "loss": 0.5088276863098145,
      "step": 8560
    },
    {
      "epoch": 0.570572569906791,
      "grad_norm": 0.9987938468366188,
      "learning_rate": 2e-05,
      "loss": 0.4802361488342285,
      "step": 8570
    },
    {
      "epoch": 0.5712383488681758,
      "grad_norm": 2.692724651838101,
      "learning_rate": 2e-05,
      "loss": 0.5072489738464355,
      "step": 8580
    },
    {
      "epoch": 0.5719041278295606,
      "grad_norm": 0.7252962407263525,
      "learning_rate": 2e-05,
      "loss": 0.4709373950958252,
      "step": 8590
    },
    {
      "epoch": 0.5725699067909454,
      "grad_norm": 1.1427211529495533,
      "learning_rate": 2e-05,
      "loss": 0.5158286094665527,
      "step": 8600
    },
    {
      "epoch": 0.5732356857523302,
      "grad_norm": 1.5735157210374937,
      "learning_rate": 2e-05,
      "loss": 0.45836973190307617,
      "step": 8610
    },
    {
      "epoch": 0.5739014647137151,
      "grad_norm": 0.8012161208163603,
      "learning_rate": 2e-05,
      "loss": 0.46275882720947265,
      "step": 8620
    },
    {
      "epoch": 0.5745672436750998,
      "grad_norm": 0.842025680069841,
      "learning_rate": 2e-05,
      "loss": 0.49875917434692385,
      "step": 8630
    },
    {
      "epoch": 0.5752330226364847,
      "grad_norm": 0.8809575447522143,
      "learning_rate": 2e-05,
      "loss": 0.4847869873046875,
      "step": 8640
    },
    {
      "epoch": 0.5758988015978695,
      "grad_norm": 0.8454915530053966,
      "learning_rate": 2e-05,
      "loss": 0.514919662475586,
      "step": 8650
    },
    {
      "epoch": 0.5765645805592543,
      "grad_norm": 0.9097915365104743,
      "learning_rate": 2e-05,
      "loss": 0.4768527984619141,
      "step": 8660
    },
    {
      "epoch": 0.5772303595206392,
      "grad_norm": 0.7398036246228707,
      "learning_rate": 2e-05,
      "loss": 0.4644573211669922,
      "step": 8670
    },
    {
      "epoch": 0.5778961384820239,
      "grad_norm": 0.8866316956389855,
      "learning_rate": 2e-05,
      "loss": 0.48565168380737306,
      "step": 8680
    },
    {
      "epoch": 0.5785619174434088,
      "grad_norm": 0.7763691637608071,
      "learning_rate": 2e-05,
      "loss": 0.487142276763916,
      "step": 8690
    },
    {
      "epoch": 0.5792276964047937,
      "grad_norm": 1.0578953375821514,
      "learning_rate": 2e-05,
      "loss": 0.5103287696838379,
      "step": 8700
    },
    {
      "epoch": 0.5798934753661784,
      "grad_norm": 8.689768120091141,
      "learning_rate": 2e-05,
      "loss": 0.4705756664276123,
      "step": 8710
    },
    {
      "epoch": 0.5805592543275633,
      "grad_norm": 0.7980330087784825,
      "learning_rate": 2e-05,
      "loss": 0.4759833812713623,
      "step": 8720
    },
    {
      "epoch": 0.581225033288948,
      "grad_norm": 0.7539221114093584,
      "learning_rate": 2e-05,
      "loss": 0.47228446006774905,
      "step": 8730
    },
    {
      "epoch": 0.5818908122503329,
      "grad_norm": 0.7730180675202525,
      "learning_rate": 2e-05,
      "loss": 0.4816134929656982,
      "step": 8740
    },
    {
      "epoch": 0.5825565912117177,
      "grad_norm": 0.810043825424621,
      "learning_rate": 2e-05,
      "loss": 0.48787784576416016,
      "step": 8750
    },
    {
      "epoch": 0.5832223701731025,
      "grad_norm": 0.9066860933351051,
      "learning_rate": 2e-05,
      "loss": 0.46769213676452637,
      "step": 8760
    },
    {
      "epoch": 0.5838881491344874,
      "grad_norm": 0.7126057053899365,
      "learning_rate": 2e-05,
      "loss": 0.4592240810394287,
      "step": 8770
    },
    {
      "epoch": 0.5845539280958721,
      "grad_norm": 0.8498860182874237,
      "learning_rate": 2e-05,
      "loss": 0.5045224189758301,
      "step": 8780
    },
    {
      "epoch": 0.585219707057257,
      "grad_norm": 0.715702260172293,
      "learning_rate": 2e-05,
      "loss": 0.4590615272521973,
      "step": 8790
    },
    {
      "epoch": 0.5858854860186418,
      "grad_norm": 0.7696533007783732,
      "learning_rate": 2e-05,
      "loss": 0.4808809280395508,
      "step": 8800
    },
    {
      "epoch": 0.5865512649800266,
      "grad_norm": 1.0044961232033298,
      "learning_rate": 2e-05,
      "loss": 0.5074020862579346,
      "step": 8810
    },
    {
      "epoch": 0.5872170439414115,
      "grad_norm": 1.0475177221350727,
      "learning_rate": 2e-05,
      "loss": 0.47940635681152344,
      "step": 8820
    },
    {
      "epoch": 0.5878828229027963,
      "grad_norm": 0.7832160232294868,
      "learning_rate": 2e-05,
      "loss": 0.5110118389129639,
      "step": 8830
    },
    {
      "epoch": 0.5885486018641811,
      "grad_norm": 0.89469697863423,
      "learning_rate": 2e-05,
      "loss": 0.49699106216430666,
      "step": 8840
    },
    {
      "epoch": 0.5892143808255659,
      "grad_norm": 1.8016203953480472,
      "learning_rate": 2e-05,
      "loss": 0.4786980628967285,
      "step": 8850
    },
    {
      "epoch": 0.5898801597869507,
      "grad_norm": 0.8041274301462753,
      "learning_rate": 2e-05,
      "loss": 0.5134645462036133,
      "step": 8860
    },
    {
      "epoch": 0.5905459387483355,
      "grad_norm": 0.7541142949713551,
      "learning_rate": 2e-05,
      "loss": 0.471388053894043,
      "step": 8870
    },
    {
      "epoch": 0.5912117177097204,
      "grad_norm": 1.144397104729034,
      "learning_rate": 2e-05,
      "loss": 0.501277494430542,
      "step": 8880
    },
    {
      "epoch": 0.5918774966711052,
      "grad_norm": 2.642426576764901,
      "learning_rate": 2e-05,
      "loss": 0.49772443771362307,
      "step": 8890
    },
    {
      "epoch": 0.59254327563249,
      "grad_norm": 0.6246936027803643,
      "learning_rate": 2e-05,
