{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.40127616862632637,
  "eval_steps": 500,
  "global_step": 5597,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007169486664754804,
      "grad_norm": 315.4236659535465,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 1.848151397705078,
      "step": 10
    },
    {
      "epoch": 0.0014338973329509608,
      "grad_norm": 672.973252055138,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 1.1725639343261718,
      "step": 20
    },
    {
      "epoch": 0.002150845999426441,
      "grad_norm": 126.65993818302222,
      "learning_rate": 3.342134831460675e-06,
      "loss": 0.7593138694763184,
      "step": 30
    },
    {
      "epoch": 0.0028677946659019216,
      "grad_norm": 29.886751429599133,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.6618130683898926,
      "step": 40
    },
    {
      "epoch": 0.0035847433323774017,
      "grad_norm": 137.5005248660405,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.6227120399475098,
      "step": 50
    },
    {
      "epoch": 0.004301691998852882,
      "grad_norm": 38.139287491655594,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.5859609127044678,
      "step": 60
    },
    {
      "epoch": 0.005018640665328362,
      "grad_norm": 51.02157400174971,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.6018378734588623,
      "step": 70
    },
    {
      "epoch": 0.005735589331803843,
      "grad_norm": 24.465251435228527,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.5624160289764404,
      "step": 80
    },
    {
      "epoch": 0.006452537998279323,
      "grad_norm": 22.828409412609748,
      "learning_rate": 1.005e-05,
      "loss": 0.5475245475769043,
      "step": 90
    },
    {
      "epoch": 0.007169486664754803,
      "grad_norm": 14.786662427088277,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.5528249263763427,
      "step": 100
    },
    {
      "epoch": 0.007886435331230283,
      "grad_norm": 75.92842280330609,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.5614291667938233,
      "step": 110
    },
    {
      "epoch": 0.008603383997705764,
      "grad_norm": 20.360896855493966,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.6078104972839355,
      "step": 120
    },
    {
      "epoch": 0.009320332664181245,
      "grad_norm": 31.43963052094272,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.5890892028808594,
      "step": 130
    },
    {
      "epoch": 0.010037281330656725,
      "grad_norm": 54.35895463776518,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.6670564174652099,
      "step": 140
    },
    {
      "epoch": 0.010754229997132205,
      "grad_norm": 33.964610072430865,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.7157962322235107,
      "step": 150
    },
    {
      "epoch": 0.011471178663607686,
      "grad_norm": 110.65639964008429,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.7455081462860107,
      "step": 160
    },
    {
      "epoch": 0.012188127330083166,
      "grad_norm": 41.868609627942014,
      "learning_rate": 1.89938202247191e-05,
      "loss": 1.1094837188720703,
      "step": 170
    },
    {
      "epoch": 0.012905075996558647,
      "grad_norm": 10.702193812872004,
      "learning_rate": 2e-05,
      "loss": 0.6436423778533935,
      "step": 180
    },
    {
      "epoch": 0.013622024663034128,
      "grad_norm": 51.71265251185425,
      "learning_rate": 2e-05,
      "loss": 0.6243439674377441,
      "step": 190
    },
    {
      "epoch": 0.014338973329509607,
      "grad_norm": 154.8177418260545,
      "learning_rate": 2e-05,
      "loss": 0.610740852355957,
      "step": 200
    },
    {
      "epoch": 0.015055921995985088,
      "grad_norm": 52.282960752120125,
      "learning_rate": 2e-05,
      "loss": 0.6687478065490723,
      "step": 210
    },
    {
      "epoch": 0.015772870662460567,
      "grad_norm": 3.8483090765907435,
      "learning_rate": 2e-05,
      "loss": 0.579859733581543,
      "step": 220
    },
    {
      "epoch": 0.016489819328936048,
      "grad_norm": 3.6109561986511833,
      "learning_rate": 2e-05,
      "loss": 0.6074420928955078,
      "step": 230
    },
    {
      "epoch": 0.01720676799541153,
      "grad_norm": 5.053213094247748,
      "learning_rate": 2e-05,
      "loss": 0.5664905071258545,
      "step": 240
    },
    {
      "epoch": 0.01792371666188701,
      "grad_norm": 3.0138514562209093,
      "learning_rate": 2e-05,
      "loss": 0.5948857307434082,
      "step": 250
    },
    {
      "epoch": 0.01864066532836249,
      "grad_norm": 13.449705648790312,
      "learning_rate": 2e-05,
      "loss": 0.6020577907562256,
      "step": 260
    },
    {
      "epoch": 0.019357613994837968,
      "grad_norm": 3.6227046108363377,
      "learning_rate": 2e-05,
      "loss": 0.5548747062683106,
      "step": 270
    },
    {
      "epoch": 0.02007456266131345,
      "grad_norm": 4.112840464071606,
      "learning_rate": 2e-05,
      "loss": 0.5805762767791748,
      "step": 280
    },
    {
      "epoch": 0.02079151132778893,
      "grad_norm": 1.9209785496382135,
      "learning_rate": 2e-05,
      "loss": 0.5842366218566895,
      "step": 290
    },
    {
      "epoch": 0.02150845999426441,
      "grad_norm": 3.0684228704996004,
      "learning_rate": 2e-05,
      "loss": 0.5580601692199707,
      "step": 300
    },
    {
      "epoch": 0.022225408660739892,
      "grad_norm": 1.1909889223759265,
      "learning_rate": 2e-05,
      "loss": 0.5615853786468505,
      "step": 310
    },
    {
      "epoch": 0.022942357327215373,
      "grad_norm": 2.644802472025187,
      "learning_rate": 2e-05,
      "loss": 0.5611689567565918,
      "step": 320
    },
    {
      "epoch": 0.02365930599369085,
      "grad_norm": 20.263309516988937,
      "learning_rate": 2e-05,
      "loss": 0.5905077934265137,
      "step": 330
    },
    {
      "epoch": 0.02437625466016633,
      "grad_norm": 5.118749842105238,
      "learning_rate": 2e-05,
      "loss": 0.5447088241577148,
      "step": 340
    },
    {
      "epoch": 0.025093203326641812,
      "grad_norm": 14.83085444344932,
      "learning_rate": 2e-05,
      "loss": 0.5638551235198974,
      "step": 350
    },
    {
      "epoch": 0.025810151993117293,
      "grad_norm": 32.69304087592208,
      "learning_rate": 2e-05,
      "loss": 0.6271306037902832,
      "step": 360
    },
    {
      "epoch": 0.026527100659592774,
      "grad_norm": 7.062265311502185,
      "learning_rate": 2e-05,
      "loss": 0.5702651023864747,
      "step": 370
    },
    {
      "epoch": 0.027244049326068255,
      "grad_norm": 30.12218728774149,
      "learning_rate": 2e-05,
      "loss": 0.6142964839935303,
      "step": 380
    },
    {
      "epoch": 0.027960997992543733,
      "grad_norm": 4.161661997042373,
      "learning_rate": 2e-05,
      "loss": 0.6302419662475586,
      "step": 390
    },
    {
      "epoch": 0.028677946659019213,
      "grad_norm": 3.3921348737529113,
      "learning_rate": 2e-05,
      "loss": 0.5987069129943847,
      "step": 400
    },
    {
      "epoch": 0.029394895325494694,
      "grad_norm": 1.4838082708559646,
      "learning_rate": 2e-05,
      "loss": 0.602195930480957,
      "step": 410
    },
    {
      "epoch": 0.030111843991970175,
      "grad_norm": 3.3128177340785627,
      "learning_rate": 2e-05,
      "loss": 0.5845789909362793,
      "step": 420
    },
    {
      "epoch": 0.030828792658445656,
      "grad_norm": 2.1866628727610182,
      "learning_rate": 2e-05,
      "loss": 0.5644162654876709,
      "step": 430
    },
    {
      "epoch": 0.031545741324921134,
      "grad_norm": 11.884198120767302,
      "learning_rate": 2e-05,
      "loss": 0.5629150390625,
      "step": 440
    },
    {
      "epoch": 0.032262689991396615,
      "grad_norm": 18.16899260160586,
      "learning_rate": 2e-05,
      "loss": 0.5555632591247559,
      "step": 450
    },
    {
      "epoch": 0.032979638657872096,
      "grad_norm": 5.462581933210383,
      "learning_rate": 2e-05,
      "loss": 0.5991993904113769,
      "step": 460
    },
    {
      "epoch": 0.03369658732434758,
      "grad_norm": 3.045413742803253,
      "learning_rate": 2e-05,
      "loss": 0.5700560569763183,
      "step": 470
    },
    {
      "epoch": 0.03441353599082306,
      "grad_norm": 4.246757908127417,
      "learning_rate": 2e-05,
      "loss": 0.5917448997497559,
      "step": 480
    },
    {
      "epoch": 0.03513048465729854,
      "grad_norm": 5.424046887719749,
      "learning_rate": 2e-05,
      "loss": 0.5816346645355225,
      "step": 490
    },
    {
      "epoch": 0.03584743332377402,
      "grad_norm": 6.583800218516907,
      "learning_rate": 2e-05,
      "loss": 0.5913246154785157,
      "step": 500
    },
    {
      "epoch": 0.0365643819902495,
      "grad_norm": 3.482147587454769,
      "learning_rate": 2e-05,
      "loss": 0.5650829315185547,
      "step": 510
    },
    {
      "epoch": 0.03728133065672498,
      "grad_norm": 2.808982864585292,
      "learning_rate": 2e-05,
      "loss": 0.5790875911712646,
      "step": 520
    },
    {
      "epoch": 0.03799827932320046,
      "grad_norm": 4.388524538222509,
      "learning_rate": 2e-05,
      "loss": 0.5646647453308106,
      "step": 530
    },
    {
      "epoch": 0.038715227989675936,
      "grad_norm": 3.929514313642325,
      "learning_rate": 2e-05,
      "loss": 0.5568113327026367,
      "step": 540
    },
    {
      "epoch": 0.03943217665615142,
      "grad_norm": 1.928176676276484,
      "learning_rate": 2e-05,
      "loss": 0.5727216720581054,
      "step": 550
    },
    {
      "epoch": 0.0401491253226269,
      "grad_norm": 6.68335008196921,
      "learning_rate": 2e-05,
      "loss": 0.5850999355316162,
      "step": 560
    },
    {
      "epoch": 0.04086607398910238,
      "grad_norm": 15.837152933980025,
      "learning_rate": 2e-05,
      "loss": 0.5926525115966796,
      "step": 570
    },
    {
      "epoch": 0.04158302265557786,
      "grad_norm": 3.6114569877104423,
      "learning_rate": 2e-05,
      "loss": 0.5768462657928467,
      "step": 580
    },
    {
      "epoch": 0.04229997132205334,
      "grad_norm": 2.164511077764764,
      "learning_rate": 2e-05,
      "loss": 0.567426061630249,
      "step": 590
    },
    {
      "epoch": 0.04301691998852882,
      "grad_norm": 2.948019819021623,
      "learning_rate": 2e-05,
      "loss": 0.5795512199401855,
      "step": 600
    },
    {
      "epoch": 0.0437338686550043,
      "grad_norm": 39.2052027751844,
      "learning_rate": 2e-05,
      "loss": 0.5852450370788574,
      "step": 610
    },
    {
      "epoch": 0.044450817321479784,
      "grad_norm": 11.115023206584201,
      "learning_rate": 2e-05,
      "loss": 0.5830046653747558,
      "step": 620
    },
    {
      "epoch": 0.045167765987955265,
      "grad_norm": 11.130215580191274,
      "learning_rate": 2e-05,
      "loss": 0.55075101852417,
      "step": 630
    },
    {
      "epoch": 0.045884714654430746,
      "grad_norm": 2.2401242450660765,
      "learning_rate": 2e-05,
      "loss": 0.5412608623504639,
      "step": 640
    },
    {
      "epoch": 0.04660166332090622,
      "grad_norm": 25.8476496640342,
      "learning_rate": 2e-05,
      "loss": 0.5914183616638183,
      "step": 650
    },
    {
      "epoch": 0.0473186119873817,
      "grad_norm": 3.955944237962116,
      "learning_rate": 2e-05,
      "loss": 0.5842460632324219,
      "step": 660
    },
    {
      "epoch": 0.04803556065385718,
      "grad_norm": 2.509395729576886,
      "learning_rate": 2e-05,
      "loss": 0.5712038993835449,
      "step": 670
    },
    {
      "epoch": 0.04875250932033266,
      "grad_norm": 2.4532942736851084,
      "learning_rate": 2e-05,
      "loss": 0.5865258693695068,
      "step": 680
    },
    {
      "epoch": 0.049469457986808144,
      "grad_norm": 3.6555036797810754,
      "learning_rate": 2e-05,
      "loss": 0.5788607597351074,
      "step": 690
    },
    {
      "epoch": 0.050186406653283624,
      "grad_norm": 5.581489874452689,
