{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8024806423860051,
  "eval_steps": 500,
  "global_step": 11193,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007169486664754804,
      "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": 0.864994504068432,
      "learning_rate": 2e-05,
      "loss": 0.5199171900749207,
      "step": 5090
    },
    {
      "epoch": 0.365643819902495,
      "grad_norm": 0.9353904567143638,
      "learning_rate": 2e-05,
      "loss": 0.49895362854003905,
      "step": 5100
    },
    {
      "epoch": 0.3663607685689705,
      "grad_norm": 0.8937035108769994,
      "learning_rate": 2e-05,
      "loss": 0.5320062637329102,
      "step": 5110
    },
    {
      "epoch": 0.36707771723544597,
      "grad_norm": 1.3925407204529694,
      "learning_rate": 2e-05,
      "loss": 0.5510077953338623,
      "step": 5120
    },
    {
      "epoch": 0.36779466590192145,
      "grad_norm": 0.8826191960506601,
      "learning_rate": 2e-05,
      "loss": 0.5219274520874023,
      "step": 5130
    },
    {
      "epoch": 0.36851161456839693,
      "grad_norm": 3.784229977914671,
      "learning_rate": 2e-05,
      "loss": 0.540579891204834,
      "step": 5140
    },
    {
      "epoch": 0.3692285632348724,
      "grad_norm": 1.8847205421565885,
      "learning_rate": 2e-05,
      "loss": 0.5202635765075684,
      "step": 5150
    },
    {
      "epoch": 0.3699455119013479,
      "grad_norm": 1.8964469227316363,
      "learning_rate": 2e-05,
      "loss": 0.4954378128051758,
      "step": 5160
    },
    {
      "epoch": 0.37066246056782337,
      "grad_norm": 7.105699471702284,
      "learning_rate": 2e-05,
      "loss": 0.5307998657226562,
      "step": 5170
    },
    {
      "epoch": 0.37137940923429885,
      "grad_norm": 1.5605666028408527,
      "learning_rate": 2e-05,
      "loss": 0.5215501308441162,
      "step": 5180
    },
    {
      "epoch": 0.3720963579007743,
      "grad_norm": 1.248655625279853,
      "learning_rate": 2e-05,
      "loss": 0.5216545581817627,
      "step": 5190
    },
    {
      "epoch": 0.37281330656724976,
      "grad_norm": 0.8639682954950599,
      "learning_rate": 2e-05,
      "loss": 0.5109705448150634,
      "step": 5200
    },
    {
      "epoch": 0.37353025523372524,
      "grad_norm": 1.3003584830761596,
      "learning_rate": 2e-05,
      "loss": 0.5524150848388671,
      "step": 5210
    },
    {
      "epoch": 0.3742472039002007,
      "grad_norm": 18.420650329515507,
      "learning_rate": 2e-05,
      "loss": 0.525451374053955,
      "step": 5220
    },
    {
      "epoch": 0.3749641525666762,
      "grad_norm": 1.1092195972835406,
      "learning_rate": 2e-05,
      "loss": 0.514319896697998,
      "step": 5230
    },
    {
      "epoch": 0.3756811012331517,
      "grad_norm": 0.8860387514578969,
      "learning_rate": 2e-05,
      "loss": 0.5160833835601807,
      "step": 5240
    },
    {
      "epoch": 0.37639804989962716,
      "grad_norm": 1.2156063548721754,
      "learning_rate": 2e-05,
      "loss": 0.5175872325897217,
      "step": 5250
    },
    {
      "epoch": 0.37711499856610264,
      "grad_norm": 0.8059751383379508,
      "learning_rate": 2e-05,
      "loss": 0.49125003814697266,
      "step": 5260
    },
    {
      "epoch": 0.3778319472325781,
      "grad_norm": 1.887055417834278,
      "learning_rate": 2e-05,
      "loss": 0.538143253326416,
      "step": 5270
    },
    {
      "epoch": 0.3785488958990536,
      "grad_norm": 1.0501527241398467,
      "learning_rate": 2e-05,
      "loss": 0.5258872032165527,
      "step": 5280
    },
    {
      "epoch": 0.3792658445655291,
      "grad_norm": 1.8578903112735026,
      "learning_rate": 2e-05,
      "loss": 0.5523202419281006,
      "step": 5290
    },
    {
      "epoch": 0.37998279323200457,
      "grad_norm": 1.1097121587552596,
      "learning_rate": 2e-05,
      "loss": 0.5173210144042969,
      "step": 5300
    },
    {
      "epoch": 0.38069974189848005,
      "grad_norm": 1.7105195569374911,
      "learning_rate": 2e-05,
      "loss": 0.501483964920044,
      "step": 5310
    },
    {
      "epoch": 0.38141669056495553,
      "grad_norm": 1.0585651693643918,
      "learning_rate": 2e-05,
      "loss": 0.5276995658874511,
      "step": 5320
    },
    {
      "epoch": 0.382133639231431,
      "grad_norm": 1.4773940003636923,
      "learning_rate": 2e-05,
      "loss": 0.5233052730560303,
      "step": 5330
    },
    {
      "epoch": 0.3828505878979065,
      "grad_norm": 0.9613272702664429,
      "learning_rate": 2e-05,
      "loss": 0.530787181854248,
      "step": 5340
    },
    {
      "epoch": 0.383567536564382,
      "grad_norm": 0.7708259033505594,
      "learning_rate": 2e-05,
      "loss": 0.5258699417114258,
      "step": 5350
    },
    {
      "epoch": 0.38428448523085745,
      "grad_norm": 0.776015888213791,
      "learning_rate": 2e-05,
      "loss": 0.5277801036834717,
      "step": 5360
    },
    {
      "epoch": 0.38500143389733293,
      "grad_norm": 0.7351475287170331,
      "learning_rate": 2e-05,
      "loss": 0.5188160896301269,
      "step": 5370
    },
    {
      "epoch": 0.3857183825638084,
      "grad_norm": 0.8788762580554667,
      "learning_rate": 2e-05,
      "loss": 0.4874452590942383,
      "step": 5380
    },
    {
      "epoch": 0.3864353312302839,
      "grad_norm": 0.7626767104672794,
      "learning_rate": 2e-05,
      "loss": 0.514564037322998,
      "step": 5390
    },
    {
      "epoch": 0.3871522798967594,
      "grad_norm": 0.6889170582375465,
      "learning_rate": 2e-05,
      "loss": 0.510062313079834,
      "step": 5400
    },
    {
      "epoch": 0.38786922856323486,
      "grad_norm": 0.744309146148602,
      "learning_rate": 2e-05,
      "loss": 0.5457504272460938,
      "step": 5410
    },
    {
      "epoch": 0.38858617722971034,
      "grad_norm": 1.486244127693706,
      "learning_rate": 2e-05,
      "loss": 0.505613899230957,
      "step": 5420
    },
    {
      "epoch": 0.3893031258961858,
      "grad_norm": 1.1859530671051346,
      "learning_rate": 2e-05,
      "loss": 0.5069845199584961,
      "step": 5430
    },
    {
      "epoch": 0.3900200745626613,
      "grad_norm": 0.9290787980440666,
      "learning_rate": 2e-05,
      "loss": 0.5472968101501465,
      "step": 5440
    },
    {
      "epoch": 0.3907370232291368,
      "grad_norm": 1.0013702492980339,
      "learning_rate": 2e-05,
      "loss": 0.534746503829956,
      "step": 5450
    },
    {
      "epoch": 0.39145397189561226,
      "grad_norm": 0.9249328081138126,
      "learning_rate": 2e-05,