      "loss": 0.4302422046661377,
      "step": 8900
    },
    {
      "epoch": 0.5932090545938749,
      "grad_norm": 0.7555572487097605,
      "learning_rate": 2e-05,
      "loss": 0.5157959938049317,
      "step": 8910
    },
    {
      "epoch": 0.5938748335552596,
      "grad_norm": 1.6581390037180415,
      "learning_rate": 2e-05,
      "loss": 0.48918676376342773,
      "step": 8920
    },
    {
      "epoch": 0.5945406125166445,
      "grad_norm": 1.4096846959478908,
      "learning_rate": 2e-05,
      "loss": 0.4437868595123291,
      "step": 8930
    },
    {
      "epoch": 0.5952063914780293,
      "grad_norm": 0.7775271685759331,
      "learning_rate": 2e-05,
      "loss": 0.4972583293914795,
      "step": 8940
    },
    {
      "epoch": 0.5958721704394141,
      "grad_norm": 0.7873987194767993,
      "learning_rate": 2e-05,
      "loss": 0.4626293659210205,
      "step": 8950
    },
    {
      "epoch": 0.596537949400799,
      "grad_norm": 0.8473862429848624,
      "learning_rate": 2e-05,
      "loss": 0.48300938606262206,
      "step": 8960
    },
    {
      "epoch": 0.5972037283621837,
      "grad_norm": 2.1686668976531585,
      "learning_rate": 2e-05,
      "loss": 0.4696963310241699,
      "step": 8970
    },
    {
      "epoch": 0.5978695073235686,
      "grad_norm": 0.804638811443482,
      "learning_rate": 2e-05,
      "loss": 0.47777504920959474,
      "step": 8980
    },
    {
      "epoch": 0.5985352862849534,
      "grad_norm": 0.8524933728610727,
      "learning_rate": 2e-05,
      "loss": 0.4884025573730469,
      "step": 8990
    },
    {
      "epoch": 0.5992010652463382,
      "grad_norm": 0.8378155916527288,
      "learning_rate": 2e-05,
      "loss": 0.4769148826599121,
      "step": 9000
    },
    {
      "epoch": 0.5998668442077231,
      "grad_norm": 0.7729997887672165,
      "learning_rate": 2e-05,
      "loss": 0.4607945442199707,
      "step": 9010
    },
    {
      "epoch": 0.6005326231691078,
      "grad_norm": 0.7900526805478449,
      "learning_rate": 2e-05,
      "loss": 0.46518535614013673,
      "step": 9020
    },
    {
      "epoch": 0.6011984021304927,
      "grad_norm": 1.3615958609311933,
      "learning_rate": 2e-05,
      "loss": 0.45151820182800295,
      "step": 9030
    },
    {
      "epoch": 0.6018641810918774,
      "grad_norm": 0.7513620937209765,
      "learning_rate": 2e-05,
      "loss": 0.4551080226898193,
      "step": 9040
    },
    {
      "epoch": 0.6025299600532623,
      "grad_norm": 0.7062211321860575,
      "learning_rate": 2e-05,
      "loss": 0.4844379901885986,
      "step": 9050
    },
    {
      "epoch": 0.6031957390146472,
      "grad_norm": 0.7982336597052364,
      "learning_rate": 2e-05,
      "loss": 0.4716363430023193,
      "step": 9060
    },
    {
      "epoch": 0.6038615179760319,
      "grad_norm": 0.6624870991134735,
      "learning_rate": 2e-05,
      "loss": 0.43236989974975587,
      "step": 9070
    },
    {
      "epoch": 0.6045272969374168,
      "grad_norm": 0.6787841191082821,
      "learning_rate": 2e-05,
      "loss": 0.46657757759094237,
      "step": 9080
    },
    {
      "epoch": 0.6051930758988016,
      "grad_norm": 0.7806378901599479,
      "learning_rate": 2e-05,
      "loss": 0.46466598510742185,
      "step": 9090
    },
    {
      "epoch": 0.6058588548601864,
      "grad_norm": 1.9101449955865821,
      "learning_rate": 2e-05,
      "loss": 0.4862531185150146,
      "step": 9100
    },
    {
      "epoch": 0.6065246338215713,
      "grad_norm": 0.8086871705059603,
      "learning_rate": 2e-05,
      "loss": 0.48087053298950194,
      "step": 9110
    },
    {
      "epoch": 0.607190412782956,
      "grad_norm": 1.3562492286101033,
      "learning_rate": 2e-05,
      "loss": 0.4847963809967041,
      "step": 9120
    },
    {
      "epoch": 0.6078561917443409,
      "grad_norm": 1.6302047634931895,
      "learning_rate": 2e-05,
      "loss": 0.5078322410583496,
      "step": 9130
    },
    {
      "epoch": 0.6085219707057257,
      "grad_norm": 1.3686320838182926,
      "learning_rate": 2e-05,
      "loss": 0.47059993743896483,
      "step": 9140
    },
    {
      "epoch": 0.6091877496671105,
      "grad_norm": 1.3143047719505099,
      "learning_rate": 2e-05,
      "loss": 0.5115951538085938,
      "step": 9150
    },
    {
      "epoch": 0.6098535286284953,
      "grad_norm": 1.967743369209231,
      "learning_rate": 2e-05,
      "loss": 0.47182130813598633,
      "step": 9160
    },
    {
      "epoch": 0.6105193075898802,
      "grad_norm": 1.0523664972844085,
      "learning_rate": 2e-05,
      "loss": 0.456136417388916,
      "step": 9170
    },
    {
      "epoch": 0.611185086551265,
      "grad_norm": 0.8410719946182716,
      "learning_rate": 2e-05,
      "loss": 0.473177433013916,
      "step": 9180
    },
    {
      "epoch": 0.6118508655126498,
      "grad_norm": 1.0094720815589038,
      "learning_rate": 2e-05,
      "loss": 0.4880485534667969,
      "step": 9190
    },
    {
      "epoch": 0.6125166444740346,
      "grad_norm": 1.7224707223587616,
      "learning_rate": 2e-05,
      "loss": 0.4858388900756836,
      "step": 9200
    },
    {
      "epoch": 0.6131824234354194,
      "grad_norm": 0.9169468428513387,
      "learning_rate": 2e-05,
      "loss": 0.48983259201049806,
      "step": 9210
    },
    {
      "epoch": 0.6138482023968043,
      "grad_norm": 0.8044715671034054,
      "learning_rate": 2e-05,
      "loss": 0.48929810523986816,
      "step": 9220
    },
    {
      "epoch": 0.6145139813581891,
      "grad_norm": 1.0297207170453084,
      "learning_rate": 2e-05,
      "loss": 0.4921379089355469,
      "step": 9230
    },
    {
      "epoch": 0.6151797603195739,
      "grad_norm": 0.8037811160585597,