      "learning_rate": 2e-05,
      "loss": 0.5549694061279297,
      "step": 700
    },
    {
      "epoch": 0.050903355319759105,
      "grad_norm": 7.448830832646055,
      "learning_rate": 2e-05,
      "loss": 0.5698171138763428,
      "step": 710
    },
    {
      "epoch": 0.051620303986234586,
      "grad_norm": 3.6608686716559906,
      "learning_rate": 2e-05,
      "loss": 0.554749584197998,
      "step": 720
    },
    {
      "epoch": 0.05233725265271007,
      "grad_norm": 4.307177970225383,
      "learning_rate": 2e-05,
      "loss": 0.5440076351165771,
      "step": 730
    },
    {
      "epoch": 0.05305420131918555,
      "grad_norm": 4.418311679242279,
      "learning_rate": 2e-05,
      "loss": 0.5399693489074707,
      "step": 740
    },
    {
      "epoch": 0.05377114998566103,
      "grad_norm": 5.044806271145424,
      "learning_rate": 2e-05,
      "loss": 0.5785567283630371,
      "step": 750
    },
    {
      "epoch": 0.05448809865213651,
      "grad_norm": 3.4544327147592595,
      "learning_rate": 2e-05,
      "loss": 0.6100111961364746,
      "step": 760
    },
    {
      "epoch": 0.055205047318611984,
      "grad_norm": 2.9036817715442393,
      "learning_rate": 2e-05,
      "loss": 0.5828133583068847,
      "step": 770
    },
    {
      "epoch": 0.055921995985087465,
      "grad_norm": 1.7522351082267351,
      "learning_rate": 2e-05,
      "loss": 0.5638906478881835,
      "step": 780
    },
    {
      "epoch": 0.056638944651562946,
      "grad_norm": 2.209946265416314,
      "learning_rate": 2e-05,
      "loss": 0.590320348739624,
      "step": 790
    },
    {
      "epoch": 0.05735589331803843,
      "grad_norm": 2.6684001731598195,
      "learning_rate": 2e-05,
      "loss": 0.5646738052368164,
      "step": 800
    },
    {
      "epoch": 0.05807284198451391,
      "grad_norm": 5.223266428619974,
      "learning_rate": 2e-05,
      "loss": 0.5637637138366699,
      "step": 810
    },
    {
      "epoch": 0.05878979065098939,
      "grad_norm": 1.9142769139433,
      "learning_rate": 2e-05,
      "loss": 0.5934517860412598,
      "step": 820
    },
    {
      "epoch": 0.05950673931746487,
      "grad_norm": 2.6149464203468504,
      "learning_rate": 2e-05,
      "loss": 0.5700265884399414,
      "step": 830
    },
    {
      "epoch": 0.06022368798394035,
      "grad_norm": 3.96105780832147,
      "learning_rate": 2e-05,
      "loss": 0.5675205230712891,
      "step": 840
    },
    {
      "epoch": 0.06094063665041583,
      "grad_norm": 2.5400423051210974,
      "learning_rate": 2e-05,
      "loss": 0.5908157348632812,
      "step": 850
    },
    {
      "epoch": 0.06165758531689131,
      "grad_norm": 2.6344860048789127,
      "learning_rate": 2e-05,
      "loss": 0.5718094825744628,
      "step": 860
    },
    {
      "epoch": 0.062374533983366794,
      "grad_norm": 2.600050881708375,
      "learning_rate": 2e-05,
      "loss": 0.5780810832977294,
      "step": 870
    },
    {
      "epoch": 0.06309148264984227,
      "grad_norm": 3.134985069799007,
      "learning_rate": 2e-05,
      "loss": 0.567197322845459,
      "step": 880
    },
    {
      "epoch": 0.06380843131631775,
      "grad_norm": 2.0285438198733443,
      "learning_rate": 2e-05,
      "loss": 0.5852311611175537,
      "step": 890
    },
    {
      "epoch": 0.06452537998279323,
      "grad_norm": 5.248736255854688,
      "learning_rate": 2e-05,
      "loss": 0.5891732215881348,
      "step": 900
    },
    {
      "epoch": 0.06524232864926871,
      "grad_norm": 1.8727144020099094,
      "learning_rate": 2e-05,
      "loss": 0.5862767696380615,
      "step": 910
    },
    {
      "epoch": 0.06595927731574419,
      "grad_norm": 1.711765938123259,
      "learning_rate": 2e-05,
      "loss": 0.6182688713073731,
      "step": 920
    },
    {
      "epoch": 0.06667622598221967,
      "grad_norm": 1.763371319742271,
      "learning_rate": 2e-05,
      "loss": 0.5868721008300781,
      "step": 930
    },
    {
      "epoch": 0.06739317464869515,
      "grad_norm": 1.2813315472585687,
      "learning_rate": 2e-05,
      "loss": 0.5589268684387207,
      "step": 940
    },
    {
      "epoch": 0.06811012331517063,
      "grad_norm": 2.751663887567051,
      "learning_rate": 2e-05,
      "loss": 0.5824700355529785,
      "step": 950
    },
    {
      "epoch": 0.06882707198164612,
      "grad_norm": 4.993093109327952,
      "learning_rate": 2e-05,
      "loss": 0.5773461818695068,
      "step": 960
    },
    {
      "epoch": 0.0695440206481216,
      "grad_norm": 6.264445826159941,
      "learning_rate": 2e-05,
      "loss": 0.5533229827880859,
      "step": 970
    },
    {
      "epoch": 0.07026096931459708,
      "grad_norm": 3.2290134324615365,
      "learning_rate": 2e-05,
      "loss": 0.5897500514984131,
      "step": 980
    },
    {
      "epoch": 0.07097791798107256,
      "grad_norm": 3.0414609504263934,
      "learning_rate": 2e-05,
      "loss": 0.5628997802734375,
      "step": 990
    },
    {
      "epoch": 0.07169486664754804,
      "grad_norm": 2.0651674928177504,
      "learning_rate": 2e-05,
      "loss": 0.5647363662719727,
      "step": 1000
    },
    {
      "epoch": 0.07241181531402352,
      "grad_norm": 2.325752137633854,
      "learning_rate": 2e-05,
      "loss": 0.574032211303711,
      "step": 1010
    },
    {
      "epoch": 0.073128763980499,
      "grad_norm": 2.1263355903408008,
      "learning_rate": 2e-05,
      "loss": 0.5692534446716309,
      "step": 1020
    },
    {
      "epoch": 0.07384571264697448,
      "grad_norm": 24.27180244689631,
      "learning_rate": 2e-05,
      "loss": 0.5639516353607178,
      "step": 1030
    },
    {
      "epoch": 0.07456266131344996,
      "grad_norm": 1.0790226331900736,
      "learning_rate": 2e-05,
      "loss": 0.5968218803405761,
      "step": 1040
    },
    {
      "epoch": 0.07527960997992544,
      "grad_norm": 1.0623420342326966,
      "learning_rate": 2e-05,
      "loss": 0.5545104026794434,
      "step": 1050
    },
    {
      "epoch": 0.07599655864640092,
      "grad_norm": 2.342778745540576,
      "learning_rate": 2e-05,
      "loss": 0.5392826080322266,
      "step": 1060
    },
    {
      "epoch": 0.07671350731287639,
      "grad_norm": 2.003241443878693,
      "learning_rate": 2e-05,
      "loss": 0.5342343807220459,
      "step": 1070
    },
    {
      "epoch": 0.07743045597935187,
      "grad_norm": 2.027419316411222,
      "learning_rate": 2e-05,
      "loss": 0.5778005123138428,
      "step": 1080
    },
    {
      "epoch": 0.07814740464582735,
      "grad_norm": 1.896028667711285,
      "learning_rate": 2e-05,
      "loss": 0.5735631465911866,
      "step": 1090
    },
    {
      "epoch": 0.07886435331230283,
      "grad_norm": 2.0470674801420214,
      "learning_rate": 2e-05,
      "loss": 0.580590009689331,
      "step": 1100
    },
    {
      "epoch": 0.07958130197877832,
      "grad_norm": 2.7415867683239825,
      "learning_rate": 2e-05,
      "loss": 0.5596016883850098,
      "step": 1110
    },
    {
      "epoch": 0.0802982506452538,
      "grad_norm": 1.6531033173482321,
      "learning_rate": 2e-05,
      "loss": 0.5827084541320801,
      "step": 1120
    },
    {
      "epoch": 0.08101519931172928,
      "grad_norm": 1.2823997995083143,
      "learning_rate": 2e-05,
      "loss": 0.5864274024963378,
      "step": 1130
    },
    {
      "epoch": 0.08173214797820476,
      "grad_norm": 1.3642460710168849,
      "learning_rate": 2e-05,
      "loss": 0.5656659603118896,
      "step": 1140
    },
    {
      "epoch": 0.08244909664468024,
      "grad_norm": 2.6194064364228318,
      "learning_rate": 2e-05,
      "loss": 0.5763952255249023,
      "step": 1150
    },
    {
      "epoch": 0.08316604531115572,
      "grad_norm": 1.2781451995694622,
      "learning_rate": 2e-05,
      "loss": 0.5566621780395508,
      "step": 1160
    },
    {
      "epoch": 0.0838829939776312,
      "grad_norm": 1.5589065178330486,
      "learning_rate": 2e-05,
      "loss": 0.53390474319458,
      "step": 1170
    },
    {
      "epoch": 0.08459994264410668,
      "grad_norm": 7.459041859593309,
      "learning_rate": 2e-05,
      "loss": 0.5653298854827881,
      "step": 1180
    },
    {
      "epoch": 0.08531689131058216,
      "grad_norm": 2.9904694002471066,
      "learning_rate": 2e-05,
      "loss": 0.560931396484375,
      "step": 1190
    },
    {
      "epoch": 0.08603383997705764,
      "grad_norm": 1.5979927096992035,
      "learning_rate": 2e-05,
      "loss": 0.530519676208496,
      "step": 1200
    },
    {
      "epoch": 0.08675078864353312,
      "grad_norm": 8.353200078910842,
      "learning_rate": 2e-05,
      "loss": 0.5533581733703613,
      "step": 1210
    },
    {
      "epoch": 0.0874677373100086,
      "grad_norm": 37.21010749725826,
      "learning_rate": 2e-05,
      "loss": 0.5901515007019043,
      "step": 1220
    },
    {
      "epoch": 0.08818468597648409,
      "grad_norm": 5.122367920805441,
      "learning_rate": 2e-05,
      "loss": 0.5856198310852051,
      "step": 1230
    },
    {
      "epoch": 0.08890163464295957,
      "grad_norm": 3.3489073569661105,
      "learning_rate": 2e-05,
      "loss": 0.5682595729827881,
      "step": 1240
    },
    {
      "epoch": 0.08961858330943505,
      "grad_norm": 2.327849536892137,
      "learning_rate": 2e-05,
      "loss": 0.5584495067596436,
      "step": 1250
    },
    {
      "epoch": 0.09033553197591053,
      "grad_norm": 22.72509490275124,
      "learning_rate": 2e-05,
      "loss": 0.5762526512145996,
      "step": 1260
    },
    {
      "epoch": 0.09105248064238601,
      "grad_norm": 7.3344483954259285,
      "learning_rate": 2e-05,
      "loss": 0.5725028991699219,
      "step": 1270
    },
    {
      "epoch": 0.09176942930886149,
      "grad_norm": 3.9521562517230238,
      "learning_rate": 2e-05,
      "loss": 0.5834849357604981,
      "step": 1280
    },
    {
      "epoch": 0.09248637797533697,
      "grad_norm": 6.43496222827271,
      "learning_rate": 2e-05,
      "loss": 0.5840559005737305,
      "step": 1290
    },
    {
      "epoch": 0.09320332664181244,
      "grad_norm": 3.8438872068840104,
      "learning_rate": 2e-05,
      "loss": 0.6151006698608399,
      "step": 1300
    },
    {
      "epoch": 0.09392027530828792,
      "grad_norm": 1.4662403573164653,
      "learning_rate": 2e-05,
      "loss": 0.5667572021484375,
      "step": 1310
    },
    {
      "epoch": 0.0946372239747634,
      "grad_norm": 17.21523102720178,
      "learning_rate": 2e-05,
      "loss": 0.5740190505981445,
      "step": 1320
    },
    {
      "epoch": 0.09535417264123888,
      "grad_norm": 1.766413981347864,
      "learning_rate": 2e-05,
      "loss": 0.5626147270202637,
      "step": 1330
    },
    {
      "epoch": 0.09607112130771436,
      "grad_norm": 2.6800514021980004,
      "learning_rate": 2e-05,
      "loss": 0.5712190628051758,
      "step": 1340
    },
    {
      "epoch": 0.09678806997418984,
      "grad_norm": 2.961182111187328,
      "learning_rate": 2e-05,
      "loss": 0.5644628047943115,
      "step": 1350
    },
    {
      "epoch": 0.09750501864066533,
      "grad_norm": 2.454375644899256,
      "learning_rate": 2e-05,
      "loss": 0.5722562313079834,
      "step": 1360
    },
    {
      "epoch": 0.0982219673071408,
      "grad_norm": 1.6873399339759625,
      "learning_rate": 2e-05,
      "loss": 0.5544299125671387,
      "step": 1370
    },
    {
      "epoch": 0.09893891597361629,
      "grad_norm": 1.1555872203445525,
      "learning_rate": 2e-05,
      "loss": 0.5385549068450928,
      "step": 1380
    },
    {
      "epoch": 0.09965586464009177,
      "grad_norm": 1.5548337351700245,
      "learning_rate": 2e-05,
      "loss": 0.5589668273925781,
      "step": 1390
    },
    {
      "epoch": 0.10037281330656725,
      "grad_norm": 1.2505995954976357,
      "learning_rate": 2e-05,
      "loss": 0.57919921875,
      "step": 1400
    },
    {
      "epoch": 0.10108976197304273,
      "grad_norm": 5.206526378064766,
      "learning_rate": 2e-05,