      "loss": 0.503115463256836,
      "step": 5460
    },
    {
      "epoch": 0.39217092056208774,
      "grad_norm": 0.971960524681807,
      "learning_rate": 2e-05,
      "loss": 0.5420567989349365,
      "step": 5470
    },
    {
      "epoch": 0.3928878692285632,
      "grad_norm": 0.8571582898936956,
      "learning_rate": 2e-05,
      "loss": 0.5200867652893066,
      "step": 5480
    },
    {
      "epoch": 0.3936048178950387,
      "grad_norm": 1.7778617570586506,
      "learning_rate": 2e-05,
      "loss": 0.5102245330810546,
      "step": 5490
    },
    {
      "epoch": 0.3943217665615142,
      "grad_norm": 0.7766818638445314,
      "learning_rate": 2e-05,
      "loss": 0.514164924621582,
      "step": 5500
    },
    {
      "epoch": 0.39503871522798967,
      "grad_norm": 1.2112565086286067,
      "learning_rate": 2e-05,
      "loss": 0.5177792072296142,
      "step": 5510
    },
    {
      "epoch": 0.39575566389446515,
      "grad_norm": 1.1009938868494242,
      "learning_rate": 2e-05,
      "loss": 0.5230871200561523,
      "step": 5520
    },
    {
      "epoch": 0.39647261256094063,
      "grad_norm": 1.1445964618271325,
      "learning_rate": 2e-05,
      "loss": 0.5187901496887207,
      "step": 5530
    },
    {
      "epoch": 0.3971895612274161,
      "grad_norm": 1.4951681609749377,
      "learning_rate": 2e-05,
      "loss": 0.5384046077728272,
      "step": 5540
    },
    {
      "epoch": 0.3979065098938916,
      "grad_norm": 1.7304859124897563,
      "learning_rate": 2e-05,
      "loss": 0.5273252487182617,
      "step": 5550
    },
    {
      "epoch": 0.39862345856036707,
      "grad_norm": 2.7510677983159515,
      "learning_rate": 2e-05,
      "loss": 0.5297118186950683,
      "step": 5560
    },
    {
      "epoch": 0.39934040722684255,
      "grad_norm": 1.349177687511358,
      "learning_rate": 2e-05,
      "loss": 0.5020649433135986,
      "step": 5570
    },
    {
      "epoch": 0.40005735589331803,
      "grad_norm": 1.1322086645326035,
      "learning_rate": 2e-05,
      "loss": 0.5103409767150879,
      "step": 5580
    },
    {
      "epoch": 0.4007743045597935,
      "grad_norm": 1.136768654097528,
      "learning_rate": 2e-05,
      "loss": 0.5352713584899902,
      "step": 5590
    },
    {
      "epoch": 0.401491253226269,
      "grad_norm": 50.20629136517997,
      "learning_rate": 2e-05,
      "loss": 0.5078218460083008,
      "step": 5600
    },
    {
      "epoch": 0.4022082018927445,
      "grad_norm": 6.113867051116225,
      "learning_rate": 2e-05,
      "loss": 0.5380892753601074,
      "step": 5610
    },
    {
      "epoch": 0.40292515055921996,
      "grad_norm": 0.9246495982565801,
      "learning_rate": 2e-05,
      "loss": 0.5029634952545166,
      "step": 5620
    },
    {
      "epoch": 0.40364209922569544,
      "grad_norm": 1.3022984853933472,
      "learning_rate": 2e-05,
      "loss": 0.5281048774719238,
      "step": 5630
    },
    {
      "epoch": 0.4043590478921709,
      "grad_norm": 1.0384478142461055,
      "learning_rate": 2e-05,
      "loss": 0.5221714019775391,
      "step": 5640
    },
    {
      "epoch": 0.4050759965586464,
      "grad_norm": 0.901129239782602,
      "learning_rate": 2e-05,
      "loss": 0.5202292442321778,
      "step": 5650
    },
    {
      "epoch": 0.4057929452251219,
      "grad_norm": 1.3589696267940317,
      "learning_rate": 2e-05,
      "loss": 0.5326308727264404,
      "step": 5660
    },
    {
      "epoch": 0.40650989389159736,
      "grad_norm": 1.431519902730432,
      "learning_rate": 2e-05,
      "loss": 0.5103866577148437,
      "step": 5670
    },
    {
      "epoch": 0.40722684255807284,
      "grad_norm": 0.9529341285568881,
      "learning_rate": 2e-05,
      "loss": 0.5200950622558593,
      "step": 5680
    },
    {
      "epoch": 0.4079437912245483,
      "grad_norm": 3.4753598646791004,
      "learning_rate": 2e-05,
      "loss": 0.5181207656860352,
      "step": 5690
    },
    {
      "epoch": 0.4086607398910238,
      "grad_norm": 1.1184695617183087,
      "learning_rate": 2e-05,
      "loss": 0.5047992706298828,
      "step": 5700
    },
    {
      "epoch": 0.4093776885574993,
      "grad_norm": 0.8419305815039531,
      "learning_rate": 2e-05,
      "loss": 0.5123749256134034,
      "step": 5710
    },
    {
      "epoch": 0.41009463722397477,
      "grad_norm": 0.8825670902307253,
      "learning_rate": 2e-05,
      "loss": 0.5207160949707031,
      "step": 5720
    },
    {
      "epoch": 0.41081158589045025,
      "grad_norm": 0.6826003108203159,
      "learning_rate": 2e-05,
      "loss": 0.5178203105926513,
      "step": 5730
    },
    {
      "epoch": 0.41152853455692573,
      "grad_norm": 1.1462668425525397,
      "learning_rate": 2e-05,
      "loss": 0.502413558959961,
      "step": 5740
    },
    {
      "epoch": 0.4122454832234012,
      "grad_norm": 1.1791741705216856,
      "learning_rate": 2e-05,
      "loss": 0.5222821235656738,
      "step": 5750
    },
    {
      "epoch": 0.4129624318898767,
      "grad_norm": 0.9909977792005786,
      "learning_rate": 2e-05,
      "loss": 0.4915468215942383,
      "step": 5760
    },
    {
      "epoch": 0.41367938055635217,
      "grad_norm": 2.063794771402862,
      "learning_rate": 2e-05,
      "loss": 0.5063940048217773,
      "step": 5770
    },
    {
      "epoch": 0.41439632922282765,
      "grad_norm": 0.8718294261645427,
      "learning_rate": 2e-05,
      "loss": 0.5175788402557373,
      "step": 5780
    },
    {
      "epoch": 0.41511327788930313,
      "grad_norm": 1.219747929393184,
      "learning_rate": 2e-05,
      "loss": 0.5270315170288086,
      "step": 5790
    },
    {
      "epoch": 0.4158302265557786,
      "grad_norm": 1.5420676989227868,
      "learning_rate": 2e-05,
      "loss": 0.49914159774780276,
      "step": 5800
    },
    {
      "epoch": 0.4165471752222541,
      "grad_norm": 1.0293524025677767,
      "learning_rate": 2e-05,
      "loss": 0.5259191513061523,
      "step": 5810
    },
    {
      "epoch": 0.4172641238887296,
      "grad_norm": 0.9539809160660647,
      "learning_rate": 2e-05,
      "loss": 0.5287211418151856,
      "step": 5820
    },
    {
      "epoch": 0.41798107255520506,
      "grad_norm": 0.9409037795512198,
      "learning_rate": 2e-05,
      "loss": 0.5249923229217529,
      "step": 5830
    },
    {
      "epoch": 0.41869802122168054,
      "grad_norm": 0.7280446188743951,
      "learning_rate": 2e-05,
      "loss": 0.5126391410827636,
      "step": 5840
    },
    {
      "epoch": 0.419414969888156,
      "grad_norm": 0.9614442201865259,