      "learning_rate": 2e-05,
      "loss": 0.49055066108703616,
      "step": 9240
    },
    {
      "epoch": 0.6158455392809588,
      "grad_norm": 1.165747051594704,
      "learning_rate": 2e-05,
      "loss": 0.48770503997802733,
      "step": 9250
    },
    {
      "epoch": 0.6165113182423435,
      "grad_norm": 1.0501769781565717,
      "learning_rate": 2e-05,
      "loss": 0.46318635940551756,
      "step": 9260
    },
    {
      "epoch": 0.6171770972037284,
      "grad_norm": 5.250182609909293,
      "learning_rate": 2e-05,
      "loss": 0.48468685150146484,
      "step": 9270
    },
    {
      "epoch": 0.6178428761651131,
      "grad_norm": 0.7912253205261757,
      "learning_rate": 2e-05,
      "loss": 0.487294340133667,
      "step": 9280
    },
    {
      "epoch": 0.618508655126498,
      "grad_norm": 0.7095838464170107,
      "learning_rate": 2e-05,
      "loss": 0.487591552734375,
      "step": 9290
    },
    {
      "epoch": 0.6191744340878829,
      "grad_norm": 0.7952486495115798,
      "learning_rate": 2e-05,
      "loss": 0.49062275886535645,
      "step": 9300
    },
    {
      "epoch": 0.6198402130492676,
      "grad_norm": 0.9030345914206299,
      "learning_rate": 2e-05,
      "loss": 0.48517804145812987,
      "step": 9310
    },
    {
      "epoch": 0.6205059920106525,
      "grad_norm": 0.9781946188870019,
      "learning_rate": 2e-05,
      "loss": 0.48894314765930175,
      "step": 9320
    },
    {
      "epoch": 0.6211717709720372,
      "grad_norm": 1.5392418164657125,
      "learning_rate": 2e-05,
      "loss": 0.4934073448181152,
      "step": 9330
    },
    {
      "epoch": 0.6218375499334221,
      "grad_norm": 0.8667863232470085,
      "learning_rate": 2e-05,
      "loss": 0.5014958381652832,
      "step": 9340
    },
    {
      "epoch": 0.622503328894807,
      "grad_norm": 0.9161171372411799,
      "learning_rate": 2e-05,
      "loss": 0.5040398597717285,
      "step": 9350
    },
    {
      "epoch": 0.6231691078561917,
      "grad_norm": 0.9482618191307345,
      "learning_rate": 2e-05,
      "loss": 0.4622523784637451,
      "step": 9360
    },
    {
      "epoch": 0.6238348868175766,
      "grad_norm": 1.4781929952275674,
      "learning_rate": 2e-05,
      "loss": 0.47809767723083496,
      "step": 9370
    },
    {
      "epoch": 0.6245006657789614,
      "grad_norm": 0.7513647940422048,
      "learning_rate": 2e-05,
      "loss": 0.47103567123413087,
      "step": 9380
    },
    {
      "epoch": 0.6251664447403462,
      "grad_norm": 0.979194111528927,
      "learning_rate": 2e-05,
      "loss": 0.4828365325927734,
      "step": 9390
    },
    {
      "epoch": 0.625832223701731,
      "grad_norm": 1.177541690698747,
      "learning_rate": 2e-05,
      "loss": 0.4829904556274414,
      "step": 9400
    },
    {
      "epoch": 0.6264980026631158,
      "grad_norm": 1.0124039730308856,
      "learning_rate": 2e-05,
      "loss": 0.4666259765625,
      "step": 9410
    },
    {
      "epoch": 0.6271637816245007,
      "grad_norm": 0.801598243707843,
      "learning_rate": 2e-05,
      "loss": 0.48631134033203127,
      "step": 9420
    },
    {
      "epoch": 0.6278295605858855,
      "grad_norm": 0.7438336747527146,
      "learning_rate": 2e-05,
      "loss": 0.4872234344482422,
      "step": 9430
    },
    {
      "epoch": 0.6284953395472703,
      "grad_norm": 0.9595603516444775,
      "learning_rate": 2e-05,
      "loss": 0.4646885395050049,
      "step": 9440
    },
    {
      "epoch": 0.6291611185086551,
      "grad_norm": 0.5966813367267466,
      "learning_rate": 2e-05,
      "loss": 0.4537227153778076,
      "step": 9450
    },
    {
      "epoch": 0.62982689747004,
      "grad_norm": 1.0075606013749805,
      "learning_rate": 2e-05,
      "loss": 0.4962646007537842,
      "step": 9460
    },
    {
      "epoch": 0.6304926764314248,
      "grad_norm": 0.7161683701296329,
      "learning_rate": 2e-05,
      "loss": 0.4832143306732178,
      "step": 9470
    },
    {
      "epoch": 0.6311584553928096,
      "grad_norm": 1.0254611704992567,
      "learning_rate": 2e-05,
      "loss": 0.4976398468017578,
      "step": 9480
    },
    {
      "epoch": 0.6318242343541944,
      "grad_norm": 0.7646248692915562,
      "learning_rate": 2e-05,
      "loss": 0.46971492767333983,
      "step": 9490
    },
    {
      "epoch": 0.6324900133155792,
      "grad_norm": 0.7757154307207489,
      "learning_rate": 2e-05,
      "loss": 0.4705049514770508,
      "step": 9500
    },
    {
      "epoch": 0.6331557922769641,
      "grad_norm": 0.8243441424244377,
      "learning_rate": 2e-05,
      "loss": 0.5105357646942139,
      "step": 9510
    },
    {
      "epoch": 0.6338215712383488,
      "grad_norm": 0.9946512396072108,
      "learning_rate": 2e-05,
      "loss": 0.4937573432922363,
      "step": 9520
    },
    {
      "epoch": 0.6344873501997337,
      "grad_norm": 1.2584616903009944,
      "learning_rate": 2e-05,
      "loss": 0.4904512882232666,
      "step": 9530
    },
    {
      "epoch": 0.6351531291611185,
      "grad_norm": 1.3698556040618295,
      "learning_rate": 2e-05,
      "loss": 0.44481625556945803,
      "step": 9540
    },
    {
      "epoch": 0.6358189081225033,
      "grad_norm": 0.889463833869457,
      "learning_rate": 2e-05,
      "loss": 0.48227267265319823,
      "step": 9550
    },
    {
      "epoch": 0.6364846870838882,
      "grad_norm": 0.8033515569718702,
      "learning_rate": 2e-05,
      "loss": 0.4772143840789795,
      "step": 9560
    },
    {
      "epoch": 0.6371504660452729,
      "grad_norm": 0.7758175015138523,
      "learning_rate": 2e-05,
      "loss": 0.49745640754699705,
      "step": 9570
    },
    {
      "epoch": 0.6378162450066578,