      "loss": 0.5539649963378906,
      "step": 1410
    },
    {
      "epoch": 0.10180671063951821,
      "grad_norm": 2.610432174521263,
      "learning_rate": 2e-05,
      "loss": 0.592057180404663,
      "step": 1420
    },
    {
      "epoch": 0.10252365930599369,
      "grad_norm": 3.3133808987579276,
      "learning_rate": 2e-05,
      "loss": 0.5443053245544434,
      "step": 1430
    },
    {
      "epoch": 0.10324060797246917,
      "grad_norm": 2.0455832858183682,
      "learning_rate": 2e-05,
      "loss": 0.546113920211792,
      "step": 1440
    },
    {
      "epoch": 0.10395755663894465,
      "grad_norm": 1.7592817899937072,
      "learning_rate": 2e-05,
      "loss": 0.551348066329956,
      "step": 1450
    },
    {
      "epoch": 0.10467450530542013,
      "grad_norm": 2.943775668256852,
      "learning_rate": 2e-05,
      "loss": 0.5249824523925781,
      "step": 1460
    },
    {
      "epoch": 0.10539145397189562,
      "grad_norm": 2.140971602034373,
      "learning_rate": 2e-05,
      "loss": 0.5666528224945069,
      "step": 1470
    },
    {
      "epoch": 0.1061084026383711,
      "grad_norm": 4.331861493161884,
      "learning_rate": 2e-05,
      "loss": 0.5543347835540772,
      "step": 1480
    },
    {
      "epoch": 0.10682535130484658,
      "grad_norm": 6.353962200216883,
      "learning_rate": 2e-05,
      "loss": 0.5454977035522461,
      "step": 1490
    },
    {
      "epoch": 0.10754229997132206,
      "grad_norm": 41.865001318361585,
      "learning_rate": 2e-05,
      "loss": 0.5601842880249024,
      "step": 1500
    },
    {
      "epoch": 0.10825924863779754,
      "grad_norm": 1.7010891103793604,
      "learning_rate": 2e-05,
      "loss": 0.5699163436889648,
      "step": 1510
    },
    {
      "epoch": 0.10897619730427302,
      "grad_norm": 2.8044652038050373,
      "learning_rate": 2e-05,
      "loss": 0.5804720878601074,
      "step": 1520
    },
    {
      "epoch": 0.1096931459707485,
      "grad_norm": 2.8044499219399186,
      "learning_rate": 2e-05,
      "loss": 0.5642358779907226,
      "step": 1530
    },
    {
      "epoch": 0.11041009463722397,
      "grad_norm": 1.6404567243971169,
      "learning_rate": 2e-05,
      "loss": 0.547740125656128,
      "step": 1540
    },
    {
      "epoch": 0.11112704330369945,
      "grad_norm": 5.8842210253367595,
      "learning_rate": 2e-05,
      "loss": 0.5724470138549804,
      "step": 1550
    },
    {
      "epoch": 0.11184399197017493,
      "grad_norm": 2.0263887469590465,
      "learning_rate": 2e-05,
      "loss": 0.5523801803588867,
      "step": 1560
    },
    {
      "epoch": 0.11256094063665041,
      "grad_norm": 1.3512109762818472,
      "learning_rate": 2e-05,
      "loss": 0.5701728343963623,
      "step": 1570
    },
    {
      "epoch": 0.11327788930312589,
      "grad_norm": 2.961754694676749,
      "learning_rate": 2e-05,
      "loss": 0.5410641670227051,
      "step": 1580
    },
    {
      "epoch": 0.11399483796960137,
      "grad_norm": 9.38177531956118,
      "learning_rate": 2e-05,
      "loss": 0.5361968994140625,
      "step": 1590
    },
    {
      "epoch": 0.11471178663607685,
      "grad_norm": 3.1019434885984083,
      "learning_rate": 2e-05,
      "loss": 0.5702048301696777,
      "step": 1600
    },
    {
      "epoch": 0.11542873530255233,
      "grad_norm": 1.8141527074697572,
      "learning_rate": 2e-05,
      "loss": 0.5201643466949463,
      "step": 1610
    },
    {
      "epoch": 0.11614568396902782,
      "grad_norm": 2.81574948493408,
      "learning_rate": 2e-05,
      "loss": 0.5578659057617188,
      "step": 1620
    },
    {
      "epoch": 0.1168626326355033,
      "grad_norm": 5.40288732260851,
      "learning_rate": 2e-05,
      "loss": 0.5688968658447265,
      "step": 1630
    },
    {
      "epoch": 0.11757958130197878,
      "grad_norm": 2.2012274195157153,
      "learning_rate": 2e-05,
      "loss": 0.5558701515197754,
      "step": 1640
    },
    {
      "epoch": 0.11829652996845426,
      "grad_norm": 1.730570521948697,
      "learning_rate": 2e-05,
      "loss": 0.5515780448913574,
      "step": 1650
    },
    {
      "epoch": 0.11901347863492974,
      "grad_norm": 3.083940659262359,
      "learning_rate": 2e-05,
      "loss": 0.5680615901947021,
      "step": 1660
    },
    {
      "epoch": 0.11973042730140522,
      "grad_norm": 1.8293486250460052,
      "learning_rate": 2e-05,
      "loss": 0.5579127311706543,
      "step": 1670
    },
    {
      "epoch": 0.1204473759678807,
      "grad_norm": 2.691528194596175,
      "learning_rate": 2e-05,
      "loss": 0.5666030406951904,
      "step": 1680
    },
    {
      "epoch": 0.12116432463435618,
      "grad_norm": 2.3017548820591376,
      "learning_rate": 2e-05,
      "loss": 0.5542361259460449,
      "step": 1690
    },
    {
      "epoch": 0.12188127330083166,
      "grad_norm": 1.3982396451314711,
      "learning_rate": 2e-05,
      "loss": 0.5412776470184326,
      "step": 1700
    },
    {
      "epoch": 0.12259822196730714,
      "grad_norm": 2.8633211849346836,
      "learning_rate": 2e-05,
      "loss": 0.5596530914306641,
      "step": 1710
    },
    {
      "epoch": 0.12331517063378263,
      "grad_norm": 1.2756481909481026,
      "learning_rate": 2e-05,
      "loss": 0.5272120475769043,
      "step": 1720
    },
    {
      "epoch": 0.1240321193002581,
      "grad_norm": 1.884108459767808,
      "learning_rate": 2e-05,
      "loss": 0.5613333702087402,
      "step": 1730
    },
    {
      "epoch": 0.12474906796673359,
      "grad_norm": 5.273226308180893,
      "learning_rate": 2e-05,
      "loss": 0.5628758430480957,
      "step": 1740
    },
    {
      "epoch": 0.12546601663320905,
      "grad_norm": 3.0995531259526348,
      "learning_rate": 2e-05,
      "loss": 0.5465875148773194,
      "step": 1750
    },
    {
      "epoch": 0.12618296529968454,
      "grad_norm": 3.2626826548420476,
      "learning_rate": 2e-05,
      "loss": 0.5627561092376709,
      "step": 1760
    },
    {
      "epoch": 0.12689991396616002,
      "grad_norm": 2.364412227942818,
      "learning_rate": 2e-05,
      "loss": 0.5472114562988282,
      "step": 1770
    },
    {
      "epoch": 0.1276168626326355,
      "grad_norm": 2.4263037850994746,
      "learning_rate": 2e-05,
      "loss": 0.5574633598327636,
      "step": 1780
    },
    {
      "epoch": 0.12833381129911098,
      "grad_norm": 2.309082260471337,
      "learning_rate": 2e-05,
      "loss": 0.5340644836425781,
      "step": 1790
    },
    {
      "epoch": 0.12905075996558646,
      "grad_norm": 2.426509513022813,
      "learning_rate": 2e-05,
      "loss": 0.5799418926239014,
      "step": 1800
    },
    {
      "epoch": 0.12976770863206194,
      "grad_norm": 1.8704066062154519,
      "learning_rate": 2e-05,
      "loss": 0.5632325649261475,
      "step": 1810
    },
    {
      "epoch": 0.13048465729853742,
      "grad_norm": 1.406035471805342,
      "learning_rate": 2e-05,
      "loss": 0.5656003952026367,
      "step": 1820
    },
    {
      "epoch": 0.1312016059650129,
      "grad_norm": 1.544336567362451,
      "learning_rate": 2e-05,
      "loss": 0.5451941967010498,
      "step": 1830
    },
    {
      "epoch": 0.13191855463148838,
      "grad_norm": 2.0799351359110485,
      "learning_rate": 2e-05,
      "loss": 0.5621081352233886,
      "step": 1840
    },
    {
      "epoch": 0.13263550329796386,
      "grad_norm": 1.7299640932134708,
      "learning_rate": 2e-05,
      "loss": 0.5848733425140381,
      "step": 1850
    },
    {
      "epoch": 0.13335245196443934,
      "grad_norm": 1.1961375728783252,
      "learning_rate": 2e-05,
      "loss": 0.5659356117248535,
      "step": 1860
    },
    {
      "epoch": 0.13406940063091483,
      "grad_norm": 2.9075264586093783,
      "learning_rate": 2e-05,
      "loss": 0.5509964942932128,
      "step": 1870
    },
    {
      "epoch": 0.1347863492973903,
      "grad_norm": 1.3125243339991093,
      "learning_rate": 2e-05,
      "loss": 0.5528433799743653,
      "step": 1880
    },
    {
      "epoch": 0.1355032979638658,
      "grad_norm": 1.3727193915645148,
      "learning_rate": 2e-05,
      "loss": 0.564556884765625,
      "step": 1890
    },
    {
      "epoch": 0.13622024663034127,
      "grad_norm": 3.7953721625845933,
      "learning_rate": 2e-05,
      "loss": 0.5450845718383789,
      "step": 1900
    },
    {
      "epoch": 0.13693719529681675,
      "grad_norm": 0.9639735389876055,
      "learning_rate": 2e-05,
      "loss": 0.5541686534881591,
      "step": 1910
    },
    {
      "epoch": 0.13765414396329223,
      "grad_norm": 1.003569244043582,
      "learning_rate": 2e-05,
      "loss": 0.5625730514526367,
      "step": 1920
    },
    {
      "epoch": 0.1383710926297677,
      "grad_norm": 0.8880655840375095,
      "learning_rate": 2e-05,
      "loss": 0.5609973907470703,
      "step": 1930
    },
    {
      "epoch": 0.1390880412962432,
      "grad_norm": 1.8013979122929717,
      "learning_rate": 2e-05,
      "loss": 0.5645060539245605,
      "step": 1940
    },
    {
      "epoch": 0.13980498996271867,
      "grad_norm": 1.534112281642532,
      "learning_rate": 2e-05,
      "loss": 0.5428046226501465,
      "step": 1950
    },
    {
      "epoch": 0.14052193862919415,
      "grad_norm": 3.630130753040599,
      "learning_rate": 2e-05,
      "loss": 0.5524575233459472,
      "step": 1960
    },
    {
      "epoch": 0.14123888729566964,
      "grad_norm": 1.334941948733717,
      "learning_rate": 2e-05,
      "loss": 0.5768626213073731,
      "step": 1970
    },
    {
      "epoch": 0.14195583596214512,
      "grad_norm": 6.284844482295266,
      "learning_rate": 2e-05,
      "loss": 0.5756556987762451,
      "step": 1980
    },
    {
      "epoch": 0.1426727846286206,
      "grad_norm": 2.7603549710378648,
      "learning_rate": 2e-05,
      "loss": 0.5478507995605468,
      "step": 1990
    },
    {
      "epoch": 0.14338973329509608,
      "grad_norm": 2.711164302375169,
      "learning_rate": 2e-05,
      "loss": 0.5454816341400146,
      "step": 2000
    },
    {
      "epoch": 0.14410668196157156,
      "grad_norm": 1.6373091053111855,
      "learning_rate": 2e-05,
      "loss": 0.56094331741333,
      "step": 2010
    },
    {
      "epoch": 0.14482363062804704,
      "grad_norm": 1.3231963543618217,
      "learning_rate": 2e-05,
      "loss": 0.5709875106811524,
      "step": 2020
    },
    {
      "epoch": 0.14554057929452252,
      "grad_norm": 2.6013946026230177,
      "learning_rate": 2e-05,
      "loss": 0.5760745048522949,
      "step": 2030
    },
    {
      "epoch": 0.146257527960998,
      "grad_norm": 1.0760891963639665,
      "learning_rate": 2e-05,
      "loss": 0.540398359298706,
      "step": 2040
    },
    {
      "epoch": 0.14697447662747348,
      "grad_norm": 3.806990207982052,
      "learning_rate": 2e-05,
      "loss": 0.5452116966247559,
      "step": 2050
    },
    {
      "epoch": 0.14769142529394896,
      "grad_norm": 0.8770094294050765,
      "learning_rate": 2e-05,
      "loss": 0.516285514831543,
      "step": 2060
    },
    {
      "epoch": 0.14840837396042444,
      "grad_norm": 1.64509283352414,
      "learning_rate": 2e-05,
      "loss": 0.5785709381103515,
      "step": 2070
    },
    {
      "epoch": 0.14912532262689993,
      "grad_norm": 1.0501298033510167,
      "learning_rate": 2e-05,
      "loss": 0.5674317836761474,
      "step": 2080
    },
    {
      "epoch": 0.1498422712933754,
      "grad_norm": 1.0810242187793186,
      "learning_rate": 2e-05,
      "loss": 0.5573361873626709,
      "step": 2090
    },
    {
      "epoch": 0.1505592199598509,
      "grad_norm": 1.3235649158062135,
      "learning_rate": 2e-05,
      "loss": 0.5721589088439941,
      "step": 2100
    },
    {
      "epoch": 0.15127616862632637,
      "grad_norm": 2.61075713737253,
      "learning_rate": 2e-05,
      "loss": 0.5381807327270508,
      "step": 2110
    },
    {
      "epoch": 0.15199311729280185,
      "grad_norm": 2.0071819238060655,