      "learning_rate": 2e-05,
      "loss": 0.5067748069763184,
      "step": 5850
    },
    {
      "epoch": 0.4201319185546315,
      "grad_norm": 1.1014748163383388,
      "learning_rate": 2e-05,
      "loss": 0.5247032165527343,
      "step": 5860
    },
    {
      "epoch": 0.420848867221107,
      "grad_norm": 1.089635249785812,
      "learning_rate": 2e-05,
      "loss": 0.5491992950439453,
      "step": 5870
    },
    {
      "epoch": 0.42156581588758246,
      "grad_norm": 1.1187967553201985,
      "learning_rate": 2e-05,
      "loss": 0.5372167587280273,
      "step": 5880
    },
    {
      "epoch": 0.42228276455405794,
      "grad_norm": 0.974875907150638,
      "learning_rate": 2e-05,
      "loss": 0.50736083984375,
      "step": 5890
    },
    {
      "epoch": 0.4229997132205334,
      "grad_norm": 1.9539186535762432,
      "learning_rate": 2e-05,
      "loss": 0.5319615364074707,
      "step": 5900
    },
    {
      "epoch": 0.4237166618870089,
      "grad_norm": 0.8126966218940485,
      "learning_rate": 2e-05,
      "loss": 0.5033891201019287,
      "step": 5910
    },
    {
      "epoch": 0.4244336105534844,
      "grad_norm": 3.9590995039161436,
      "learning_rate": 2e-05,
      "loss": 0.502863597869873,
      "step": 5920
    },
    {
      "epoch": 0.42515055921995987,
      "grad_norm": 1.8412691840684865,
      "learning_rate": 2e-05,
      "loss": 0.49304885864257814,
      "step": 5930
    },
    {
      "epoch": 0.42586750788643535,
      "grad_norm": 1.1325636263624932,
      "learning_rate": 2e-05,
      "loss": 0.5105993270874023,
      "step": 5940
    },
    {
      "epoch": 0.42658445655291083,
      "grad_norm": 0.9632122363114018,
      "learning_rate": 2e-05,
      "loss": 0.5272776126861572,
      "step": 5950
    },
    {
      "epoch": 0.4273014052193863,
      "grad_norm": 1.9304004187369532,
      "learning_rate": 2e-05,
      "loss": 0.5189320564270019,
      "step": 5960
    },
    {
      "epoch": 0.4280183538858618,
      "grad_norm": 0.6436799993897718,
      "learning_rate": 2e-05,
      "loss": 0.5309199810028076,
      "step": 5970
    },
    {
      "epoch": 0.42873530255233727,
      "grad_norm": 0.8498033071396338,
      "learning_rate": 2e-05,
      "loss": 0.5111265182495117,
      "step": 5980
    },
    {
      "epoch": 0.42945225121881275,
      "grad_norm": 1.0022295645469073,
      "learning_rate": 2e-05,
      "loss": 0.4978248119354248,
      "step": 5990
    },
    {
      "epoch": 0.43016919988528823,
      "grad_norm": 1.5239616771266022,
      "learning_rate": 2e-05,
      "loss": 0.5067305088043212,
      "step": 6000
    },
    {
      "epoch": 0.4308861485517637,
      "grad_norm": 0.8728358727270638,
      "learning_rate": 2e-05,
      "loss": 0.5184773445129395,
      "step": 6010
    },
    {
      "epoch": 0.4316030972182392,
      "grad_norm": 1.718338990812282,
      "learning_rate": 2e-05,
      "loss": 0.5463886260986328,
      "step": 6020
    },
    {
      "epoch": 0.4323200458847147,
      "grad_norm": 2.003793093193922,
      "learning_rate": 2e-05,
      "loss": 0.46938066482543944,
      "step": 6030
    },
    {
      "epoch": 0.43303699455119016,
      "grad_norm": 1.2399779740063872,
      "learning_rate": 2e-05,
      "loss": 0.5034530162811279,
      "step": 6040
    },
    {
      "epoch": 0.43375394321766564,
      "grad_norm": 1.1509675594918138,
      "learning_rate": 2e-05,
      "loss": 0.5230756282806397,
      "step": 6050
    },
    {
      "epoch": 0.4344708918841411,
      "grad_norm": 0.9639190162752002,
      "learning_rate": 2e-05,
      "loss": 0.5109916687011719,
      "step": 6060
    },
    {
      "epoch": 0.4351878405506166,
      "grad_norm": 0.7499484321816493,
      "learning_rate": 2e-05,
      "loss": 0.5106131553649902,
      "step": 6070
    },
    {
      "epoch": 0.4359047892170921,
      "grad_norm": 0.9914731742336221,
      "learning_rate": 2e-05,
      "loss": 0.5066973686218261,
      "step": 6080
    },
    {
      "epoch": 0.43662173788356756,
      "grad_norm": 0.857208561537766,
      "learning_rate": 2e-05,
      "loss": 0.5407319068908691,
      "step": 6090
    },
    {
      "epoch": 0.43733868655004304,
      "grad_norm": 0.7169797083758115,
      "learning_rate": 2e-05,
      "loss": 0.5180103778839111,
      "step": 6100
    },
    {
      "epoch": 0.4380556352165185,
      "grad_norm": 0.957063545154399,
      "learning_rate": 2e-05,
      "loss": 0.5107412338256836,
      "step": 6110
    },
    {
      "epoch": 0.438772583882994,
      "grad_norm": 0.7734098458198758,
      "learning_rate": 2e-05,
      "loss": 0.5341405391693115,
      "step": 6120
    },
    {
      "epoch": 0.43948953254946943,
      "grad_norm": 1.0174519008901894,
      "learning_rate": 2e-05,
      "loss": 0.5222190380096435,
      "step": 6130
    },
    {
      "epoch": 0.4402064812159449,
      "grad_norm": 0.9187110949226082,
      "learning_rate": 2e-05,
      "loss": 0.4954657077789307,
      "step": 6140
    },
    {
      "epoch": 0.4409234298824204,
      "grad_norm": 0.9005278928144959,
      "learning_rate": 2e-05,
      "loss": 0.5304151058197022,
      "step": 6150
    },
    {
      "epoch": 0.4416403785488959,
      "grad_norm": 0.7714558554310395,
      "learning_rate": 2e-05,
      "loss": 0.527191162109375,
      "step": 6160
    },
    {
      "epoch": 0.44235732721537135,
      "grad_norm": 0.9453380387354731,
      "learning_rate": 2e-05,
      "loss": 0.5070681571960449,
      "step": 6170
    },
    {
      "epoch": 0.44307427588184684,
      "grad_norm": 0.6500831944112214,
      "learning_rate": 2e-05,
      "loss": 0.5091273307800293,
      "step": 6180
    },
    {
      "epoch": 0.4437912245483223,
      "grad_norm": 0.8402845962370686,
      "learning_rate": 2e-05,
      "loss": 0.5281148910522461,
      "step": 6190
    },
    {
      "epoch": 0.4445081732147978,
      "grad_norm": 0.9139693065906228,
      "learning_rate": 2e-05,
      "loss": 0.5047156333923339,
      "step": 6200
    },
    {
      "epoch": 0.4452251218812733,
      "grad_norm": 0.7922375946257991,
      "learning_rate": 2e-05,
      "loss": 0.5065896511077881,
      "step": 6210
    },
    {
      "epoch": 0.44594207054774876,
      "grad_norm": 1.0947797846946243,
      "learning_rate": 2e-05,
      "loss": 0.5479475975036621,
      "step": 6220
    },
    {
      "epoch": 0.44665901921422424,
      "grad_norm": 0.8359549408769132,
      "learning_rate": 2e-05,
      "loss": 0.5226608753204346,
      "step": 6230
    },
    {
      "epoch": 0.4473759678806997,