      "grad_norm": 0.6675777486335106,
      "learning_rate": 2e-05,
      "loss": 0.5016908168792724,
      "step": 9580
    },
    {
      "epoch": 0.6384820239680427,
      "grad_norm": 1.0916430517000364,
      "learning_rate": 2e-05,
      "loss": 0.47296772003173826,
      "step": 9590
    },
    {
      "epoch": 0.6391478029294274,
      "grad_norm": 0.7965177137100148,
      "learning_rate": 2e-05,
      "loss": 0.46767592430114746,
      "step": 9600
    },
    {
      "epoch": 0.6398135818908123,
      "grad_norm": 1.1268996383533525,
      "learning_rate": 2e-05,
      "loss": 0.45755720138549805,
      "step": 9610
    },
    {
      "epoch": 0.640479360852197,
      "grad_norm": 0.6958775968344562,
      "learning_rate": 2e-05,
      "loss": 0.45668625831604004,
      "step": 9620
    },
    {
      "epoch": 0.6411451398135819,
      "grad_norm": 1.0845909355060506,
      "learning_rate": 2e-05,
      "loss": 0.48232641220092776,
      "step": 9630
    },
    {
      "epoch": 0.6418109187749668,
      "grad_norm": 0.7744834037183739,
      "learning_rate": 2e-05,
      "loss": 0.45946483612060546,
      "step": 9640
    },
    {
      "epoch": 0.6424766977363515,
      "grad_norm": 0.7470718368844216,
      "learning_rate": 2e-05,
      "loss": 0.4772624492645264,
      "step": 9650
    },
    {
      "epoch": 0.6431424766977364,
      "grad_norm": 1.0547851957853716,
      "learning_rate": 2e-05,
      "loss": 0.4673303127288818,
      "step": 9660
    },
    {
      "epoch": 0.6438082556591211,
      "grad_norm": 0.9454162294630392,
      "learning_rate": 2e-05,
      "loss": 0.5032262802124023,
      "step": 9670
    },
    {
      "epoch": 0.644474034620506,
      "grad_norm": 0.7547437414303226,
      "learning_rate": 2e-05,
      "loss": 0.45669021606445315,
      "step": 9680
    },
    {
      "epoch": 0.6451398135818908,
      "grad_norm": 0.8693459518983041,
      "learning_rate": 2e-05,
      "loss": 0.4844752311706543,
      "step": 9690
    },
    {
      "epoch": 0.6458055925432756,
      "grad_norm": 0.8868481290244419,
      "learning_rate": 2e-05,
      "loss": 0.4743020057678223,
      "step": 9700
    },
    {
      "epoch": 0.6464713715046605,
      "grad_norm": 1.1522790659267501,
      "learning_rate": 2e-05,
      "loss": 0.5071447372436524,
      "step": 9710
    },
    {
      "epoch": 0.6471371504660453,
      "grad_norm": 0.7613299270091233,
      "learning_rate": 2e-05,
      "loss": 0.48708667755126955,
      "step": 9720
    },
    {
      "epoch": 0.6478029294274301,
      "grad_norm": 0.7708580993249599,
      "learning_rate": 2e-05,
      "loss": 0.49457292556762694,
      "step": 9730
    },
    {
      "epoch": 0.6484687083888149,
      "grad_norm": 0.858967687519222,
      "learning_rate": 2e-05,
      "loss": 0.45243682861328127,
      "step": 9740
    },
    {
      "epoch": 0.6491344873501997,
      "grad_norm": 1.3452034994383124,
      "learning_rate": 2e-05,
      "loss": 0.4672114372253418,
      "step": 9750
    },
    {
      "epoch": 0.6498002663115846,
      "grad_norm": 0.6939577038731806,
      "learning_rate": 2e-05,
      "loss": 0.4688560485839844,
      "step": 9760
    },
    {
      "epoch": 0.6504660452729694,
      "grad_norm": 0.7778113769763999,
      "learning_rate": 2e-05,
      "loss": 0.4393949508666992,
      "step": 9770
    },
    {
      "epoch": 0.6511318242343542,
      "grad_norm": 0.9810136607738305,
      "learning_rate": 2e-05,
      "loss": 0.4802661895751953,
      "step": 9780
    },
    {
      "epoch": 0.651797603195739,
      "grad_norm": 1.0984856620279801,
      "learning_rate": 2e-05,
      "loss": 0.4913919925689697,
      "step": 9790
    },
    {
      "epoch": 0.6524633821571239,
      "grad_norm": 0.6931454381859076,
      "learning_rate": 2e-05,
      "loss": 0.4843465328216553,
      "step": 9800
    },
    {
      "epoch": 0.6531291611185086,
      "grad_norm": 0.7612343278363043,
      "learning_rate": 2e-05,
      "loss": 0.45316119194030763,
      "step": 9810
    },
    {
      "epoch": 0.6537949400798935,
      "grad_norm": 0.8519864659413243,
      "learning_rate": 2e-05,
      "loss": 0.5139319896697998,
      "step": 9820
    },
    {
      "epoch": 0.6544607190412783,
      "grad_norm": 1.0195890278574065,
      "learning_rate": 2e-05,
      "loss": 0.48363456726074217,
      "step": 9830
    },
    {
      "epoch": 0.6551264980026631,
      "grad_norm": 0.7097928866969417,
      "learning_rate": 2e-05,
      "loss": 0.448469352722168,
      "step": 9840
    },
    {
      "epoch": 0.655792276964048,
      "grad_norm": 0.6968403269564801,
      "learning_rate": 2e-05,
      "loss": 0.48908109664916993,
      "step": 9850
    },
    {
      "epoch": 0.6564580559254327,
      "grad_norm": 0.6809630680379472,
      "learning_rate": 2e-05,
      "loss": 0.4721801280975342,
      "step": 9860
    },
    {
      "epoch": 0.6571238348868176,
      "grad_norm": 1.0041400673775855,
      "learning_rate": 2e-05,
      "loss": 0.47836127281188967,
      "step": 9870
    },
    {
      "epoch": 0.6577896138482024,
      "grad_norm": 0.8154978025387766,
      "learning_rate": 2e-05,
      "loss": 0.47530488967895507,
      "step": 9880
    },
    {
      "epoch": 0.6584553928095872,
      "grad_norm": 0.6505200966200821,
      "learning_rate": 2e-05,
      "loss": 0.46505041122436525,
      "step": 9890
    },
    {
      "epoch": 0.6591211717709721,
      "grad_norm": 0.7247440998093667,
      "learning_rate": 2e-05,
      "loss": 0.4454700469970703,
      "step": 9900
    },
    {
      "epoch": 0.6597869507323568,
      "grad_norm": 0.6086771091408818,
      "learning_rate": 2e-05,
      "loss": 0.4784823417663574,
      "step": 9910
    },
    {