      "learning_rate": 2e-05,
      "loss": 0.5428504943847656,
      "step": 2120
    },
    {
      "epoch": 0.1527100659592773,
      "grad_norm": 1.9874329432228202,
      "learning_rate": 2e-05,
      "loss": 0.5540834426879883,
      "step": 2130
    },
    {
      "epoch": 0.15342701462575278,
      "grad_norm": 1.0751634954880047,
      "learning_rate": 2e-05,
      "loss": 0.5777225494384766,
      "step": 2140
    },
    {
      "epoch": 0.15414396329222826,
      "grad_norm": 2.247684458013884,
      "learning_rate": 2e-05,
      "loss": 0.5361541748046875,
      "step": 2150
    },
    {
      "epoch": 0.15486091195870375,
      "grad_norm": 1.5069655343366213,
      "learning_rate": 2e-05,
      "loss": 0.5514012336730957,
      "step": 2160
    },
    {
      "epoch": 0.15557786062517923,
      "grad_norm": 1.4677351079218146,
      "learning_rate": 2e-05,
      "loss": 0.5583604335784912,
      "step": 2170
    },
    {
      "epoch": 0.1562948092916547,
      "grad_norm": 1.1533300901460213,
      "learning_rate": 2e-05,
      "loss": 0.5542792797088623,
      "step": 2180
    },
    {
      "epoch": 0.1570117579581302,
      "grad_norm": 1.7028225571829634,
      "learning_rate": 2e-05,
      "loss": 0.5386368751525878,
      "step": 2190
    },
    {
      "epoch": 0.15772870662460567,
      "grad_norm": 1.3309927566018185,
      "learning_rate": 2e-05,
      "loss": 0.5358123302459716,
      "step": 2200
    },
    {
      "epoch": 0.15844565529108115,
      "grad_norm": 8.920239112234068,
      "learning_rate": 2e-05,
      "loss": 0.5610211849212646,
      "step": 2210
    },
    {
      "epoch": 0.15916260395755663,
      "grad_norm": 0.9909451029281341,
      "learning_rate": 2e-05,
      "loss": 0.5076250076293946,
      "step": 2220
    },
    {
      "epoch": 0.1598795526240321,
      "grad_norm": 1.2993026276344535,
      "learning_rate": 2e-05,
      "loss": 0.540493392944336,
      "step": 2230
    },
    {
      "epoch": 0.1605965012905076,
      "grad_norm": 2.3512466519648587,
      "learning_rate": 2e-05,
      "loss": 0.5421253204345703,
      "step": 2240
    },
    {
      "epoch": 0.16131344995698307,
      "grad_norm": 3.215608826413565,
      "learning_rate": 2e-05,
      "loss": 0.5742346763610839,
      "step": 2250
    },
    {
      "epoch": 0.16203039862345855,
      "grad_norm": 2.1540313753980445,
      "learning_rate": 2e-05,
      "loss": 0.5484731674194336,
      "step": 2260
    },
    {
      "epoch": 0.16274734728993404,
      "grad_norm": 1.844563278680931,
      "learning_rate": 2e-05,
      "loss": 0.5414963722229004,
      "step": 2270
    },
    {
      "epoch": 0.16346429595640952,
      "grad_norm": 1.2178930369073198,
      "learning_rate": 2e-05,
      "loss": 0.5368363380432128,
      "step": 2280
    },
    {
      "epoch": 0.164181244622885,
      "grad_norm": 0.8834306050985979,
      "learning_rate": 2e-05,
      "loss": 0.5298836708068848,
      "step": 2290
    },
    {
      "epoch": 0.16489819328936048,
      "grad_norm": 1.5799606162459594,
      "learning_rate": 2e-05,
      "loss": 0.5445155143737793,
      "step": 2300
    },
    {
      "epoch": 0.16561514195583596,
      "grad_norm": 1.106839067943099,
      "learning_rate": 2e-05,
      "loss": 0.5452072143554687,
      "step": 2310
    },
    {
      "epoch": 0.16633209062231144,
      "grad_norm": 2.598513025278755,
      "learning_rate": 2e-05,
      "loss": 0.5509998321533203,
      "step": 2320
    },
    {
      "epoch": 0.16704903928878692,
      "grad_norm": 2.825348671573568,
      "learning_rate": 2e-05,
      "loss": 0.5631167411804199,
      "step": 2330
    },
    {
      "epoch": 0.1677659879552624,
      "grad_norm": 1.7145872571500615,
      "learning_rate": 2e-05,
      "loss": 0.5552741050720215,
      "step": 2340
    },
    {
      "epoch": 0.16848293662173788,
      "grad_norm": 14.03401035026617,
      "learning_rate": 2e-05,
      "loss": 0.5463265895843505,
      "step": 2350
    },
    {
      "epoch": 0.16919988528821336,
      "grad_norm": 2.0847220323880165,
      "learning_rate": 2e-05,
      "loss": 0.5252552986145019,
      "step": 2360
    },
    {
      "epoch": 0.16991683395468885,
      "grad_norm": 1.011338519703196,
      "learning_rate": 2e-05,
      "loss": 0.5438662528991699,
      "step": 2370
    },
    {
      "epoch": 0.17063378262116433,
      "grad_norm": 1.1683565008517496,
      "learning_rate": 2e-05,
      "loss": 0.5535166740417481,
      "step": 2380
    },
    {
      "epoch": 0.1713507312876398,
      "grad_norm": 5.08246929812119,
      "learning_rate": 2e-05,
      "loss": 0.5406145095825196,
      "step": 2390
    },
    {
      "epoch": 0.1720676799541153,
      "grad_norm": 1.088285394309193,
      "learning_rate": 2e-05,
      "loss": 0.5489974021911621,
      "step": 2400
    },
    {
      "epoch": 0.17278462862059077,
      "grad_norm": 0.7764690731022792,
      "learning_rate": 2e-05,
      "loss": 0.5712192058563232,
      "step": 2410
    },
    {
      "epoch": 0.17350157728706625,
      "grad_norm": 1.4063769617297237,
      "learning_rate": 2e-05,
      "loss": 0.5649506568908691,
      "step": 2420
    },
    {
      "epoch": 0.17421852595354173,
      "grad_norm": 0.8841185813513404,
      "learning_rate": 2e-05,
      "loss": 0.5460720062255859,
      "step": 2430
    },
    {
      "epoch": 0.1749354746200172,
      "grad_norm": 1.0771999567027115,
      "learning_rate": 2e-05,
      "loss": 0.568940544128418,
      "step": 2440
    },
    {
      "epoch": 0.1756524232864927,
      "grad_norm": 1.2155770616956478,
      "learning_rate": 2e-05,
      "loss": 0.5199066638946533,
      "step": 2450
    },
    {
      "epoch": 0.17636937195296817,
      "grad_norm": 1.677688151590649,
      "learning_rate": 2e-05,
      "loss": 0.556132459640503,
      "step": 2460
    },
    {
      "epoch": 0.17708632061944365,
      "grad_norm": 0.9244792122316684,
      "learning_rate": 2e-05,
      "loss": 0.5273470878601074,
      "step": 2470
    },
    {
      "epoch": 0.17780326928591914,
      "grad_norm": 4.24012694050476,
      "learning_rate": 2e-05,
      "loss": 0.5406831741333008,
      "step": 2480
    },
    {
      "epoch": 0.17852021795239462,
      "grad_norm": 0.9989434456238647,
      "learning_rate": 2e-05,
      "loss": 0.5287195682525635,
      "step": 2490
    },
    {
      "epoch": 0.1792371666188701,
      "grad_norm": 1.2294548845023252,
      "learning_rate": 2e-05,
      "loss": 0.5238923072814942,
      "step": 2500
    },
    {
      "epoch": 0.17995411528534558,
      "grad_norm": 1.1680245691471733,
      "learning_rate": 2e-05,
      "loss": 0.5360855102539063,
      "step": 2510
    },
    {
      "epoch": 0.18067106395182106,
      "grad_norm": 3.259902952148628,
      "learning_rate": 2e-05,
      "loss": 0.5570301055908203,
      "step": 2520
    },
    {
      "epoch": 0.18138801261829654,
      "grad_norm": 2.790265316658619,
      "learning_rate": 2e-05,
      "loss": 0.532185411453247,
      "step": 2530
    },
    {
      "epoch": 0.18210496128477202,
      "grad_norm": 1.2744671754626793,
      "learning_rate": 2e-05,
      "loss": 0.5266745567321778,
      "step": 2540
    },
    {
      "epoch": 0.1828219099512475,
      "grad_norm": 0.9695743661168896,
      "learning_rate": 2e-05,
      "loss": 0.5322426046643939,
      "step": 2550
    },
    {
      "epoch": 0.18353885861772298,
      "grad_norm": 1.975696273374771,
      "learning_rate": 2e-05,
      "loss": 0.5236278057098389,
      "step": 2560
    },
    {
      "epoch": 0.18425580728419846,
      "grad_norm": 1.3548366922913517,
      "learning_rate": 2e-05,
      "loss": 0.5559279441833496,
      "step": 2570
    },
    {
      "epoch": 0.18497275595067394,
      "grad_norm": 1.9747413551068616,
      "learning_rate": 2e-05,
      "loss": 0.5727641105651855,
      "step": 2580
    },
    {
      "epoch": 0.18568970461714943,
      "grad_norm": 0.906622028066299,
      "learning_rate": 2e-05,
      "loss": 0.531487512588501,
      "step": 2590
    },
    {
      "epoch": 0.18640665328362488,
      "grad_norm": 1.4879888618089294,
      "learning_rate": 2e-05,
      "loss": 0.5519673824310303,
      "step": 2600
    },
    {
      "epoch": 0.18712360195010036,
      "grad_norm": 1.254187226503217,
      "learning_rate": 2e-05,
      "loss": 0.565938138961792,
      "step": 2610
    },
    {
      "epoch": 0.18784055061657584,
      "grad_norm": 0.8793117177954728,
      "learning_rate": 2e-05,
      "loss": 0.5405161857604981,
      "step": 2620
    },
    {
      "epoch": 0.18855749928305132,
      "grad_norm": 1.9261302890906202,
      "learning_rate": 2e-05,
      "loss": 0.5784532070159912,
      "step": 2630
    },
    {
      "epoch": 0.1892744479495268,
      "grad_norm": 1.3022370866614288,
      "learning_rate": 2e-05,
      "loss": 0.5467851638793946,
      "step": 2640
    },
    {
      "epoch": 0.18999139661600228,
      "grad_norm": 10.70761513211328,
      "learning_rate": 2e-05,
      "loss": 0.5216636180877685,
      "step": 2650
    },
    {
      "epoch": 0.19070834528247776,
      "grad_norm": 1.720852701911088,
      "learning_rate": 2e-05,
      "loss": 0.5543615341186523,
      "step": 2660
    },
    {
      "epoch": 0.19142529394895325,
      "grad_norm": 1.5679154309102394,
      "learning_rate": 2e-05,
      "loss": 0.5322744369506835,
      "step": 2670
    },
    {
      "epoch": 0.19214224261542873,
      "grad_norm": 1.1577374495962562,
      "learning_rate": 2e-05,
      "loss": 0.5546241760253906,
      "step": 2680
    },
    {
      "epoch": 0.1928591912819042,
      "grad_norm": 2.078480326424917,
      "learning_rate": 2e-05,
      "loss": 0.5530313014984131,
      "step": 2690
    },
    {
      "epoch": 0.1935761399483797,
      "grad_norm": 1.718894223380425,
      "learning_rate": 2e-05,
      "loss": 0.5648358345031739,
      "step": 2700
    },
    {
      "epoch": 0.19429308861485517,
      "grad_norm": 4.36678086638391,
      "learning_rate": 2e-05,
      "loss": 0.5371812343597412,
      "step": 2710
    },
    {
      "epoch": 0.19501003728133065,
      "grad_norm": 1.986070398230652,
      "learning_rate": 2e-05,
      "loss": 0.5654563903808594,
      "step": 2720
    },
    {
      "epoch": 0.19572698594780613,
      "grad_norm": 1.083731261882624,
      "learning_rate": 2e-05,
      "loss": 0.5358358383178711,
      "step": 2730
    },
    {
      "epoch": 0.1964439346142816,
      "grad_norm": 3.523753168920136,
      "learning_rate": 2e-05,
      "loss": 0.5595098018646241,
      "step": 2740
    },
    {
      "epoch": 0.1971608832807571,
      "grad_norm": 1.4381686976503263,
      "learning_rate": 2e-05,
      "loss": 0.5390559196472168,
      "step": 2750
    },
    {
      "epoch": 0.19787783194723257,
      "grad_norm": 5.526157432332793,
      "learning_rate": 2e-05,
      "loss": 0.5542815208435059,
      "step": 2760
    },
    {
      "epoch": 0.19859478061370806,
      "grad_norm": 23.27422669047157,
      "learning_rate": 2e-05,
      "loss": 0.561483097076416,
      "step": 2770
    },
    {
      "epoch": 0.19931172928018354,
      "grad_norm": 38.96983775818154,
      "learning_rate": 2e-05,
      "loss": 0.5560777187347412,
      "step": 2780
    },
    {
      "epoch": 0.20002867794665902,
      "grad_norm": 11.667933873668167,
      "learning_rate": 2e-05,
      "loss": 0.5505549907684326,
      "step": 2790
    },
    {
      "epoch": 0.2007456266131345,
      "grad_norm": 2.956362296432195,
      "learning_rate": 2e-05,
      "loss": 0.562765645980835,
      "step": 2800
    },
    {
      "epoch": 0.20146257527960998,
      "grad_norm": 1.1719897457764303,
      "learning_rate": 2e-05,
      "loss": 0.5576314449310302,
      "step": 2810
    },
    {
      "epoch": 0.20217952394608546,
      "grad_norm": 4.082839867104697,
      "learning_rate": 2e-05,
      "loss": 0.5498549461364746,
      "step": 2820
    },
    {
      "epoch": 0.20289647261256094,
      "grad_norm": 1.7541756539952584,