      "grad_norm": 1.3584630170634788,
      "learning_rate": 2e-05,
      "loss": 0.5117042064666748,
      "step": 6240
    },
    {
      "epoch": 0.4480929165471752,
      "grad_norm": 1.1864821243333057,
      "learning_rate": 2e-05,
      "loss": 0.5072595596313476,
      "step": 6250
    },
    {
      "epoch": 0.4488098652136507,
      "grad_norm": 0.769155791812538,
      "learning_rate": 2e-05,
      "loss": 0.5028779983520508,
      "step": 6260
    },
    {
      "epoch": 0.44952681388012616,
      "grad_norm": 0.8088338614240499,
      "learning_rate": 2e-05,
      "loss": 0.5130536556243896,
      "step": 6270
    },
    {
      "epoch": 0.45024376254660164,
      "grad_norm": 0.7949187532822332,
      "learning_rate": 2e-05,
      "loss": 0.5382438659667969,
      "step": 6280
    },
    {
      "epoch": 0.4509607112130771,
      "grad_norm": 0.7594721444935496,
      "learning_rate": 2e-05,
      "loss": 0.5151664257049561,
      "step": 6290
    },
    {
      "epoch": 0.4516776598795526,
      "grad_norm": 0.869309280691975,
      "learning_rate": 2e-05,
      "loss": 0.5416646480560303,
      "step": 6300
    },
    {
      "epoch": 0.4523946085460281,
      "grad_norm": 0.8042678629341743,
      "learning_rate": 2e-05,
      "loss": 0.5285287380218506,
      "step": 6310
    },
    {
      "epoch": 0.45311155721250357,
      "grad_norm": 0.7572634298219694,
      "learning_rate": 2e-05,
      "loss": 0.5035251617431641,
      "step": 6320
    },
    {
      "epoch": 0.45382850587897905,
      "grad_norm": 0.9840000607984183,
      "learning_rate": 2e-05,
      "loss": 0.5499188899993896,
      "step": 6330
    },
    {
      "epoch": 0.45454545454545453,
      "grad_norm": 0.7798960769056715,
      "learning_rate": 2e-05,
      "loss": 0.5174559593200684,
      "step": 6340
    },
    {
      "epoch": 0.45526240321193,
      "grad_norm": 1.036069709981091,
      "learning_rate": 2e-05,
      "loss": 0.5178956031799317,
      "step": 6350
    },
    {
      "epoch": 0.4559793518784055,
      "grad_norm": 2.0305440286196146,
      "learning_rate": 2e-05,
      "loss": 0.5299837112426757,
      "step": 6360
    },
    {
      "epoch": 0.456696300544881,
      "grad_norm": 2.2829479939094894,
      "learning_rate": 2e-05,
      "loss": 0.5273044109344482,
      "step": 6370
    },
    {
      "epoch": 0.45741324921135645,
      "grad_norm": 1.0684306096639176,
      "learning_rate": 2e-05,
      "loss": 0.5194907188415527,
      "step": 6380
    },
    {
      "epoch": 0.45813019787783194,
      "grad_norm": 1.096021155531072,
      "learning_rate": 2e-05,
      "loss": 0.514397382736206,
      "step": 6390
    },
    {
      "epoch": 0.4588471465443074,
      "grad_norm": 0.7690059084411529,
      "learning_rate": 2e-05,
      "loss": 0.515052604675293,
      "step": 6400
    },
    {
      "epoch": 0.4595640952107829,
      "grad_norm": 0.7662606585791335,
      "learning_rate": 2e-05,
      "loss": 0.5275700092315674,
      "step": 6410
    },
    {
      "epoch": 0.4602810438772584,
      "grad_norm": 1.031540544035458,
      "learning_rate": 2e-05,
      "loss": 0.5342594146728515,
      "step": 6420
    },
    {
      "epoch": 0.46099799254373386,
      "grad_norm": 0.8421682140680002,
      "learning_rate": 2e-05,
      "loss": 0.5030712127685547,
      "step": 6430
    },
    {
      "epoch": 0.46171494121020934,
      "grad_norm": 0.8559456653868788,
      "learning_rate": 2e-05,
      "loss": 0.5293164253234863,
      "step": 6440
    },
    {
      "epoch": 0.4624318898766848,
      "grad_norm": 2.2463749603040957,
      "learning_rate": 2e-05,
      "loss": 0.526069974899292,
      "step": 6450
    },
    {
      "epoch": 0.4631488385431603,
      "grad_norm": 8.066005333995465,
      "learning_rate": 2e-05,
      "loss": 0.5071203231811523,
      "step": 6460
    },
    {
      "epoch": 0.4638657872096358,
      "grad_norm": 0.9319640204441687,
      "learning_rate": 2e-05,
      "loss": 0.5263157367706299,
      "step": 6470
    },
    {
      "epoch": 0.46458273587611126,
      "grad_norm": 1.3856440442419842,
      "learning_rate": 2e-05,
      "loss": 0.5465081691741943,
      "step": 6480
    },
    {
      "epoch": 0.46529968454258674,
      "grad_norm": 1.0412256357594378,
      "learning_rate": 2e-05,
      "loss": 0.5231462001800538,
      "step": 6490
    },
    {
      "epoch": 0.4660166332090622,
      "grad_norm": 1.6775152417909436,
      "learning_rate": 2e-05,
      "loss": 0.5453738689422607,
      "step": 6500
    },
    {
      "epoch": 0.4667335818755377,
      "grad_norm": 1.3250783374984014,
      "learning_rate": 2e-05,
      "loss": 0.5328183650970459,
      "step": 6510
    },
    {
      "epoch": 0.4674505305420132,
      "grad_norm": 2.1311474377212067,
      "learning_rate": 2e-05,
      "loss": 0.4897273540496826,
      "step": 6520
    },
    {
      "epoch": 0.46816747920848867,
      "grad_norm": 1.0026120622130248,
      "learning_rate": 2e-05,
      "loss": 0.5073579311370849,
      "step": 6530
    },
    {
      "epoch": 0.46888442787496415,
      "grad_norm": 0.914245437287077,
      "learning_rate": 2e-05,
      "loss": 0.5042996883392334,
      "step": 6540
    },
    {
      "epoch": 0.46960137654143963,
      "grad_norm": 1.114511447216895,
      "learning_rate": 2e-05,
      "loss": 0.5273276329040527,
      "step": 6550
    },
    {
      "epoch": 0.4703183252079151,
      "grad_norm": 0.8744276871783305,
      "learning_rate": 2e-05,
      "loss": 0.5131682395935059,
      "step": 6560
    },
    {
      "epoch": 0.4710352738743906,
      "grad_norm": 0.8567413997334786,
      "learning_rate": 2e-05,
      "loss": 0.5231666564941406,
      "step": 6570
    },
    {
      "epoch": 0.4717522225408661,
      "grad_norm": 1.9413217684940853,
      "learning_rate": 2e-05,
      "loss": 0.5153170108795166,
      "step": 6580
    },
    {
      "epoch": 0.47246917120734155,
      "grad_norm": 0.9262964065819531,
      "learning_rate": 2e-05,
      "loss": 0.5102490901947021,
      "step": 6590
    },
    {
      "epoch": 0.47318611987381703,
      "grad_norm": 1.223548208374445,
      "learning_rate": 2e-05,
      "loss": 0.5442815303802491,
      "step": 6600
    },
    {
      "epoch": 0.4739030685402925,
      "grad_norm": 0.8003303106594456,
      "learning_rate": 2e-05,
      "loss": 0.527792501449585,
      "step": 6610
    },
    {
      "epoch": 0.474620017206768,
      "grad_norm": 0.6893012633076322,
      "learning_rate": 2e-05,
      "loss": 0.5056752681732177,
      "step": 6620
    },
    {