      "epoch": 0.6604527296937417,
      "grad_norm": 6.62019833843309,
      "learning_rate": 2e-05,
      "loss": 0.48889665603637694,
      "step": 9920
    },
    {
      "epoch": 0.6611185086551264,
      "grad_norm": 4.964373982365685,
      "learning_rate": 2e-05,
      "loss": 0.46590538024902345,
      "step": 9930
    },
    {
      "epoch": 0.6617842876165113,
      "grad_norm": 0.779333010385721,
      "learning_rate": 2e-05,
      "loss": 0.4830604553222656,
      "step": 9940
    },
    {
      "epoch": 0.6624500665778962,
      "grad_norm": 0.693012268982187,
      "learning_rate": 2e-05,
      "loss": 0.5141133785247802,
      "step": 9950
    },
    {
      "epoch": 0.6631158455392809,
      "grad_norm": 0.8000662094332319,
      "learning_rate": 2e-05,
      "loss": 0.471113920211792,
      "step": 9960
    },
    {
      "epoch": 0.6637816245006658,
      "grad_norm": 0.8872089561413331,
      "learning_rate": 2e-05,
      "loss": 0.47097229957580566,
      "step": 9970
    },
    {
      "epoch": 0.6644474034620506,
      "grad_norm": 0.7217954673946949,
      "learning_rate": 2e-05,
      "loss": 0.4358534812927246,
      "step": 9980
    },
    {
      "epoch": 0.6651131824234354,
      "grad_norm": 0.7134805611336171,
      "learning_rate": 2e-05,
      "loss": 0.4974659442901611,
      "step": 9990
    },
    {
      "epoch": 0.6657789613848203,
      "grad_norm": 0.6276739079728207,
      "learning_rate": 2e-05,
      "loss": 0.467667293548584,
      "step": 10000
    },
    {
      "epoch": 0.666444740346205,
      "grad_norm": 0.826834927393318,
      "learning_rate": 2e-05,
      "loss": 0.4490972518920898,
      "step": 10010
    },
    {
      "epoch": 0.6671105193075899,
      "grad_norm": 0.7498052649269377,
      "learning_rate": 2e-05,
      "loss": 0.47895336151123047,
      "step": 10020
    },
    {
      "epoch": 0.6677762982689747,
      "grad_norm": 0.8773099318719606,
      "learning_rate": 2e-05,
      "loss": 0.48343348503112793,
      "step": 10030
    },
    {
      "epoch": 0.6684420772303595,
      "grad_norm": 0.6760815139776695,
      "learning_rate": 2e-05,
      "loss": 0.4524684429168701,
      "step": 10040
    },
    {
      "epoch": 0.6691078561917443,
      "grad_norm": 0.8654811873003859,
      "learning_rate": 2e-05,
      "loss": 0.4666222095489502,
      "step": 10050
    },
    {
      "epoch": 0.6697736351531292,
      "grad_norm": 1.215307159652945,
      "learning_rate": 2e-05,
      "loss": 0.4701251983642578,
      "step": 10060
    },
    {
      "epoch": 0.670439414114514,
      "grad_norm": 0.77244522032638,
      "learning_rate": 2e-05,
      "loss": 0.4524833679199219,
      "step": 10070
    },
    {
      "epoch": 0.6711051930758988,
      "grad_norm": 0.7043959605384333,
      "learning_rate": 2e-05,
      "loss": 0.44318299293518065,
      "step": 10080
    },
    {
      "epoch": 0.6717709720372836,
      "grad_norm": 0.6533875886178754,
      "learning_rate": 2e-05,
      "loss": 0.4938386917114258,
      "step": 10090
    },
    {
      "epoch": 0.6724367509986684,
      "grad_norm": 0.592771626675693,
      "learning_rate": 2e-05,
      "loss": 0.46810016632080076,
      "step": 10100
    },
    {
      "epoch": 0.6731025299600533,
      "grad_norm": 0.7772811199034018,
      "learning_rate": 2e-05,
      "loss": 0.48443264961242677,
      "step": 10110
    },
    {
      "epoch": 0.6737683089214381,
      "grad_norm": 0.6803574175887839,
      "learning_rate": 2e-05,
      "loss": 0.4779493808746338,
      "step": 10120
    },
    {
      "epoch": 0.6744340878828229,
      "grad_norm": 0.8820074714155928,
      "learning_rate": 2e-05,
      "loss": 0.4595679759979248,
      "step": 10130
    },
    {
      "epoch": 0.6750998668442078,
      "grad_norm": 0.6136217995607606,
      "learning_rate": 2e-05,
      "loss": 0.46224546432495117,
      "step": 10140
    },
    {
      "epoch": 0.6757656458055925,
      "grad_norm": 0.5906435665451198,
      "learning_rate": 2e-05,
      "loss": 0.4798443794250488,
      "step": 10150
    },
    {
      "epoch": 0.6764314247669774,
      "grad_norm": 0.7382245127304095,
      "learning_rate": 2e-05,
      "loss": 0.4696209907531738,
      "step": 10160
    },
    {
      "epoch": 0.6770972037283621,
      "grad_norm": 0.7063976546779468,
      "learning_rate": 2e-05,
      "loss": 0.4784710884094238,
      "step": 10170
    },
    {
      "epoch": 0.677762982689747,
      "grad_norm": 0.8546931795982535,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.47204333543777466,
      "step": 10180
    },
    {
      "epoch": 0.6784287616511319,
      "grad_norm": 0.7599622943177448,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.46112399101257323,
      "step": 10190
    },
    {
      "epoch": 0.6790945406125166,
      "grad_norm": 0.7988027940413726,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.4842371940612793,
      "step": 10200
    },
    {
      "epoch": 0.6797603195739015,
      "grad_norm": 0.7263018668999435,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.46210379600524903,
      "step": 10210
    },
    {
      "epoch": 0.6804260985352862,
      "grad_norm": 0.8248393608316446,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.45780439376831056,
      "step": 10220
    },
    {
      "epoch": 0.6810918774966711,
      "grad_norm": 0.6550312384503294,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.5122216224670411,
      "step": 10230
    },
    {
      "epoch": 0.681757656458056,
      "grad_norm": 0.7295014348264358,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.4734334945678711,
      "step": 10240
    },
    {
      "epoch": 0.6824234354194407,