      "learning_rate": 2e-05,
      "loss": 0.5313169479370117,
      "step": 2830
    },
    {
      "epoch": 0.20361342127903642,
      "grad_norm": 1.0974198282288676,
      "learning_rate": 2e-05,
      "loss": 0.5603452682495117,
      "step": 2840
    },
    {
      "epoch": 0.2043303699455119,
      "grad_norm": 1.2795235608523798,
      "learning_rate": 2e-05,
      "loss": 0.5323123455047607,
      "step": 2850
    },
    {
      "epoch": 0.20504731861198738,
      "grad_norm": 4.441527476440911,
      "learning_rate": 2e-05,
      "loss": 0.5210596084594726,
      "step": 2860
    },
    {
      "epoch": 0.20576426727846286,
      "grad_norm": 1.8996709872890165,
      "learning_rate": 2e-05,
      "loss": 0.5504693508148193,
      "step": 2870
    },
    {
      "epoch": 0.20648121594493835,
      "grad_norm": 1.5342075306767362,
      "learning_rate": 2e-05,
      "loss": 0.5548974990844726,
      "step": 2880
    },
    {
      "epoch": 0.20719816461141383,
      "grad_norm": 3.822518798218387,
      "learning_rate": 2e-05,
      "loss": 0.5298895835876465,
      "step": 2890
    },
    {
      "epoch": 0.2079151132778893,
      "grad_norm": 1.5660242014134182,
      "learning_rate": 2e-05,
      "loss": 0.5920438289642334,
      "step": 2900
    },
    {
      "epoch": 0.2086320619443648,
      "grad_norm": 1.734303117822142,
      "learning_rate": 2e-05,
      "loss": 0.5546172142028809,
      "step": 2910
    },
    {
      "epoch": 0.20934901061084027,
      "grad_norm": 0.9989875856357617,
      "learning_rate": 2e-05,
      "loss": 0.5633022308349609,
      "step": 2920
    },
    {
      "epoch": 0.21006595927731575,
      "grad_norm": 1.5732534502376367,
      "learning_rate": 2e-05,
      "loss": 0.5416759490966797,
      "step": 2930
    },
    {
      "epoch": 0.21078290794379123,
      "grad_norm": 23.507681546072988,
      "learning_rate": 2e-05,
      "loss": 0.528422737121582,
      "step": 2940
    },
    {
      "epoch": 0.2114998566102667,
      "grad_norm": 1.2017706772985,
      "learning_rate": 2e-05,
      "loss": 0.5437676906585693,
      "step": 2950
    },
    {
      "epoch": 0.2122168052767422,
      "grad_norm": 1.6041209991424712,
      "learning_rate": 2e-05,
      "loss": 0.5334288597106933,
      "step": 2960
    },
    {
      "epoch": 0.21293375394321767,
      "grad_norm": 1.0106595569520083,
      "learning_rate": 2e-05,
      "loss": 0.5421632289886474,
      "step": 2970
    },
    {
      "epoch": 0.21365070260969315,
      "grad_norm": 1.515518985224875,
      "learning_rate": 2e-05,
      "loss": 0.5262147903442382,
      "step": 2980
    },
    {
      "epoch": 0.21436765127616864,
      "grad_norm": 0.9320183070635978,
      "learning_rate": 2e-05,
      "loss": 0.5572122573852539,
      "step": 2990
    },
    {
      "epoch": 0.21508459994264412,
      "grad_norm": 1.681011210707191,
      "learning_rate": 2e-05,
      "loss": 0.5309662818908691,
      "step": 3000
    },
    {
      "epoch": 0.2158015486091196,
      "grad_norm": 1.492067630150501,
      "learning_rate": 2e-05,
      "loss": 0.5504055023193359,
      "step": 3010
    },
    {
      "epoch": 0.21651849727559508,
      "grad_norm": 1.7393942545798637,
      "learning_rate": 2e-05,
      "loss": 0.5454302787780761,
      "step": 3020
    },
    {
      "epoch": 0.21723544594207056,
      "grad_norm": 1.8233114646395159,
      "learning_rate": 2e-05,
      "loss": 0.5375478744506836,
      "step": 3030
    },
    {
      "epoch": 0.21795239460854604,
      "grad_norm": 1.583947446323942,
      "learning_rate": 2e-05,
      "loss": 0.5283527374267578,
      "step": 3040
    },
    {
      "epoch": 0.21866934327502152,
      "grad_norm": 1.085145001186251,
      "learning_rate": 2e-05,
      "loss": 0.5616912364959716,
      "step": 3050
    },
    {
      "epoch": 0.219386291941497,
      "grad_norm": 1.3612189500537382,
      "learning_rate": 2e-05,
      "loss": 0.5466063499450684,
      "step": 3060
    },
    {
      "epoch": 0.22010324060797246,
      "grad_norm": 1.2663612429251823,
      "learning_rate": 2e-05,
      "loss": 0.5183010101318359,
      "step": 3070
    },
    {
      "epoch": 0.22082018927444794,
      "grad_norm": 1.8445683557833539,
      "learning_rate": 2e-05,
      "loss": 0.5627210140228271,
      "step": 3080
    },
    {
      "epoch": 0.22153713794092342,
      "grad_norm": 0.9530696229399991,
      "learning_rate": 2e-05,
      "loss": 0.5286128044128418,
      "step": 3090
    },
    {
      "epoch": 0.2222540866073989,
      "grad_norm": 1.9275100831360488,
      "learning_rate": 2e-05,
      "loss": 0.5295572280883789,
      "step": 3100
    },
    {
      "epoch": 0.22297103527387438,
      "grad_norm": 1.8036902772708967,
      "learning_rate": 2e-05,
      "loss": 0.5241225719451904,
      "step": 3110
    },
    {
      "epoch": 0.22368798394034986,
      "grad_norm": 1.2752586023399894,
      "learning_rate": 2e-05,
      "loss": 0.5464012145996093,
      "step": 3120
    },
    {
      "epoch": 0.22440493260682534,
      "grad_norm": 1.5973383515727866,
      "learning_rate": 2e-05,
      "loss": 0.5185372352600097,
      "step": 3130
    },
    {
      "epoch": 0.22512188127330082,
      "grad_norm": 1.125494452349449,
      "learning_rate": 2e-05,
      "loss": 0.53779296875,
      "step": 3140
    },
    {
      "epoch": 0.2258388299397763,
      "grad_norm": 0.9803524218253822,
      "learning_rate": 2e-05,
      "loss": 0.5390335083007812,
      "step": 3150
    },
    {
      "epoch": 0.22655577860625178,
      "grad_norm": 1.1089529146650596,
      "learning_rate": 2e-05,
      "loss": 0.5619882106781006,
      "step": 3160
    },
    {
      "epoch": 0.22727272727272727,
      "grad_norm": 1.8397426876656962,
      "learning_rate": 2e-05,
      "loss": 0.5175239562988281,
      "step": 3170
    },
    {
      "epoch": 0.22798967593920275,
      "grad_norm": 25.280662553093666,
      "learning_rate": 2e-05,
      "loss": 0.5171899795532227,
      "step": 3180
    },
    {
      "epoch": 0.22870662460567823,
      "grad_norm": 1.7319517977417502,
      "learning_rate": 2e-05,
      "loss": 0.5366016387939453,
      "step": 3190
    },
    {
      "epoch": 0.2294235732721537,
      "grad_norm": 2.312012904945782,
      "learning_rate": 2e-05,
      "loss": 0.5698837280273438,
      "step": 3200
    },
    {
      "epoch": 0.2301405219386292,
      "grad_norm": 2.2350276894989465,
      "learning_rate": 2e-05,
      "loss": 0.5330371379852294,
      "step": 3210
    },
    {
      "epoch": 0.23085747060510467,
      "grad_norm": 1.9140928590362325,
      "learning_rate": 2e-05,
      "loss": 0.5491515636444092,
      "step": 3220
    },
    {
      "epoch": 0.23157441927158015,
      "grad_norm": 3.3710032513683714,
      "learning_rate": 2e-05,
      "loss": 0.5247351646423339,
      "step": 3230
    },
    {
      "epoch": 0.23229136793805563,
      "grad_norm": 2.370709462661893,
      "learning_rate": 2e-05,
      "loss": 0.5287737846374512,
      "step": 3240
    },
    {
      "epoch": 0.2330083166045311,
      "grad_norm": 1.488656649685699,
      "learning_rate": 2e-05,
      "loss": 0.5258587837219239,
      "step": 3250
    },
    {
      "epoch": 0.2337252652710066,
      "grad_norm": 11.99781492210604,
      "learning_rate": 2e-05,
      "loss": 0.5449223518371582,
      "step": 3260
    },
    {
      "epoch": 0.23444221393748207,
      "grad_norm": 1.5065644812354444,
      "learning_rate": 2e-05,
      "loss": 0.5603810787200928,
      "step": 3270
    },
    {
      "epoch": 0.23515916260395756,
      "grad_norm": 1.272746904290695,
      "learning_rate": 2e-05,
      "loss": 0.512965202331543,
      "step": 3280
    },
    {
      "epoch": 0.23587611127043304,
      "grad_norm": 2.1153209916343583,
      "learning_rate": 2e-05,
      "loss": 0.5705286979675293,
      "step": 3290
    },
    {
      "epoch": 0.23659305993690852,
      "grad_norm": 1.0644763614234334,
      "learning_rate": 2e-05,
      "loss": 0.5379728317260742,
      "step": 3300
    },
    {
      "epoch": 0.237310008603384,
      "grad_norm": 1.2223729176305438,
      "learning_rate": 2e-05,
      "loss": 0.5281596183776855,
      "step": 3310
    },
    {
      "epoch": 0.23802695726985948,
      "grad_norm": 3.0955217938333766,
      "learning_rate": 2e-05,
      "loss": 0.5670812606811524,
      "step": 3320
    },
    {
      "epoch": 0.23874390593633496,
      "grad_norm": 1.317719915367741,
      "learning_rate": 2e-05,
      "loss": 0.5240924835205079,
      "step": 3330
    },
    {
      "epoch": 0.23946085460281044,
      "grad_norm": 0.9220506420985892,
      "learning_rate": 2e-05,
      "loss": 0.5291359424591064,
      "step": 3340
    },
    {
      "epoch": 0.24017780326928592,
      "grad_norm": 0.8583333624738706,
      "learning_rate": 2e-05,
      "loss": 0.5238984107971192,
      "step": 3350
    },
    {
      "epoch": 0.2408947519357614,
      "grad_norm": 1.6162146762560434,
      "learning_rate": 2e-05,
      "loss": 0.5236201286315918,
      "step": 3360
    },
    {
      "epoch": 0.24161170060223688,
      "grad_norm": 43.79925933783028,
      "learning_rate": 2e-05,
      "loss": 0.5236298561096191,
      "step": 3370
    },
    {
      "epoch": 0.24232864926871236,
      "grad_norm": 1.2716483364395454,
      "learning_rate": 2e-05,
      "loss": 0.537201976776123,
      "step": 3380
    },
    {
      "epoch": 0.24304559793518785,
      "grad_norm": 5.149019849350892,
      "learning_rate": 2e-05,
      "loss": 0.5322999000549317,
      "step": 3390
    },
    {
      "epoch": 0.24376254660166333,
      "grad_norm": 1.8595415352393094,
      "learning_rate": 2e-05,
      "loss": 0.5268035888671875,
      "step": 3400
    },
    {
      "epoch": 0.2444794952681388,
      "grad_norm": 8.290708089453803,
      "learning_rate": 2e-05,
      "loss": 0.5156392097473145,
      "step": 3410
    },
    {
      "epoch": 0.2451964439346143,
      "grad_norm": 2.527019223050327,
      "learning_rate": 2e-05,
      "loss": 0.5337914943695068,
      "step": 3420
    },
    {
      "epoch": 0.24591339260108977,
      "grad_norm": 0.8127393510415952,
      "learning_rate": 2e-05,
      "loss": 0.5314960479736328,
      "step": 3430
    },
    {
      "epoch": 0.24663034126756525,
      "grad_norm": 0.8736699627908139,
      "learning_rate": 2e-05,
      "loss": 0.5364626884460449,
      "step": 3440
    },
    {
      "epoch": 0.24734728993404073,
      "grad_norm": 1.1013574667483945,
      "learning_rate": 2e-05,
      "loss": 0.5215117931365967,
      "step": 3450
    },
    {
      "epoch": 0.2480642386005162,
      "grad_norm": 7.72532619422504,
      "learning_rate": 2e-05,
      "loss": 0.5496770858764648,
      "step": 3460
    },
    {
      "epoch": 0.2487811872669917,
      "grad_norm": 1.1898055374205145,
      "learning_rate": 2e-05,
      "loss": 0.5156715393066407,
      "step": 3470
    },
    {
      "epoch": 0.24949813593346717,
      "grad_norm": 0.8275033836776856,
      "learning_rate": 2e-05,
      "loss": 0.5399666786193847,
      "step": 3480
    },
    {
      "epoch": 0.2502150845999426,
      "grad_norm": 2.853585482485147,
      "learning_rate": 2e-05,
      "loss": 0.498288631439209,
      "step": 3490
    },
    {
      "epoch": 0.2509320332664181,
      "grad_norm": 1.9618902397157088,
      "learning_rate": 2e-05,
      "loss": 0.501953125,
      "step": 3500
    },
    {
      "epoch": 0.2516489819328936,
      "grad_norm": 1.1910590165038817,
      "learning_rate": 2e-05,
      "loss": 0.531727933883667,
      "step": 3510
    },
    {
      "epoch": 0.25236593059936907,
      "grad_norm": 0.9271247866342498,
      "learning_rate": 2e-05,
      "loss": 0.5388535976409912,
      "step": 3520
    },
    {
      "epoch": 0.25308287926584455,
      "grad_norm": 1.19140294013902,
      "learning_rate": 2e-05,
      "loss": 0.544429874420166,
      "step": 3530
    },
    {
      "epoch": 0.25379982793232003,
      "grad_norm": 1.7221120163446813,
      "learning_rate": 2e-05,