      "epoch": 0.4753369658732435,
      "grad_norm": 1.2900210804247279,
      "learning_rate": 2e-05,
      "loss": 0.5184678554534912,
      "step": 6630
    },
    {
      "epoch": 0.47605391453971896,
      "grad_norm": 0.8156554990785718,
      "learning_rate": 2e-05,
      "loss": 0.5079127788543701,
      "step": 6640
    },
    {
      "epoch": 0.47677086320619444,
      "grad_norm": 0.9585863649258285,
      "learning_rate": 2e-05,
      "loss": 0.48990840911865235,
      "step": 6650
    },
    {
      "epoch": 0.4774878118726699,
      "grad_norm": 3.305323997447983,
      "learning_rate": 2e-05,
      "loss": 0.5125312805175781,
      "step": 6660
    },
    {
      "epoch": 0.4782047605391454,
      "grad_norm": 1.0340400827310727,
      "learning_rate": 2e-05,
      "loss": 0.5101258277893066,
      "step": 6670
    },
    {
      "epoch": 0.4789217092056209,
      "grad_norm": 1.12446504975166,
      "learning_rate": 2e-05,
      "loss": 0.5349960327148438,
      "step": 6680
    },
    {
      "epoch": 0.47963865787209636,
      "grad_norm": 0.9569680936491971,
      "learning_rate": 2e-05,
      "loss": 0.5079183578491211,
      "step": 6690
    },
    {
      "epoch": 0.48035560653857184,
      "grad_norm": 2.1527750138046735,
      "learning_rate": 2e-05,
      "loss": 0.4912437438964844,
      "step": 6700
    },
    {
      "epoch": 0.4810725552050473,
      "grad_norm": 0.9394252757754379,
      "learning_rate": 2e-05,
      "loss": 0.5158671379089356,
      "step": 6710
    },
    {
      "epoch": 0.4817895038715228,
      "grad_norm": 1.4194644774982483,
      "learning_rate": 2e-05,
      "loss": 0.5487950325012207,
      "step": 6720
    },
    {
      "epoch": 0.4825064525379983,
      "grad_norm": 1.034270137467919,
      "learning_rate": 2e-05,
      "loss": 0.5411038875579834,
      "step": 6730
    },
    {
      "epoch": 0.48322340120447377,
      "grad_norm": 0.8261537017835807,
      "learning_rate": 2e-05,
      "loss": 0.514849853515625,
      "step": 6740
    },
    {
      "epoch": 0.48394034987094925,
      "grad_norm": 2.4507797046825086,
      "learning_rate": 2e-05,
      "loss": 0.5033552169799804,
      "step": 6750
    },
    {
      "epoch": 0.48465729853742473,
      "grad_norm": 0.8429180865593078,
      "learning_rate": 2e-05,
      "loss": 0.5237858295440674,
      "step": 6760
    },
    {
      "epoch": 0.4853742472039002,
      "grad_norm": 0.8167192125691962,
      "learning_rate": 2e-05,
      "loss": 0.5268121719360351,
      "step": 6770
    },
    {
      "epoch": 0.4860911958703757,
      "grad_norm": 1.0254213611073861,
      "learning_rate": 2e-05,
      "loss": 0.5140989303588868,
      "step": 6780
    },
    {
      "epoch": 0.4868081445368512,
      "grad_norm": 0.9869884323472027,
      "learning_rate": 2e-05,
      "loss": 0.5335724830627442,
      "step": 6790
    },
    {
      "epoch": 0.48752509320332665,
      "grad_norm": 0.9590046372187695,
      "learning_rate": 2e-05,
      "loss": 0.5102479457855225,
      "step": 6800
    },
    {
      "epoch": 0.48824204186980213,
      "grad_norm": 0.9533773951459288,
      "learning_rate": 2e-05,
      "loss": 0.5216955184936524,
      "step": 6810
    },
    {
      "epoch": 0.4889589905362776,
      "grad_norm": 0.7667977878008911,
      "learning_rate": 2e-05,
      "loss": 0.4987051010131836,
      "step": 6820
    },
    {
      "epoch": 0.4896759392027531,
      "grad_norm": 1.2946013371405336,
      "learning_rate": 2e-05,
      "loss": 0.48003325462341306,
      "step": 6830
    },
    {
      "epoch": 0.4903928878692286,
      "grad_norm": 0.7763426327191401,
      "learning_rate": 2e-05,
      "loss": 0.5237167358398438,
      "step": 6840
    },
    {
      "epoch": 0.49110983653570406,
      "grad_norm": 0.9485276540895917,
      "learning_rate": 2e-05,
      "loss": 0.5073770046234131,
      "step": 6850
    },
    {
      "epoch": 0.49182678520217954,
      "grad_norm": 1.1721431122134778,
      "learning_rate": 2e-05,
      "loss": 0.5083352088928222,
      "step": 6860
    },
    {
      "epoch": 0.492543733868655,
      "grad_norm": 1.237526087468327,
      "learning_rate": 2e-05,
      "loss": 0.5023940086364747,
      "step": 6870
    },
    {
      "epoch": 0.4932606825351305,
      "grad_norm": 0.664152455742416,
      "learning_rate": 2e-05,
      "loss": 0.5155488014221191,
      "step": 6880
    },
    {
      "epoch": 0.493977631201606,
      "grad_norm": 0.8592717280854,
      "learning_rate": 2e-05,
      "loss": 0.5230101108551025,
      "step": 6890
    },
    {
      "epoch": 0.49469457986808146,
      "grad_norm": 0.9116677370127618,
      "learning_rate": 2e-05,
      "loss": 0.4980932235717773,
      "step": 6900
    },
    {
      "epoch": 0.49541152853455694,
      "grad_norm": 0.885488529461993,
      "learning_rate": 2e-05,
      "loss": 0.5346492767333985,
      "step": 6910
    },
    {
      "epoch": 0.4961284772010324,
      "grad_norm": 0.8785544562666268,
      "learning_rate": 2e-05,
      "loss": 0.48756990432739256,
      "step": 6920
    },
    {
      "epoch": 0.4968454258675079,
      "grad_norm": 1.2420959500518514,
      "learning_rate": 2e-05,
      "loss": 0.516632080078125,
      "step": 6930
    },
    {
      "epoch": 0.4975623745339834,
      "grad_norm": 0.7934228722568283,
      "learning_rate": 2e-05,
      "loss": 0.4784632682800293,
      "step": 6940
    },
    {
      "epoch": 0.49827932320045887,
      "grad_norm": 1.6273032798687854,
      "learning_rate": 2e-05,
      "loss": 0.5210093975067138,
      "step": 6950
    },
    {
      "epoch": 0.49899627186693435,
      "grad_norm": 1.0637269036425454,
      "learning_rate": 2e-05,
      "loss": 0.501754903793335,
      "step": 6960
    },
    {
      "epoch": 0.49971322053340983,
      "grad_norm": 0.9400321702645511,
      "learning_rate": 2e-05,
      "loss": 0.4960771560668945,
      "step": 6970
    },
    {
      "epoch": 0.5004301691998853,
      "grad_norm": 0.945171775450658,
      "learning_rate": 2e-05,
      "loss": 0.5388191223144532,
      "step": 6980
    },
    {
      "epoch": 0.5011471178663608,
      "grad_norm": 0.7190568899863233,
      "learning_rate": 2e-05,
      "loss": 0.49685091972351075,
      "step": 6990
    },
    {
      "epoch": 0.5018640665328362,
      "grad_norm": 0.7303379476300604,
      "learning_rate": 2e-05,
      "loss": 0.5187771797180176,
      "step": 7000
    },
    {
      "epoch": 0.5025810151993118,
      "grad_norm": 0.9431313720858298,
      "learning_rate": 2e-05,
      "loss": 0.5156877994537353,