      "grad_norm": 0.7147275003842296,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.46173667907714844,
      "step": 10250
    },
    {
      "epoch": 0.6830892143808256,
      "grad_norm": 0.5763245108309344,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.4650113105773926,
      "step": 10260
    },
    {
      "epoch": 0.6837549933422103,
      "grad_norm": 0.7623675066479597,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.48537521362304686,
      "step": 10270
    },
    {
      "epoch": 0.6844207723035952,
      "grad_norm": 0.792562740511735,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.48282480239868164,
      "step": 10280
    },
    {
      "epoch": 0.6850865512649801,
      "grad_norm": 0.638839280249028,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.4757516860961914,
      "step": 10290
    },
    {
      "epoch": 0.6857523302263648,
      "grad_norm": 0.6883456971398473,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.45363125801086424,
      "step": 10300
    },
    {
      "epoch": 0.6864181091877497,
      "grad_norm": 0.6781478297702987,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.4642007350921631,
      "step": 10310
    },
    {
      "epoch": 0.6870838881491345,
      "grad_norm": 0.6977409906661634,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.4549248695373535,
      "step": 10320
    },
    {
      "epoch": 0.6877496671105193,
      "grad_norm": 0.6717013510512636,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.48225154876708987,
      "step": 10330
    },
    {
      "epoch": 0.6884154460719041,
      "grad_norm": 0.7647934690845998,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.4632759094238281,
      "step": 10340
    },
    {
      "epoch": 0.689081225033289,
      "grad_norm": 0.6465167168930893,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.4517676830291748,
      "step": 10350
    },
    {
      "epoch": 0.6897470039946738,
      "grad_norm": 0.7573287819661285,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.48365178108215334,
      "step": 10360
    },
    {
      "epoch": 0.6904127829560586,
      "grad_norm": 0.6314867928319673,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.4771847724914551,
      "step": 10370
    },
    {
      "epoch": 0.6910785619174434,
      "grad_norm": 0.7325534786100285,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.4804633140563965,
      "step": 10380
    },
    {
      "epoch": 0.6917443408788282,
      "grad_norm": 0.6124979893509979,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.4473861217498779,
      "step": 10390
    },
    {
      "epoch": 0.6924101198402131,
      "grad_norm": 0.6523657589569143,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.44477219581604005,
      "step": 10400
    },
    {
      "epoch": 0.6930758988015979,
      "grad_norm": 0.693616682233738,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.45774664878845217,
      "step": 10410
    },
    {
      "epoch": 0.6937416777629827,
      "grad_norm": 1.4165541537511248,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.45206193923950194,
      "step": 10420
    },
    {
      "epoch": 0.6944074567243675,
      "grad_norm": 0.6991600377351948,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.48028993606567383,
      "step": 10430
    },
    {
      "epoch": 0.6950732356857523,
      "grad_norm": 0.6869371843370404,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.4837343215942383,
      "step": 10440
    },
    {
      "epoch": 0.6957390146471372,
      "grad_norm": 0.5953078081694219,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.4593203067779541,
      "step": 10450
    },
    {
      "epoch": 0.6964047936085219,
      "grad_norm": 0.6157314384663657,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.47696514129638673,
      "step": 10460
    },
    {
      "epoch": 0.6970705725699068,
      "grad_norm": 0.8689423170362172,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.46161303520202634,
      "step": 10470
    },
    {
      "epoch": 0.6977363515312917,
      "grad_norm": 0.7509273329734193,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.4898796081542969,
      "step": 10480
    },
    {
      "epoch": 0.6984021304926764,
      "grad_norm": 0.5433917506911256,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.4730968475341797,
      "step": 10490
    },
    {
      "epoch": 0.6990679094540613,
      "grad_norm": 0.6691239959972042,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.4527570724487305,
      "step": 10500
    },
    {
      "epoch": 0.699733688415446,
      "grad_norm": 0.9322670964241587,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.4872427463531494,
      "step": 10510
    },
    {
      "epoch": 0.7003994673768309,
      "grad_norm": 0.6997293391272841,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.45914716720581056,
      "step": 10520
    },
    {
      "epoch": 0.7010652463382158,
      "grad_norm": 0.7791803647754992,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.45444669723510744,
      "step": 10530
    },
    {
      "epoch": 0.7017310252996005,
      "grad_norm": 0.7505914008850666,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.48410654067993164,
      "step": 10540
    },
    {
      "epoch": 0.7023968042609854,
      "grad_norm": 0.67693664164961,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.4766782283782959,
      "step": 10550
    },
    {
      "epoch": 0.7030625832223701,
      "grad_norm": 0.8059224379335344,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.47568612098693847,
      "step": 10560
    },
    {
      "epoch": 0.703728362183755,
      "grad_norm": 0.6420069190722317,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.45975723266601565,