      "loss": 0.5195553779602051,
      "step": 3540
    },
    {
      "epoch": 0.2545167765987955,
      "grad_norm": 0.8954155006239505,
      "learning_rate": 2e-05,
      "loss": 0.5110740184783935,
      "step": 3550
    },
    {
      "epoch": 0.255233725265271,
      "grad_norm": 3.6788697348363697,
      "learning_rate": 2e-05,
      "loss": 0.5324721336364746,
      "step": 3560
    },
    {
      "epoch": 0.2559506739317465,
      "grad_norm": 2.6391234325783386,
      "learning_rate": 2e-05,
      "loss": 0.5505046844482422,
      "step": 3570
    },
    {
      "epoch": 0.25666762259822196,
      "grad_norm": 1.1599856088626521,
      "learning_rate": 2e-05,
      "loss": 0.5322173118591309,
      "step": 3580
    },
    {
      "epoch": 0.25738457126469744,
      "grad_norm": 1.0317747658748881,
      "learning_rate": 2e-05,
      "loss": 0.5402941703796387,
      "step": 3590
    },
    {
      "epoch": 0.2581015199311729,
      "grad_norm": 1.3862437836779606,
      "learning_rate": 2e-05,
      "loss": 0.538313627243042,
      "step": 3600
    },
    {
      "epoch": 0.2588184685976484,
      "grad_norm": 1.469559776655421,
      "learning_rate": 2e-05,
      "loss": 0.5158867359161377,
      "step": 3610
    },
    {
      "epoch": 0.2595354172641239,
      "grad_norm": 4.902163563323479,
      "learning_rate": 2e-05,
      "loss": 0.5293915748596192,
      "step": 3620
    },
    {
      "epoch": 0.26025236593059936,
      "grad_norm": 0.9202037499216777,
      "learning_rate": 2e-05,
      "loss": 0.5168137073516845,
      "step": 3630
    },
    {
      "epoch": 0.26096931459707484,
      "grad_norm": 1.3209791239336233,
      "learning_rate": 2e-05,
      "loss": 0.509038782119751,
      "step": 3640
    },
    {
      "epoch": 0.2616862632635503,
      "grad_norm": 1.1050190888380467,
      "learning_rate": 2e-05,
      "loss": 0.5604996681213379,
      "step": 3650
    },
    {
      "epoch": 0.2624032119300258,
      "grad_norm": 1.8280374024123087,
      "learning_rate": 2e-05,
      "loss": 0.5344246864318848,
      "step": 3660
    },
    {
      "epoch": 0.2631201605965013,
      "grad_norm": 2.0165992030973685,
      "learning_rate": 2e-05,
      "loss": 0.5287519454956054,
      "step": 3670
    },
    {
      "epoch": 0.26383710926297677,
      "grad_norm": 1.7700286478460387,
      "learning_rate": 2e-05,
      "loss": 0.5409080028533936,
      "step": 3680
    },
    {
      "epoch": 0.26455405792945225,
      "grad_norm": 0.844431810037113,
      "learning_rate": 2e-05,
      "loss": 0.5483170509338379,
      "step": 3690
    },
    {
      "epoch": 0.2652710065959277,
      "grad_norm": 5.6130020525259985,
      "learning_rate": 2e-05,
      "loss": 0.5398725032806396,
      "step": 3700
    },
    {
      "epoch": 0.2659879552624032,
      "grad_norm": 1.7896231276325862,
      "learning_rate": 2e-05,
      "loss": 0.520438289642334,
      "step": 3710
    },
    {
      "epoch": 0.2667049039288787,
      "grad_norm": 2.600352679110944,
      "learning_rate": 2e-05,
      "loss": 0.5343352317810058,
      "step": 3720
    },
    {
      "epoch": 0.26742185259535417,
      "grad_norm": 1.786897949642513,
      "learning_rate": 2e-05,
      "loss": 0.5271599769592286,
      "step": 3730
    },
    {
      "epoch": 0.26813880126182965,
      "grad_norm": 34.89744203753394,
      "learning_rate": 2e-05,
      "loss": 0.5272274017333984,
      "step": 3740
    },
    {
      "epoch": 0.26885574992830513,
      "grad_norm": 1.127095700380583,
      "learning_rate": 2e-05,
      "loss": 0.5211878776550293,
      "step": 3750
    },
    {
      "epoch": 0.2695726985947806,
      "grad_norm": 0.8849244154609072,
      "learning_rate": 2e-05,
      "loss": 0.5353400230407714,
      "step": 3760
    },
    {
      "epoch": 0.2702896472612561,
      "grad_norm": 0.8449407515977918,
      "learning_rate": 2e-05,
      "loss": 0.5179182052612304,
      "step": 3770
    },
    {
      "epoch": 0.2710065959277316,
      "grad_norm": 0.9021449533353064,
      "learning_rate": 2e-05,
      "loss": 0.5538331508636475,
      "step": 3780
    },
    {
      "epoch": 0.27172354459420706,
      "grad_norm": 1.1812221975491481,
      "learning_rate": 2e-05,
      "loss": 0.5262421607971192,
      "step": 3790
    },
    {
      "epoch": 0.27244049326068254,
      "grad_norm": 1.1275589974934705,
      "learning_rate": 2e-05,
      "loss": 0.5089552879333497,
      "step": 3800
    },
    {
      "epoch": 0.273157441927158,
      "grad_norm": 1.3028824864996773,
      "learning_rate": 2e-05,
      "loss": 0.5695682048797608,
      "step": 3810
    },
    {
      "epoch": 0.2738743905936335,
      "grad_norm": 1.7066815297352904,
      "learning_rate": 2e-05,
      "loss": 0.49418067932128906,
      "step": 3820
    },
    {
      "epoch": 0.274591339260109,
      "grad_norm": 1.035338338287883,
      "learning_rate": 2e-05,
      "loss": 0.5330035209655761,
      "step": 3830
    },
    {
      "epoch": 0.27530828792658446,
      "grad_norm": 1.1358706036918462,
      "learning_rate": 2e-05,
      "loss": 0.5501136779785156,
      "step": 3840
    },
    {
      "epoch": 0.27602523659305994,
      "grad_norm": 1.2892221657946634,
      "learning_rate": 2e-05,
      "loss": 0.5339395046234131,
      "step": 3850
    },
    {
      "epoch": 0.2767421852595354,
      "grad_norm": 0.8561313886516896,
      "learning_rate": 2e-05,
      "loss": 0.5449686050415039,
      "step": 3860
    },
    {
      "epoch": 0.2774591339260109,
      "grad_norm": 1.4646721686393163,
      "learning_rate": 2e-05,
      "loss": 0.5379161834716797,
      "step": 3870
    },
    {
      "epoch": 0.2781760825924864,
      "grad_norm": 1.9670767074011883,
      "learning_rate": 2e-05,
      "loss": 0.5255084991455078,
      "step": 3880
    },
    {
      "epoch": 0.27889303125896187,
      "grad_norm": 2.099890798442472,
      "learning_rate": 2e-05,
      "loss": 0.550378942489624,
      "step": 3890
    },
    {
      "epoch": 0.27960997992543735,
      "grad_norm": 0.8870149199510412,
      "learning_rate": 2e-05,
      "loss": 0.5343522071838379,
      "step": 3900
    },
    {
      "epoch": 0.2803269285919128,
      "grad_norm": 1.3490393011104544,
      "learning_rate": 2e-05,
      "loss": 0.5435528755187988,
      "step": 3910
    },
    {
      "epoch": 0.2810438772583883,
      "grad_norm": 3.727527648599542,
      "learning_rate": 2e-05,
      "loss": 0.5402599334716797,
      "step": 3920
    },
    {
      "epoch": 0.2817608259248638,
      "grad_norm": 1.4537289353262282,
      "learning_rate": 2e-05,
      "loss": 0.5270157337188721,
      "step": 3930
    },
    {
      "epoch": 0.28247777459133927,
      "grad_norm": 0.9047404412492325,
      "learning_rate": 2e-05,
      "loss": 0.5105515956878662,
      "step": 3940
    },
    {
      "epoch": 0.28319472325781475,
      "grad_norm": 1.1466679192065081,
      "learning_rate": 2e-05,
      "loss": 0.5366024017333985,
      "step": 3950
    },
    {
      "epoch": 0.28391167192429023,
      "grad_norm": 0.8117669846146933,
      "learning_rate": 2e-05,
      "loss": 0.5115161895751953,
      "step": 3960
    },
    {
      "epoch": 0.2846286205907657,
      "grad_norm": 1.6309324462455728,
      "learning_rate": 2e-05,
      "loss": 0.5267614841461181,
      "step": 3970
    },
    {
      "epoch": 0.2853455692572412,
      "grad_norm": 1.0495675528045048,
      "learning_rate": 2e-05,
      "loss": 0.5473695278167725,
      "step": 3980
    },
    {
      "epoch": 0.2860625179237167,
      "grad_norm": 1.3449854539899182,
      "learning_rate": 2e-05,
      "loss": 0.5569032669067383,
      "step": 3990
    },
    {
      "epoch": 0.28677946659019216,
      "grad_norm": 1.0209829023992651,
      "learning_rate": 2e-05,
      "loss": 0.5159942626953125,
      "step": 4000
    },
    {
      "epoch": 0.28749641525666764,
      "grad_norm": 1.048796891221019,
      "learning_rate": 2e-05,
      "loss": 0.5339295387268066,
      "step": 4010
    },
    {
      "epoch": 0.2882133639231431,
      "grad_norm": 1.1555551678444418,
      "learning_rate": 2e-05,
      "loss": 0.5608357906341552,
      "step": 4020
    },
    {
      "epoch": 0.2889303125896186,
      "grad_norm": 0.9272492182688843,
      "learning_rate": 2e-05,
      "loss": 0.5217494010925293,
      "step": 4030
    },
    {
      "epoch": 0.2896472612560941,
      "grad_norm": 1.3894292189077933,
      "learning_rate": 2e-05,
      "loss": 0.511276388168335,
      "step": 4040
    },
    {
      "epoch": 0.29036420992256956,
      "grad_norm": 0.7407717198082833,
      "learning_rate": 2e-05,
      "loss": 0.5470283508300782,
      "step": 4050
    },
    {
      "epoch": 0.29108115858904504,
      "grad_norm": 1.1051131429831134,
      "learning_rate": 2e-05,
      "loss": 0.537915563583374,
      "step": 4060
    },
    {
      "epoch": 0.2917981072555205,
      "grad_norm": 0.8401406329752389,
      "learning_rate": 2e-05,
      "loss": 0.5486094951629639,
      "step": 4070
    },
    {
      "epoch": 0.292515055921996,
      "grad_norm": 1.2471869650710634,
      "learning_rate": 2e-05,
      "loss": 0.5469850063323974,
      "step": 4080
    },
    {
      "epoch": 0.2932320045884715,
      "grad_norm": 0.940817110044171,
      "learning_rate": 2e-05,
      "loss": 0.5081852912902832,
      "step": 4090
    },
    {
      "epoch": 0.29394895325494697,
      "grad_norm": 0.7183349971004513,
      "learning_rate": 2e-05,
      "loss": 0.5073794364929199,
      "step": 4100
    },
    {
      "epoch": 0.29466590192142245,
      "grad_norm": 1.3762494207809675,
      "learning_rate": 2e-05,
      "loss": 0.5661502838134765,
      "step": 4110
    },
    {
      "epoch": 0.2953828505878979,
      "grad_norm": 1.0136514670453665,
      "learning_rate": 2e-05,
      "loss": 0.49536771774291993,
      "step": 4120
    },
    {
      "epoch": 0.2960997992543734,
      "grad_norm": 1.6072543316473569,
      "learning_rate": 2e-05,
      "loss": 0.5378045558929443,
      "step": 4130
    },
    {
      "epoch": 0.2968167479208489,
      "grad_norm": 0.9274375711919706,
      "learning_rate": 2e-05,
      "loss": 0.5259858131408691,
      "step": 4140
    },
    {
      "epoch": 0.29753369658732437,
      "grad_norm": 2.9464960782210206,
      "learning_rate": 2e-05,
      "loss": 0.49956450462341306,
      "step": 4150
    },
    {
      "epoch": 0.29825064525379985,
      "grad_norm": 0.9127254375786255,
      "learning_rate": 2e-05,
      "loss": 0.5101244926452637,
      "step": 4160
    },
    {
      "epoch": 0.29896759392027533,
      "grad_norm": 2.0323708483195086,
      "learning_rate": 2e-05,
      "loss": 0.5222911834716797,
      "step": 4170
    },
    {
      "epoch": 0.2996845425867508,
      "grad_norm": 0.7153397684188944,
      "learning_rate": 2e-05,
      "loss": 0.5137488365173339,
      "step": 4180
    },
    {
      "epoch": 0.3004014912532263,
      "grad_norm": 0.7537514853587618,
      "learning_rate": 2e-05,
      "loss": 0.5334841728210449,
      "step": 4190
    },
    {
      "epoch": 0.3011184399197018,
      "grad_norm": 0.7512756591177091,
      "learning_rate": 2e-05,
      "loss": 0.5073200225830078,
      "step": 4200
    },
    {
      "epoch": 0.30183538858617726,
      "grad_norm": 1.1355209944489406,
      "learning_rate": 2e-05,
      "loss": 0.5266365051269531,
      "step": 4210
    },
    {
      "epoch": 0.30255233725265274,
      "grad_norm": 0.9602154926646461,
      "learning_rate": 2e-05,
      "loss": 0.5233523368835449,
      "step": 4220
    },
    {
      "epoch": 0.3032692859191282,
      "grad_norm": 1.3085434347616332,
      "learning_rate": 2e-05,
      "loss": 0.5303298950195312,
      "step": 4230
    },
    {
      "epoch": 0.3039862345856037,
      "grad_norm": 1.2585969589091635,
      "learning_rate": 2e-05,
      "loss": 0.5370527267456054,
      "step": 4240
    },
    {
      "epoch": 0.3047031832520791,
      "grad_norm": 1.50811456467798,