      "step": 7010
    },
    {
      "epoch": 0.5032979638657872,
      "grad_norm": 1.1408701158475487,
      "learning_rate": 2e-05,
      "loss": 0.553832721710205,
      "step": 7020
    },
    {
      "epoch": 0.5040149125322627,
      "grad_norm": 1.1643370183698996,
      "learning_rate": 2e-05,
      "loss": 0.5052117347717285,
      "step": 7030
    },
    {
      "epoch": 0.5047318611987381,
      "grad_norm": 1.0947876577343114,
      "learning_rate": 2e-05,
      "loss": 0.4872382164001465,
      "step": 7040
    },
    {
      "epoch": 0.5054488098652137,
      "grad_norm": 0.9592525049860846,
      "learning_rate": 2e-05,
      "loss": 0.5279234886169434,
      "step": 7050
    },
    {
      "epoch": 0.5061657585316891,
      "grad_norm": 0.8337674209147588,
      "learning_rate": 2e-05,
      "loss": 0.4857050895690918,
      "step": 7060
    },
    {
      "epoch": 0.5068827071981646,
      "grad_norm": 0.7777814335232993,
      "learning_rate": 2e-05,
      "loss": 0.5099326133728027,
      "step": 7070
    },
    {
      "epoch": 0.5075996558646401,
      "grad_norm": 1.4905242363878188,
      "learning_rate": 2e-05,
      "loss": 0.5039416313171386,
      "step": 7080
    },
    {
      "epoch": 0.5083166045311156,
      "grad_norm": 0.8438240220887727,
      "learning_rate": 2e-05,
      "loss": 0.49697175025939944,
      "step": 7090
    },
    {
      "epoch": 0.509033553197591,
      "grad_norm": 1.1344419843292088,
      "learning_rate": 2e-05,
      "loss": 0.5170818328857422,
      "step": 7100
    },
    {
      "epoch": 0.5097505018640666,
      "grad_norm": 0.9873672165221876,
      "learning_rate": 2e-05,
      "loss": 0.49933176040649413,
      "step": 7110
    },
    {
      "epoch": 0.510467450530542,
      "grad_norm": 0.8965660031572235,
      "learning_rate": 2e-05,
      "loss": 0.5021637916564942,
      "step": 7120
    },
    {
      "epoch": 0.5111843991970175,
      "grad_norm": 1.0803463008547052,
      "learning_rate": 2e-05,
      "loss": 0.49392271041870117,
      "step": 7130
    },
    {
      "epoch": 0.511901347863493,
      "grad_norm": 0.9009334900088832,
      "learning_rate": 2e-05,
      "loss": 0.4904334545135498,
      "step": 7140
    },
    {
      "epoch": 0.5126182965299685,
      "grad_norm": 1.1798123344704177,
      "learning_rate": 2e-05,
      "loss": 0.511939287185669,
      "step": 7150
    },
    {
      "epoch": 0.5133352451964439,
      "grad_norm": 0.856507376276902,
      "learning_rate": 2e-05,
      "loss": 0.5154199123382568,
      "step": 7160
    },
    {
      "epoch": 0.5140521938629194,
      "grad_norm": 0.781291599029181,
      "learning_rate": 2e-05,
      "loss": 0.4862790584564209,
      "step": 7170
    },
    {
      "epoch": 0.5147691425293949,
      "grad_norm": 1.5893219106081728,
      "learning_rate": 2e-05,
      "loss": 0.512811279296875,
      "step": 7180
    },
    {
      "epoch": 0.5154860911958704,
      "grad_norm": 3.7677496153816405,
      "learning_rate": 2e-05,
      "loss": 0.49882097244262696,
      "step": 7190
    },
    {
      "epoch": 0.5162030398623458,
      "grad_norm": 1.093576259631217,
      "learning_rate": 2e-05,
      "loss": 0.4858896732330322,
      "step": 7200
    },
    {
      "epoch": 0.5169199885288214,
      "grad_norm": 1.337940230337571,
      "learning_rate": 2e-05,
      "loss": 0.4950740337371826,
      "step": 7210
    },
    {
      "epoch": 0.5176369371952968,
      "grad_norm": 0.8405812329941359,
      "learning_rate": 2e-05,
      "loss": 0.5115968227386475,
      "step": 7220
    },
    {
      "epoch": 0.5183538858617723,
      "grad_norm": 0.9600016078487494,
      "learning_rate": 2e-05,
      "loss": 0.4990074157714844,
      "step": 7230
    },
    {
      "epoch": 0.5190708345282478,
      "grad_norm": 0.8475271147589586,
      "learning_rate": 2e-05,
      "loss": 0.5182011604309082,
      "step": 7240
    },
    {
      "epoch": 0.5197877831947233,
      "grad_norm": 0.890306575687305,
      "learning_rate": 2e-05,
      "loss": 0.49146451950073244,
      "step": 7250
    },
    {
      "epoch": 0.5205047318611987,
      "grad_norm": 0.7590678000498192,
      "learning_rate": 2e-05,
      "loss": 0.5209455490112305,
      "step": 7260
    },
    {
      "epoch": 0.5212216805276743,
      "grad_norm": 0.8018659306404071,
      "learning_rate": 2e-05,
      "loss": 0.4961942195892334,
      "step": 7270
    },
    {
      "epoch": 0.5219386291941497,
      "grad_norm": 0.8312867101294389,
      "learning_rate": 2e-05,
      "loss": 0.5412381649017334,
      "step": 7280
    },
    {
      "epoch": 0.5226555778606252,
      "grad_norm": 1.8343427819707494,
      "learning_rate": 2e-05,
      "loss": 0.48026580810546876,
      "step": 7290
    },
    {
      "epoch": 0.5233725265271006,
      "grad_norm": 1.1390215453603159,
      "learning_rate": 2e-05,
      "loss": 0.5043776988983154,
      "step": 7300
    },
    {
      "epoch": 0.5240894751935762,
      "grad_norm": 1.5519866426562656,
      "learning_rate": 2e-05,
      "loss": 0.5075428009033203,
      "step": 7310
    },
    {
      "epoch": 0.5248064238600516,
      "grad_norm": 1.2379005839853365,
      "learning_rate": 2e-05,
      "loss": 0.49086637496948243,
      "step": 7320
    },
    {
      "epoch": 0.5255233725265271,
      "grad_norm": 3.4644995050964544,
      "learning_rate": 2e-05,
      "loss": 0.5167254447937012,
      "step": 7330
    },
    {
      "epoch": 0.5262403211930026,
      "grad_norm": 0.9435240929087276,
      "learning_rate": 2e-05,
      "loss": 0.523456335067749,
      "step": 7340
    },
    {
      "epoch": 0.5269572698594781,
      "grad_norm": 1.5417884879057366,
      "learning_rate": 2e-05,
      "loss": 0.5249011993408204,
      "step": 7350
    },
    {
      "epoch": 0.5276742185259535,
      "grad_norm": 0.9059896640451209,
      "learning_rate": 2e-05,
      "loss": 0.5151700496673584,
      "step": 7360
    },
    {
      "epoch": 0.5283911671924291,
      "grad_norm": 1.324295325606967,
      "learning_rate": 2e-05,
      "loss": 0.504646635055542,
      "step": 7370
    },
    {
      "epoch": 0.5291081158589045,
      "grad_norm": 1.6780254772859968,
      "learning_rate": 2e-05,
      "loss": 0.5104289531707764,
      "step": 7380
    },
    {
      "epoch": 0.52982506452538,
      "grad_norm": 1.267386693637113,
      "learning_rate": 2e-05,
      "loss": 0.5138986110687256,
      "step": 7390
    },
    {
      "epoch": 0.5305420131918555,
      "grad_norm": 0.9427366338405564,
      "learning_rate": 2e-05,
      "loss": 0.4873779296875,