      "step": 10570
    },
    {
      "epoch": 0.7043941411451398,
      "grad_norm": 0.6698479706847191,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.44035725593566893,
      "step": 10580
    },
    {
      "epoch": 0.7050599201065246,
      "grad_norm": 1.8426565429761042,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.4532874584197998,
      "step": 10590
    },
    {
      "epoch": 0.7057256990679095,
      "grad_norm": 0.7452890564560531,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.46527652740478515,
      "step": 10600
    },
    {
      "epoch": 0.7063914780292943,
      "grad_norm": 2.498691275864618,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.48431010246276857,
      "step": 10610
    },
    {
      "epoch": 0.7070572569906791,
      "grad_norm": 0.9267620424488591,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.47551236152648924,
      "step": 10620
    },
    {
      "epoch": 0.7077230359520639,
      "grad_norm": 0.7042516820852219,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.47756366729736327,
      "step": 10630
    },
    {
      "epoch": 0.7083888149134487,
      "grad_norm": 1.0296781340816235,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.4508479595184326,
      "step": 10640
    },
    {
      "epoch": 0.7090545938748336,
      "grad_norm": 2.335976430860986,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.4804405689239502,
      "step": 10650
    },
    {
      "epoch": 0.7097203728362184,
      "grad_norm": 0.6564919816617072,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.44637236595153806,
      "step": 10660
    },
    {
      "epoch": 0.7103861517976032,
      "grad_norm": 0.7702724565449062,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.45778779983520507,
      "step": 10670
    },
    {
      "epoch": 0.711051930758988,
      "grad_norm": 0.6549586191901234,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.4582659721374512,
      "step": 10680
    },
    {
      "epoch": 0.7117177097203728,
      "grad_norm": 0.6723082044515459,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.4466773509979248,
      "step": 10690
    },
    {
      "epoch": 0.7123834886817576,
      "grad_norm": 0.6628572427359807,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.4975440502166748,
      "step": 10700
    },
    {
      "epoch": 0.7130492676431425,
      "grad_norm": 0.6410327585395696,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.4414191246032715,
      "step": 10710
    },
    {
      "epoch": 0.7137150466045273,
      "grad_norm": 0.817169099552439,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.4847556114196777,
      "step": 10720
    },
    {
      "epoch": 0.7143808255659121,
      "grad_norm": 2.087495609564876,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.4480710029602051,
      "step": 10730
    },
    {
      "epoch": 0.715046604527297,
      "grad_norm": 0.6639045257641183,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.44744415283203126,
      "step": 10740
    },
    {
      "epoch": 0.7157123834886817,
      "grad_norm": 0.9162297496842278,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.4681581497192383,
      "step": 10750
    },
    {
      "epoch": 0.7163781624500666,
      "grad_norm": 0.7340824585638108,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.431569766998291,
      "step": 10760
    },
    {
      "epoch": 0.7170439414114514,
      "grad_norm": 0.6912859281368265,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.453917407989502,
      "step": 10770
    },
    {
      "epoch": 0.7177097203728362,
      "grad_norm": 0.7418369909185265,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.4552594184875488,
      "step": 10780
    },
    {
      "epoch": 0.7183754993342211,
      "grad_norm": 0.674790769350001,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.4463171482086182,
      "step": 10790
    },
    {
      "epoch": 0.7190412782956058,
      "grad_norm": 0.6831337612140544,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.45091848373413085,
      "step": 10800
    },
    {
      "epoch": 0.7197070572569907,
      "grad_norm": 1.233149405014556,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.457490348815918,
      "step": 10810
    },
    {
      "epoch": 0.7203728362183754,
      "grad_norm": 0.8559388263392679,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.44842047691345216,
      "step": 10820
    },
    {
      "epoch": 0.7210386151797603,
      "grad_norm": 1.306825608252663,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.4467334270477295,
      "step": 10830
    },
    {
      "epoch": 0.7217043941411452,
      "grad_norm": 0.6471545815858233,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.4559581756591797,
      "step": 10840
    },
    {
      "epoch": 0.7223701731025299,
      "grad_norm": 0.7472762146049096,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.4747025489807129,
      "step": 10850
    },
    {
      "epoch": 0.7230359520639148,
      "grad_norm": 0.8598633210990939,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.4557167053222656,
      "step": 10860
    },
    {
      "epoch": 0.7237017310252996,
      "grad_norm": 0.7545052653551135,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.4564221382141113,
      "step": 10870
    },
    {
      "epoch": 0.7243675099866844,
      "grad_norm": 1.8955459764264084,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.46774921417236326,
      "step": 10880
    },
    {
      "epoch": 0.7250332889480693,
      "grad_norm": 1.002391226403099,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.46192026138305664,
      "step": 10890
    },
    {