      "learning_rate": 2e-05,
      "loss": 0.5190601348876953,
      "step": 4250
    },
    {
      "epoch": 0.3054201319185546,
      "grad_norm": 0.9553706833022251,
      "learning_rate": 2e-05,
      "loss": 0.5523821830749511,
      "step": 4260
    },
    {
      "epoch": 0.3061370805850301,
      "grad_norm": 0.7940880255836344,
      "learning_rate": 2e-05,
      "loss": 0.5279844284057618,
      "step": 4270
    },
    {
      "epoch": 0.30685402925150557,
      "grad_norm": 2.56500990851689,
      "learning_rate": 2e-05,
      "loss": 0.5298017024993896,
      "step": 4280
    },
    {
      "epoch": 0.30757097791798105,
      "grad_norm": 0.8807319175072179,
      "learning_rate": 2e-05,
      "loss": 0.49184932708740237,
      "step": 4290
    },
    {
      "epoch": 0.30828792658445653,
      "grad_norm": 1.065194428026888,
      "learning_rate": 2e-05,
      "loss": 0.5380837440490722,
      "step": 4300
    },
    {
      "epoch": 0.309004875250932,
      "grad_norm": 0.9326833484788498,
      "learning_rate": 2e-05,
      "loss": 0.5197104454040528,
      "step": 4310
    },
    {
      "epoch": 0.3097218239174075,
      "grad_norm": 1.303500662972627,
      "learning_rate": 2e-05,
      "loss": 0.5344608306884766,
      "step": 4320
    },
    {
      "epoch": 0.31043877258388297,
      "grad_norm": 0.8375067137563889,
      "learning_rate": 2e-05,
      "loss": 0.523453950881958,
      "step": 4330
    },
    {
      "epoch": 0.31115572125035845,
      "grad_norm": 0.831091335031285,
      "learning_rate": 2e-05,
      "loss": 0.5444564342498779,
      "step": 4340
    },
    {
      "epoch": 0.31187266991683393,
      "grad_norm": 1.3242631944801873,
      "learning_rate": 2e-05,
      "loss": 0.5317873001098633,
      "step": 4350
    },
    {
      "epoch": 0.3125896185833094,
      "grad_norm": 0.8208960443370561,
      "learning_rate": 2e-05,
      "loss": 0.5183848381042481,
      "step": 4360
    },
    {
      "epoch": 0.3133065672497849,
      "grad_norm": 1.3637199071522241,
      "learning_rate": 2e-05,
      "loss": 0.5503087520599366,
      "step": 4370
    },
    {
      "epoch": 0.3140235159162604,
      "grad_norm": 1.3592162400309415,
      "learning_rate": 2e-05,
      "loss": 0.5206357002258301,
      "step": 4380
    },
    {
      "epoch": 0.31474046458273586,
      "grad_norm": 1.2425922786851173,
      "learning_rate": 2e-05,
      "loss": 0.5199076175689697,
      "step": 4390
    },
    {
      "epoch": 0.31545741324921134,
      "grad_norm": 1.2575966509834688,
      "learning_rate": 2e-05,
      "loss": 0.5268381118774415,
      "step": 4400
    },
    {
      "epoch": 0.3161743619156868,
      "grad_norm": 0.7771013448081049,
      "learning_rate": 2e-05,
      "loss": 0.5202343940734864,
      "step": 4410
    },
    {
      "epoch": 0.3168913105821623,
      "grad_norm": 1.1442966153026535,
      "learning_rate": 2e-05,
      "loss": 0.5246430397033691,
      "step": 4420
    },
    {
      "epoch": 0.3176082592486378,
      "grad_norm": 1.1232379320114845,
      "learning_rate": 2e-05,
      "loss": 0.5217968940734863,
      "step": 4430
    },
    {
      "epoch": 0.31832520791511326,
      "grad_norm": 0.9280548025649079,
      "learning_rate": 2e-05,
      "loss": 0.5315206527709961,
      "step": 4440
    },
    {
      "epoch": 0.31904215658158874,
      "grad_norm": 3.3903608855658147,
      "learning_rate": 2e-05,
      "loss": 0.5537518501281739,
      "step": 4450
    },
    {
      "epoch": 0.3197591052480642,
      "grad_norm": 2.709776716054461,
      "learning_rate": 2e-05,
      "loss": 0.5103358745574951,
      "step": 4460
    },
    {
      "epoch": 0.3204760539145397,
      "grad_norm": 1.6672898069796627,
      "learning_rate": 2e-05,
      "loss": 0.5406561851501465,
      "step": 4470
    },
    {
      "epoch": 0.3211930025810152,
      "grad_norm": 0.9271392992433009,
      "learning_rate": 2e-05,
      "loss": 0.5080900192260742,
      "step": 4480
    },
    {
      "epoch": 0.32190995124749067,
      "grad_norm": 9.658363538376468,
      "learning_rate": 2e-05,
      "loss": 0.5152578353881836,
      "step": 4490
    },
    {
      "epoch": 0.32262689991396615,
      "grad_norm": 0.9456264520111266,
      "learning_rate": 2e-05,
      "loss": 0.526423978805542,
      "step": 4500
    },
    {
      "epoch": 0.32334384858044163,
      "grad_norm": 8.836012690578155,
      "learning_rate": 2e-05,
      "loss": 0.5180334091186524,
      "step": 4510
    },
    {
      "epoch": 0.3240607972469171,
      "grad_norm": 7.530549512723217,
      "learning_rate": 2e-05,
      "loss": 0.5399107933044434,
      "step": 4520
    },
    {
      "epoch": 0.3247777459133926,
      "grad_norm": 0.7602846925025712,
      "learning_rate": 2e-05,
      "loss": 0.5090080261230469,
      "step": 4530
    },
    {
      "epoch": 0.32549469457986807,
      "grad_norm": 1.2398730679196137,
      "learning_rate": 2e-05,
      "loss": 0.5102209091186524,
      "step": 4540
    },
    {
      "epoch": 0.32621164324634355,
      "grad_norm": 1.3394334619673274,
      "learning_rate": 2e-05,
      "loss": 0.532669973373413,
      "step": 4550
    },
    {
      "epoch": 0.32692859191281903,
      "grad_norm": 1.5220581125375645,
      "learning_rate": 2e-05,
      "loss": 0.5228144645690918,
      "step": 4560
    },
    {
      "epoch": 0.3276455405792945,
      "grad_norm": 1.206295452618785,
      "learning_rate": 2e-05,
      "loss": 0.5468737602233886,
      "step": 4570
    },
    {
      "epoch": 0.32836248924577,
      "grad_norm": 1.0287167604693068,
      "learning_rate": 2e-05,
      "loss": 0.5387739181518555,
      "step": 4580
    },
    {
      "epoch": 0.3290794379122455,
      "grad_norm": 1.4347719919263757,
      "learning_rate": 2e-05,
      "loss": 0.4996960639953613,
      "step": 4590
    },
    {
      "epoch": 0.32979638657872096,
      "grad_norm": 1.2196058720654717,
      "learning_rate": 2e-05,
      "loss": 0.5508327484130859,
      "step": 4600
    },
    {
      "epoch": 0.33051333524519644,
      "grad_norm": 3.353516453922189,
      "learning_rate": 2e-05,
      "loss": 0.5273049831390381,
      "step": 4610
    },
    {
      "epoch": 0.3312302839116719,
      "grad_norm": 1.1284956895381522,
      "learning_rate": 2e-05,
      "loss": 0.5522353172302246,
      "step": 4620
    },
    {
      "epoch": 0.3319472325781474,
      "grad_norm": 1.0483044263166832,
      "learning_rate": 2e-05,
      "loss": 0.5340966701507568,
      "step": 4630
    },
    {
      "epoch": 0.3326641812446229,
      "grad_norm": 1.0505949579738072,
      "learning_rate": 2e-05,
      "loss": 0.5149962425231933,
      "step": 4640
    },
    {
      "epoch": 0.33338112991109836,
      "grad_norm": 1.2822549972767612,
      "learning_rate": 2e-05,
      "loss": 0.5428599357604981,
      "step": 4650
    },
    {
      "epoch": 0.33409807857757384,
      "grad_norm": 1.1089991469318894,
      "learning_rate": 2e-05,
      "loss": 0.5226129055023193,
      "step": 4660
    },
    {
      "epoch": 0.3348150272440493,
      "grad_norm": 1.4639392458736005,
      "learning_rate": 2e-05,
      "loss": 0.5530063152313233,
      "step": 4670
    },
    {
      "epoch": 0.3355319759105248,
      "grad_norm": 3.018038636039687,
      "learning_rate": 2e-05,
      "loss": 0.5231571197509766,
      "step": 4680
    },
    {
      "epoch": 0.3362489245770003,
      "grad_norm": 0.6835281448701501,
      "learning_rate": 2e-05,
      "loss": 0.5274917125701905,
      "step": 4690
    },
    {
      "epoch": 0.33696587324347577,
      "grad_norm": 0.7218878189946621,
      "learning_rate": 2e-05,
      "loss": 0.5089258193969727,
      "step": 4700
    },
    {
      "epoch": 0.33768282190995125,
      "grad_norm": 0.9212847944930124,
      "learning_rate": 2e-05,
      "loss": 0.5405418395996093,
      "step": 4710
    },
    {
      "epoch": 0.33839977057642673,
      "grad_norm": 1.5528579569101695,
      "learning_rate": 2e-05,
      "loss": 0.5413121700286865,
      "step": 4720
    },
    {
      "epoch": 0.3391167192429022,
      "grad_norm": 1.034387507489718,
      "learning_rate": 2e-05,
      "loss": 0.5292513847351075,
      "step": 4730
    },
    {
      "epoch": 0.3398336679093777,
      "grad_norm": 0.7357903493435658,
      "learning_rate": 2e-05,
      "loss": 0.5194945335388184,
      "step": 4740
    },
    {
      "epoch": 0.34055061657585317,
      "grad_norm": 1.1175454766901796,
      "learning_rate": 2e-05,
      "loss": 0.5257889270782471,
      "step": 4750
    },
    {
      "epoch": 0.34126756524232865,
      "grad_norm": 0.7346263204759098,
      "learning_rate": 2e-05,
      "loss": 0.5154601097106933,
      "step": 4760
    },
    {
      "epoch": 0.34198451390880413,
      "grad_norm": 10.374016819073029,
      "learning_rate": 2e-05,
      "loss": 0.5131362438201904,
      "step": 4770
    },
    {
      "epoch": 0.3427014625752796,
      "grad_norm": 1.2487085511789497,
      "learning_rate": 2e-05,
      "loss": 0.552493953704834,
      "step": 4780
    },
    {
      "epoch": 0.3434184112417551,
      "grad_norm": 0.9040807123788447,
      "learning_rate": 2e-05,
      "loss": 0.5034106254577637,
      "step": 4790
    },
    {
      "epoch": 0.3441353599082306,
      "grad_norm": 1.6325792499153733,
      "learning_rate": 2e-05,
      "loss": 0.5084205150604248,
      "step": 4800
    },
    {
      "epoch": 0.34485230857470606,
      "grad_norm": 0.8692253592992182,
      "learning_rate": 2e-05,
      "loss": 0.5287261962890625,
      "step": 4810
    },
    {
      "epoch": 0.34556925724118154,
      "grad_norm": 0.9060034542394609,
      "learning_rate": 2e-05,
      "loss": 0.5296453475952149,
      "step": 4820
    },
    {
      "epoch": 0.346286205907657,
      "grad_norm": 0.8186345752185697,
      "learning_rate": 2e-05,
      "loss": 0.5161468029022217,
      "step": 4830
    },
    {
      "epoch": 0.3470031545741325,
      "grad_norm": 1.8560013083847244,
      "learning_rate": 2e-05,
      "loss": 0.4909986972808838,
      "step": 4840
    },
    {
      "epoch": 0.347720103240608,
      "grad_norm": 0.7298525572293141,
      "learning_rate": 2e-05,
      "loss": 0.5153409481048584,
      "step": 4850
    },
    {
      "epoch": 0.34843705190708346,
      "grad_norm": 1.2015085844052533,
      "learning_rate": 2e-05,
      "loss": 0.5443366050720215,
      "step": 4860
    },
    {
      "epoch": 0.34915400057355894,
      "grad_norm": 4.635716606384744,
      "learning_rate": 2e-05,
      "loss": 0.5138832092285156,
      "step": 4870
    },
    {
      "epoch": 0.3498709492400344,
      "grad_norm": 0.7659313498165575,
      "learning_rate": 2e-05,
      "loss": 0.5509651184082032,
      "step": 4880
    },
    {
      "epoch": 0.3505878979065099,
      "grad_norm": 1.0403141271061953,
      "learning_rate": 2e-05,
      "loss": 0.5317223548889161,
      "step": 4890
    },
    {
      "epoch": 0.3513048465729854,
      "grad_norm": 0.6513852176682293,
      "learning_rate": 2e-05,
      "loss": 0.5015994071960449,
      "step": 4900
    },
    {
      "epoch": 0.35202179523946087,
      "grad_norm": 1.0705840868706666,
      "learning_rate": 2e-05,
      "loss": 0.5081709861755371,
      "step": 4910
    },
    {
      "epoch": 0.35273874390593635,
      "grad_norm": 1.749688192389044,
      "learning_rate": 2e-05,
      "loss": 0.5129611015319824,
      "step": 4920
    },
    {
      "epoch": 0.35345569257241183,
      "grad_norm": 0.8572715991630108,
      "learning_rate": 2e-05,
      "loss": 0.5112964153289795,
      "step": 4930
    },
    {
      "epoch": 0.3541726412388873,
      "grad_norm": 0.8468238818907539,
      "learning_rate": 2e-05,
      "loss": 0.49866714477539065,
      "step": 4940
    },
    {
      "epoch": 0.3548895899053628,
      "grad_norm": 1.0232806723693195,
      "learning_rate": 2e-05,
      "loss": 0.5378201961517334,
      "step": 4950
    },
    {
      "epoch": 0.35560653857183827,
      "grad_norm": 0.8907452319948949,