      "step": 7400
    },
    {
      "epoch": 0.531258961858331,
      "grad_norm": 2.7358585969705085,
      "learning_rate": 2e-05,
      "loss": 0.5050318241119385,
      "step": 7410
    },
    {
      "epoch": 0.5319759105248064,
      "grad_norm": 1.0315463354983205,
      "learning_rate": 2e-05,
      "loss": 0.5021390914916992,
      "step": 7420
    },
    {
      "epoch": 0.532692859191282,
      "grad_norm": 2.07369160314182,
      "learning_rate": 2e-05,
      "loss": 0.48949599266052246,
      "step": 7430
    },
    {
      "epoch": 0.5334098078577574,
      "grad_norm": 2.534291561348822,
      "learning_rate": 2e-05,
      "loss": 0.5038779258728028,
      "step": 7440
    },
    {
      "epoch": 0.5341267565242329,
      "grad_norm": 1.437447219235605,
      "learning_rate": 2e-05,
      "loss": 0.5140152931213379,
      "step": 7450
    },
    {
      "epoch": 0.5348437051907083,
      "grad_norm": 3.197835803720907,
      "learning_rate": 2e-05,
      "loss": 0.5290528297424316,
      "step": 7460
    },
    {
      "epoch": 0.5355606538571839,
      "grad_norm": 1.1160841503782417,
      "learning_rate": 2e-05,
      "loss": 0.49622058868408203,
      "step": 7470
    },
    {
      "epoch": 0.5362776025236593,
      "grad_norm": 0.8875648487048132,
      "learning_rate": 2e-05,
      "loss": 0.48895511627197263,
      "step": 7480
    },
    {
      "epoch": 0.5369945511901348,
      "grad_norm": 1.2018909735585965,
      "learning_rate": 2e-05,
      "loss": 0.5085635185241699,
      "step": 7490
    },
    {
      "epoch": 0.5377114998566103,
      "grad_norm": 1.355543534689094,
      "learning_rate": 2e-05,
      "loss": 0.5200148105621338,
      "step": 7500
    },
    {
      "epoch": 0.5384284485230858,
      "grad_norm": 20.100588463033027,
      "learning_rate": 2e-05,
      "loss": 0.4857450008392334,
      "step": 7510
    },
    {
      "epoch": 0.5391453971895612,
      "grad_norm": 0.827240431959633,
      "learning_rate": 2e-05,
      "loss": 0.5287868022918701,
      "step": 7520
    },
    {
      "epoch": 0.5398623458560368,
      "grad_norm": 1.8886206676195867,
      "learning_rate": 2e-05,
      "loss": 0.5059417724609375,
      "step": 7530
    },
    {
      "epoch": 0.5405792945225122,
      "grad_norm": 1.2542636055509073,
      "learning_rate": 2e-05,
      "loss": 0.5080095291137695,
      "step": 7540
    },
    {
      "epoch": 0.5412962431889876,
      "grad_norm": 1.1427117560470867,
      "learning_rate": 2e-05,
      "loss": 0.533676290512085,
      "step": 7550
    },
    {
      "epoch": 0.5420131918554632,
      "grad_norm": 0.7080848159122278,
      "learning_rate": 2e-05,
      "loss": 0.5090786933898925,
      "step": 7560
    },
    {
      "epoch": 0.5427301405219386,
      "grad_norm": 8.025546452517998,
      "learning_rate": 2e-05,
      "loss": 0.49716987609863283,
      "step": 7570
    },
    {
      "epoch": 0.5434470891884141,
      "grad_norm": 1.4862624559610198,
      "learning_rate": 2e-05,
      "loss": 0.48635663986206057,
      "step": 7580
    },
    {
      "epoch": 0.5441640378548895,
      "grad_norm": 3.715783763583897,
      "learning_rate": 2e-05,
      "loss": 0.4934207916259766,
      "step": 7590
    },
    {
      "epoch": 0.5448809865213651,
      "grad_norm": 0.9627483737514896,
      "learning_rate": 2e-05,
      "loss": 0.5474603176116943,
      "step": 7600
    },
    {
      "epoch": 0.5455979351878405,
      "grad_norm": 0.8094429098423439,
      "learning_rate": 2e-05,
      "loss": 0.5129518508911133,
      "step": 7610
    },
    {
      "epoch": 0.546314883854316,
      "grad_norm": 1.3597577388116278,
      "learning_rate": 2e-05,
      "loss": 0.49813342094421387,
      "step": 7620
    },
    {
      "epoch": 0.5470318325207915,
      "grad_norm": 0.6922016015666678,
      "learning_rate": 2e-05,
      "loss": 0.4728724002838135,
      "step": 7630
    },
    {
      "epoch": 0.547748781187267,
      "grad_norm": 8.681587574928516,
      "learning_rate": 2e-05,
      "loss": 0.4948934018611908,
      "step": 7640
    },
    {
      "epoch": 0.5484657298537424,
      "grad_norm": 0.9499721042015875,
      "learning_rate": 2e-05,
      "loss": 0.5053561687469482,
      "step": 7650
    },
    {
      "epoch": 0.549182678520218,
      "grad_norm": 1.0701062984129603,
      "learning_rate": 2e-05,
      "loss": 0.5295053005218506,
      "step": 7660
    },
    {
      "epoch": 0.5498996271866934,
      "grad_norm": 2.4545644183265307,
      "learning_rate": 2e-05,
      "loss": 0.5243900299072266,
      "step": 7670
    },
    {
      "epoch": 0.5506165758531689,
      "grad_norm": 0.7795778206962374,
      "learning_rate": 2e-05,
      "loss": 0.5420596122741699,
      "step": 7680
    },
    {
      "epoch": 0.5513335245196443,
      "grad_norm": 0.8578511181968267,
      "learning_rate": 2e-05,
      "loss": 0.5127090454101563,
      "step": 7690
    },
    {
      "epoch": 0.5520504731861199,
      "grad_norm": 0.7258690549896266,
      "learning_rate": 2e-05,
      "loss": 0.5011856079101562,
      "step": 7700
    },
    {
      "epoch": 0.5527674218525953,
      "grad_norm": 2.066618497995809,
      "learning_rate": 2e-05,
      "loss": 0.47744102478027345,
      "step": 7710
    },
    {
      "epoch": 0.5534843705190708,
      "grad_norm": 0.7298994427238686,
      "learning_rate": 2e-05,
      "loss": 0.5072267532348633,
      "step": 7720
    },
    {
      "epoch": 0.5542013191855463,
      "grad_norm": 0.6224070479805619,
      "learning_rate": 2e-05,
      "loss": 0.5199421882629395,
      "step": 7730
    },
    {
      "epoch": 0.5549182678520218,
      "grad_norm": 0.9925933461325204,
      "learning_rate": 2e-05,
      "loss": 0.4945536613464355,
      "step": 7740
    },
    {
      "epoch": 0.5556352165184972,
      "grad_norm": 0.796405985828255,
      "learning_rate": 2e-05,
      "loss": 0.5323121070861816,
      "step": 7750
    },
    {
      "epoch": 0.5563521651849728,
      "grad_norm": 6.163455961995889,
      "learning_rate": 2e-05,
      "loss": 0.5118585109710694,
      "step": 7760
    },
    {
      "epoch": 0.5570691138514482,
      "grad_norm": 0.741935113930321,
      "learning_rate": 2e-05,
      "loss": 0.5000914096832275,
      "step": 7770
    },
    {
      "epoch": 0.5577860625179237,
      "grad_norm": 1.0005105992745862,
      "learning_rate": 2e-05,
      "loss": 0.49091429710388185,
      "step": 7780
    },
    {
      "epoch": 0.5585030111843992,
      "grad_norm": 0.716793829233197,
      "learning_rate": 2e-05,