      "epoch": 0.725699067909454,
      "grad_norm": 1.2705595902988907,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.4081739902496338,
      "step": 10900
    },
    {
      "epoch": 0.7263648468708389,
      "grad_norm": 0.6349738550386484,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.4479313850402832,
      "step": 10910
    },
    {
      "epoch": 0.7270306258322237,
      "grad_norm": 0.9237925236673558,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.47368502616882324,
      "step": 10920
    },
    {
      "epoch": 0.7276964047936085,
      "grad_norm": 0.6744095097303442,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.40831432342529295,
      "step": 10930
    },
    {
      "epoch": 0.7283621837549934,
      "grad_norm": 0.5589303293849958,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.4400953769683838,
      "step": 10940
    },
    {
      "epoch": 0.7290279627163782,
      "grad_norm": 0.7703937750261066,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.4725142478942871,
      "step": 10950
    },
    {
      "epoch": 0.729693741677763,
      "grad_norm": 0.696563510039128,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.44761247634887696,
      "step": 10960
    },
    {
      "epoch": 0.7303595206391478,
      "grad_norm": 1.2023226303187329,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.4471686840057373,
      "step": 10970
    },
    {
      "epoch": 0.7310252996005326,
      "grad_norm": 1.215051862813975,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.4541342735290527,
      "step": 10980
    },
    {
      "epoch": 0.7316910785619174,
      "grad_norm": 0.8687147542814291,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.43675765991210935,
      "step": 10990
    },
    {
      "epoch": 0.7323568575233023,
      "grad_norm": 0.7780048662295824,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.45372533798217773,
      "step": 11000
    },
    {
      "epoch": 0.7330226364846871,
      "grad_norm": 0.7532250822272248,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.4386638641357422,
      "step": 11010
    },
    {
      "epoch": 0.7336884154460719,
      "grad_norm": 1.1883658285842893,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.40180211067199706,
      "step": 11020
    },
    {
      "epoch": 0.7343541944074568,
      "grad_norm": 0.7857234026605496,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.43234801292419434,
      "step": 11030
    },
    {
      "epoch": 0.7350199733688415,
      "grad_norm": 1.0711971782813519,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.47655425071716306,
      "step": 11040
    },
    {
      "epoch": 0.7356857523302264,
      "grad_norm": 0.7287541019652507,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.4324789047241211,
      "step": 11050
    },
    {
      "epoch": 0.7363515312916112,
      "grad_norm": 0.7544605510035809,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.45863890647888184,
      "step": 11060
    },
    {
      "epoch": 0.737017310252996,
      "grad_norm": 0.6837892960838515,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.4200753688812256,
      "step": 11070
    },
    {
      "epoch": 0.7376830892143809,
      "grad_norm": 1.7596585220286869,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.4633835792541504,
      "step": 11080
    },
    {
      "epoch": 0.7383488681757656,
      "grad_norm": 0.7471454468188036,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.45568141937255857,
      "step": 11090
    },
    {
      "epoch": 0.7390146471371505,
      "grad_norm": 0.7228749405775631,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.4319135665893555,
      "step": 11100
    },
    {
      "epoch": 0.7396804260985352,
      "grad_norm": 0.7985716176544165,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.4638084411621094,
      "step": 11110
    },
    {
      "epoch": 0.7403462050599201,
      "grad_norm": 0.8431930304521831,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.4475558280944824,
      "step": 11120
    },
    {
      "epoch": 0.741011984021305,
      "grad_norm": 1.0135010995067715,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.424169921875,
      "step": 11130
    },
    {
      "epoch": 0.7416777629826897,
      "grad_norm": 1.0175150386502556,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.4249323844909668,
      "step": 11140
    },
    {
      "epoch": 0.7423435419440746,
      "grad_norm": 0.8039649191189346,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.43068456649780273,
      "step": 11150
    },
    {
      "epoch": 0.7430093209054593,
      "grad_norm": 0.6558293921299998,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.4265101909637451,
      "step": 11160
    },
    {
      "epoch": 0.7436750998668442,
      "grad_norm": 0.9062797723135472,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.4652695655822754,
      "step": 11170
    },
    {
      "epoch": 0.7443408788282291,
      "grad_norm": 0.7430930765004248,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.459283447265625,
      "step": 11180
    },
    {
      "epoch": 0.7450066577896138,
      "grad_norm": 0.8536761800290265,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.41311111450195315,
      "step": 11190
    },
    {
      "epoch": 0.7452063914780293,
      "step": 11193,
      "total_flos": 5300110976090112.0,
      "train_loss": 0.04163894166173622,
      "train_runtime": 14725.4524,
      "train_samples_per_second": 9.121,
      "train_steps_per_second": 0.76
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "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": 5300110976090112.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}