      "learning_rate": 2e-05,
      "loss": 0.5180432796478271,
      "step": 4960
    },
    {
      "epoch": 0.35632348723831375,
      "grad_norm": 0.6780175475680283,
      "learning_rate": 2e-05,
      "loss": 0.5277535438537597,
      "step": 4970
    },
    {
      "epoch": 0.35704043590478923,
      "grad_norm": 0.8384796037599304,
      "learning_rate": 2e-05,
      "loss": 0.5275757312774658,
      "step": 4980
    },
    {
      "epoch": 0.3577573845712647,
      "grad_norm": 1.0410192454532412,
      "learning_rate": 2e-05,
      "loss": 0.5304989814758301,
      "step": 4990
    },
    {
      "epoch": 0.3584743332377402,
      "grad_norm": 1.2169254648330499,
      "learning_rate": 2e-05,
      "loss": 0.49916853904724123,
      "step": 5000
    },
    {
      "epoch": 0.3591912819042157,
      "grad_norm": 1.5296852971547796,
      "learning_rate": 2e-05,
      "loss": 0.5326381206512452,
      "step": 5010
    },
    {
      "epoch": 0.35990823057069116,
      "grad_norm": 0.9266812995951265,
      "learning_rate": 2e-05,
      "loss": 0.5403494834899902,
      "step": 5020
    },
    {
      "epoch": 0.36062517923716664,
      "grad_norm": 0.66962033280218,
      "learning_rate": 2e-05,
      "loss": 0.5185005187988281,
      "step": 5030
    },
    {
      "epoch": 0.3613421279036421,
      "grad_norm": 0.8241740934101547,
      "learning_rate": 2e-05,
      "loss": 0.5594071865081787,
      "step": 5040
    },
    {
      "epoch": 0.3620590765701176,
      "grad_norm": 4.017854901284413,
      "learning_rate": 2e-05,
      "loss": 0.5494575500488281,
      "step": 5050
    },
    {
      "epoch": 0.3627760252365931,
      "grad_norm": 0.8824091133462023,
      "learning_rate": 2e-05,
      "loss": 0.5229904174804687,
      "step": 5060
    },
    {
      "epoch": 0.36349297390306856,
      "grad_norm": 1.0534906764883571,
      "learning_rate": 2e-05,
      "loss": 0.5292678356170655,
      "step": 5070
    },
    {
      "epoch": 0.36420992256954404,
      "grad_norm": 2.350566824767454,
      "learning_rate": 2e-05,
      "loss": 0.5245790958404541,
      "step": 5080
    },
    {
      "epoch": 0.3649268712360195,
      "grad_norm": 0.8649945040601594,
      "learning_rate": 1.9999810479633703e-05,
      "loss": 0.5199171900749207,
      "step": 5090
    },
    {
      "epoch": 0.365643819902495,
      "grad_norm": 1.4192095928948318,
      "learning_rate": 1.997707677020786e-05,
      "loss": 0.49900259971618655,
      "step": 5100
    },
    {
      "epoch": 0.3663607685689705,
      "grad_norm": 0.9746019647644449,
      "learning_rate": 1.9916538196124637e-05,
      "loss": 0.5325488090515137,
      "step": 5110
    },
    {
      "epoch": 0.36707771723544597,
      "grad_norm": 1.7122816697169039,
      "learning_rate": 1.9818425303863402e-05,
      "loss": 0.5501921653747559,
      "step": 5120
    },
    {
      "epoch": 0.36779466590192145,
      "grad_norm": 1.2289217007512785,
      "learning_rate": 1.9683111732582887e-05,
      "loss": 0.5223170757293701,
      "step": 5130
    },
    {
      "epoch": 0.36851161456839693,
      "grad_norm": 0.8406254313805156,
      "learning_rate": 1.951111279120706e-05,
      "loss": 0.5416806221008301,
      "step": 5140
    },
    {
      "epoch": 0.3692285632348724,
      "grad_norm": 1.0475856830529862,
      "learning_rate": 1.9303083495996028e-05,
      "loss": 0.5196474552154541,
      "step": 5150
    },
    {
      "epoch": 0.3699455119013479,
      "grad_norm": 1.345272041848446,
      "learning_rate": 1.9059816076075482e-05,
      "loss": 0.4953624248504639,
      "step": 5160
    },
    {
      "epoch": 0.37066246056782337,
      "grad_norm": 1.360880101518971,
      "learning_rate": 1.8782236956424142e-05,
      "loss": 0.5308409690856933,
      "step": 5170
    },
    {
      "epoch": 0.37137940923429885,
      "grad_norm": 0.8801048064681738,
      "learning_rate": 1.8471403229808897e-05,
      "loss": 0.5192177772521973,
      "step": 5180
    },
    {
      "epoch": 0.3720963579007743,
      "grad_norm": 0.6853883444578611,
      "learning_rate": 1.8128498631103333e-05,
      "loss": 0.5177340984344483,
      "step": 5190
    },
    {
      "epoch": 0.37281330656724976,
      "grad_norm": 0.86925517413074,
      "learning_rate": 1.7754829029320507e-05,
      "loss": 0.5070761680603028,
      "step": 5200
    },
    {
      "epoch": 0.37353025523372524,
      "grad_norm": 0.7989368161507507,
      "learning_rate": 1.7351817454527473e-05,
      "loss": 0.5509113311767578,
      "step": 5210
    },
    {
      "epoch": 0.3742472039002007,
      "grad_norm": 1.9731799130717005,
      "learning_rate": 1.6920998678580345e-05,
      "loss": 0.5241506099700928,
      "step": 5220
    },
    {
      "epoch": 0.3749641525666762,
      "grad_norm": 1.5847123468994317,
      "learning_rate": 1.6464013370317854e-05,
      "loss": 0.5109374523162842,
      "step": 5230
    },
    {
      "epoch": 0.3756811012331517,
      "grad_norm": 4.642507689054365,
      "learning_rate": 1.5982601847471915e-05,
      "loss": 0.509030818939209,
      "step": 5240
    },
    {
      "epoch": 0.37639804989962716,
      "grad_norm": 2.385998066497452,
      "learning_rate": 1.5478597449089553e-05,
      "loss": 0.5123504638671875,
      "step": 5250
    },
    {
      "epoch": 0.37711499856610264,
      "grad_norm": 7.346705053207243,
      "learning_rate": 1.4953919553705725e-05,
      "loss": 0.48823232650756837,
      "step": 5260
    },
    {
      "epoch": 0.3778319472325781,
      "grad_norm": 2.0307120231081592,
      "learning_rate": 1.441056626985559e-05,
      "loss": 0.5341516017913819,
      "step": 5270
    },
    {
      "epoch": 0.3785488958990536,
      "grad_norm": 0.9615534550772974,
      "learning_rate": 1.3850606826762668e-05,
      "loss": 0.5223010063171387,
      "step": 5280
    },
    {
      "epoch": 0.3792658445655291,
      "grad_norm": 1.0403541443844786,
      "learning_rate": 1.3276173694181128e-05,
      "loss": 0.5469321250915528,
      "step": 5290
    },
    {
      "epoch": 0.37998279323200457,
      "grad_norm": 1.3161696526805364,
      "learning_rate": 1.2689454461401822e-05,
      "loss": 0.5105757713317871,
      "step": 5300
    },
    {
      "epoch": 0.38069974189848005,
      "grad_norm": 1.248636342927182,
      "learning_rate": 1.2092683506348939e-05,
      "loss": 0.4952629566192627,
      "step": 5310
    },
    {
      "epoch": 0.38141669056495553,
      "grad_norm": 1.6902923207133436,
      "learning_rate": 1.14881334864935e-05,
      "loss": 0.5214015483856201,
      "step": 5320
    },
    {
      "epoch": 0.382133639231431,
      "grad_norm": 1.6866015732980495,
      "learning_rate": 1.0878106683988481e-05,
      "loss": 0.5151557445526123,
      "step": 5330
    },
    {
      "epoch": 0.3828505878979065,
      "grad_norm": 3.4997624334240705,
      "learning_rate": 1.026492623798551e-05,
      "loss": 0.5235134601593018,
      "step": 5340
    },
    {
      "epoch": 0.383567536564382,
      "grad_norm": 0.8623408652782719,
      "learning_rate": 9.650927297522743e-06,
      "loss": 0.5171483993530274,
      "step": 5350
    },
    {
      "epoch": 0.38428448523085745,
      "grad_norm": 1.4889899731073444,
      "learning_rate": 9.038448128676012e-06,
      "loss": 0.5175115585327148,
      "step": 5360
    },
    {
      "epoch": 0.38500143389733293,
      "grad_norm": 1.5004230126653615,
      "learning_rate": 8.42982120983949e-06,
      "loss": 0.5092827320098877,
      "step": 5370
    },
    {
      "epoch": 0.3857183825638084,
      "grad_norm": 1.0435552649650932,
      "learning_rate": 7.827364349047303e-06,
      "loss": 0.47614574432373047,
      "step": 5380
    },
    {
      "epoch": 0.3864353312302839,
      "grad_norm": 1.7416428195538385,
      "learning_rate": 7.233371857163706e-06,
      "loss": 0.5028523921966552,
      "step": 5390
    },
    {
      "epoch": 0.3871522798967594,
      "grad_norm": 0.9820783436162582,
      "learning_rate": 6.650105810556492e-06,
      "loss": 0.4988051414489746,
      "step": 5400
    },
    {
      "epoch": 0.38786922856323486,
      "grad_norm": 1.2205397704233307,
      "learning_rate": 6.079787436527681e-06,
      "loss": 0.534299659729004,
      "step": 5410
    },
    {
      "epoch": 0.38858617722971034,
      "grad_norm": 1.0832545565682885,
      "learning_rate": 5.524588654307961e-06,
      "loss": 0.4938631057739258,
      "step": 5420
    },
    {
      "epoch": 0.3893031258961858,
      "grad_norm": 1.9486783254386961,
      "learning_rate": 4.986623803828985e-06,
      "loss": 0.4956830978393555,
      "step": 5430
    },
    {
      "epoch": 0.3900200745626613,
      "grad_norm": 2.1985386591442264,
      "learning_rate": 4.467941593772423e-06,
      "loss": 0.5335124015808106,
      "step": 5440
    },
    {
      "epoch": 0.3907370232291368,
      "grad_norm": 2.1133993347227773,
      "learning_rate": 3.9705172995597606e-06,
      "loss": 0.5205741882324219,
      "step": 5450
    },
    {
      "epoch": 0.39145397189561226,
      "grad_norm": 1.6298215574609096,
      "learning_rate": 3.4962452409948346e-06,
      "loss": 0.487537145614624,
      "step": 5460
    },
    {
      "epoch": 0.39217092056208774,
      "grad_norm": 1.810517045965952,
      "learning_rate": 3.046931568206211e-06,
      "loss": 0.5252443313598633,
      "step": 5470
    },
    {
      "epoch": 0.3928878692285632,
      "grad_norm": 1.710939893099786,
      "learning_rate": 2.624287383362404e-06,
      "loss": 0.5052958965301514,
      "step": 5480
    },
    {
      "epoch": 0.3936048178950387,
      "grad_norm": 4.020060201108611,
      "learning_rate": 2.2299222243541845e-06,
      "loss": 0.4951027393341064,
      "step": 5490
    },
    {
      "epoch": 0.3943217665615142,
      "grad_norm": 1.0342185764611262,
      "learning_rate": 1.8653379352598246e-06,
      "loss": 0.49745979309082033,
      "step": 5500
    },
    {
      "epoch": 0.39503871522798967,
      "grad_norm": 0.9922701117398024,
      "learning_rate": 1.5319229469361075e-06,
      "loss": 0.5045078277587891,
      "step": 5510
    },
    {
      "epoch": 0.39575566389446515,
      "grad_norm": 0.996918073948356,
      "learning_rate": 1.2309469895160917e-06,
      "loss": 0.5075387001037598,
      "step": 5520
    },
    {
      "epoch": 0.39647261256094063,
      "grad_norm": 1.0191115635269217,
      "learning_rate": 9.635562569497974e-07,
      "loss": 0.5054811477661133,
      "step": 5530
    },
    {
      "epoch": 0.3971895612274161,
      "grad_norm": 1.3810942488945044,
      "learning_rate": 7.307690420024867e-07,
      "loss": 0.5235503196716309,
      "step": 5540
    },
    {
      "epoch": 0.3979065098938916,
      "grad_norm": 6.442226356557042,
      "learning_rate": 5.334718583335942e-07,
      "loss": 0.512938117980957,
      "step": 5550
    },
    {
      "epoch": 0.39862345856036707,
      "grad_norm": 1.3413525459690592,
      "learning_rate": 3.7241606442442763e-07,
      "loss": 0.5129291534423828,
      "step": 5560
    },
    {
      "epoch": 0.39934040722684255,
      "grad_norm": 2.15963018675158,
      "learning_rate": 2.482150022115879e-07,
      "loss": 0.4841459274291992,
      "step": 5570
    },
    {
      "epoch": 0.40005735589331803,
      "grad_norm": 0.8347912385470394,
      "learning_rate": 1.6134166132293174e-07,
      "loss": 0.4935769557952881,
      "step": 5580
    },
    {
      "epoch": 0.4007743045597935,
      "grad_norm": 1.1685888074715836,
      "learning_rate": 1.1212687781127357e-07,
      "loss": 0.5193821907043457,
      "step": 5590
    },
    {
      "epoch": 0.40127616862632637,
      "step": 5597,
      "total_flos": 2650292248117248.0,
      "train_loss": 0.0466171020862565,
      "train_runtime": 7428.5435,
      "train_samples_per_second": 9.041,
      "train_steps_per_second": 0.753
    }
  ],
  "logging_steps": 10,
  "max_steps": 5597,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2650292248117248.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}