      "loss": 0.5042989730834961,
      "step": 7790
    },
    {
      "epoch": 0.5592199598508747,
      "grad_norm": 0.819696566963475,
      "learning_rate": 2e-05,
      "loss": 0.5183305740356445,
      "step": 7800
    },
    {
      "epoch": 0.5599369085173501,
      "grad_norm": 0.8735036052305644,
      "learning_rate": 2e-05,
      "loss": 0.5013885498046875,
      "step": 7810
    },
    {
      "epoch": 0.5606538571838257,
      "grad_norm": 0.7666990754545703,
      "learning_rate": 2e-05,
      "loss": 0.495221471786499,
      "step": 7820
    },
    {
      "epoch": 0.5613708058503011,
      "grad_norm": 0.9252070500581573,
      "learning_rate": 2e-05,
      "loss": 0.5057270050048828,
      "step": 7830
    },
    {
      "epoch": 0.5620877545167766,
      "grad_norm": 1.0218238586691153,
      "learning_rate": 2e-05,
      "loss": 0.5163712501525879,
      "step": 7840
    },
    {
      "epoch": 0.562804703183252,
      "grad_norm": 0.7028839254677526,
      "learning_rate": 2e-05,
      "loss": 0.5028861045837403,
      "step": 7850
    },
    {
      "epoch": 0.5635216518497276,
      "grad_norm": 2.601591819955894,
      "learning_rate": 2e-05,
      "loss": 0.49895524978637695,
      "step": 7860
    },
    {
      "epoch": 0.564238600516203,
      "grad_norm": 0.6988627563079403,
      "learning_rate": 2e-05,
      "loss": 0.4983854293823242,
      "step": 7870
    },
    {
      "epoch": 0.5649555491826785,
      "grad_norm": 0.6658695966659398,
      "learning_rate": 2e-05,
      "loss": 0.5105714321136474,
      "step": 7880
    },
    {
      "epoch": 0.565672497849154,
      "grad_norm": 0.8054889262131978,
      "learning_rate": 2e-05,
      "loss": 0.495974063873291,
      "step": 7890
    },
    {
      "epoch": 0.5663894465156295,
      "grad_norm": 0.858103305203879,
      "learning_rate": 2e-05,
      "loss": 0.506614875793457,
      "step": 7900
    },
    {
      "epoch": 0.5671063951821049,
      "grad_norm": 0.7786854580466089,
      "learning_rate": 2e-05,
      "loss": 0.4936980247497559,
      "step": 7910
    },
    {
      "epoch": 0.5678233438485805,
      "grad_norm": 0.8454915905547884,
      "learning_rate": 2e-05,
      "loss": 0.4840365409851074,
      "step": 7920
    },
    {
      "epoch": 0.5685402925150559,
      "grad_norm": 1.2049235512992624,
      "learning_rate": 2e-05,
      "loss": 0.5055620193481445,
      "step": 7930
    },
    {
      "epoch": 0.5692572411815314,
      "grad_norm": 0.7382077609639357,
      "learning_rate": 2e-05,
      "loss": 0.4920199394226074,
      "step": 7940
    },
    {
      "epoch": 0.5699741898480069,
      "grad_norm": 1.3776398921529516,
      "learning_rate": 2e-05,
      "loss": 0.49503183364868164,
      "step": 7950
    },
    {
      "epoch": 0.5706911385144824,
      "grad_norm": 0.9734022459029245,
      "learning_rate": 2e-05,
      "loss": 0.4866753578186035,
      "step": 7960
    },
    {
      "epoch": 0.5714080871809578,
      "grad_norm": 2.3877179062613396,
      "learning_rate": 2e-05,
      "loss": 0.5234427452087402,
      "step": 7970
    },
    {
      "epoch": 0.5721250358474333,
      "grad_norm": 0.7900779792874013,
      "learning_rate": 2e-05,
      "loss": 0.49785447120666504,
      "step": 7980
    },
    {
      "epoch": 0.5728419845139088,
      "grad_norm": 0.8104191321625626,
      "learning_rate": 2e-05,
      "loss": 0.49553427696228025,
      "step": 7990
    },
    {
      "epoch": 0.5735589331803843,
      "grad_norm": 0.8641416303353244,
      "learning_rate": 2e-05,
      "loss": 0.48295068740844727,
      "step": 8000
    },
    {
      "epoch": 0.5742758818468597,
      "grad_norm": 1.6184658035015733,
      "learning_rate": 2e-05,
      "loss": 0.47038559913635253,
      "step": 8010
    },
    {
      "epoch": 0.5749928305133353,
      "grad_norm": 0.7643017271043779,
      "learning_rate": 2e-05,
      "loss": 0.48970727920532225,
      "step": 8020
    },
    {
      "epoch": 0.5757097791798107,
      "grad_norm": 3.781943237858509,
      "learning_rate": 2e-05,
      "loss": 0.4899750709533691,
      "step": 8030
    },
    {
      "epoch": 0.5764267278462862,
      "grad_norm": 2.6197669905508394,
      "learning_rate": 2e-05,
      "loss": 0.48921966552734375,
      "step": 8040
    },
    {
      "epoch": 0.5771436765127617,
      "grad_norm": 1.00360664864747,
      "learning_rate": 2e-05,
      "loss": 0.4868148326873779,
      "step": 8050
    },
    {
      "epoch": 0.5778606251792372,
      "grad_norm": 0.8586499460755999,
      "learning_rate": 2e-05,
      "loss": 0.5158542156219482,
      "step": 8060
    },
    {
      "epoch": 0.5785775738457126,
      "grad_norm": 1.4426759297736524,
      "learning_rate": 2e-05,
      "loss": 0.5257744789123535,
      "step": 8070
    },
    {
      "epoch": 0.5792945225121882,
      "grad_norm": 4.891395588030371,
      "learning_rate": 2e-05,
      "loss": 0.49913716316223145,
      "step": 8080
    },
    {
      "epoch": 0.5800114711786636,
      "grad_norm": 1.5822701222576934,
      "learning_rate": 2e-05,
      "loss": 0.48695898056030273,
      "step": 8090
    },
    {
      "epoch": 0.5807284198451391,
      "grad_norm": 0.9229493174456277,
      "learning_rate": 2e-05,
      "loss": 0.5049197196960449,
      "step": 8100
    },
    {
      "epoch": 0.5814453685116145,
      "grad_norm": 0.7994274701027114,
      "learning_rate": 2e-05,
      "loss": 0.49312148094177244,
      "step": 8110
    },
    {
      "epoch": 0.5821623171780901,
      "grad_norm": 0.8026390861675082,
      "learning_rate": 2e-05,
      "loss": 0.501448106765747,
      "step": 8120
    },
    {
      "epoch": 0.5828792658445655,
      "grad_norm": 0.789714745609719,
      "learning_rate": 2e-05,
      "loss": 0.4937129974365234,
      "step": 8130
    },
    {
      "epoch": 0.583596214511041,
      "grad_norm": 8.245781775839426,
      "learning_rate": 2e-05,
      "loss": 0.5383259773254394,
      "step": 8140
    },
    {
      "epoch": 0.5843131631775165,
      "grad_norm": 1.5813264704882253,
      "learning_rate": 2e-05,
      "loss": 0.47626848220825196,
      "step": 8150
    },
    {
      "epoch": 0.585030111843992,
      "grad_norm": 0.6953878672301901,
      "learning_rate": 2e-05,
      "loss": 0.5081315040588379,
      "step": 8160
    },
    {
      "epoch": 0.5857470605104674,
      "grad_norm": 0.6987724213958244,
      "learning_rate": 2e-05,
      "loss": 0.5077291011810303,
      "step": 8170
    },
    {
      "epoch": 0.586464009176943,
      "grad_norm": 0.7111041391782621,
      "learning_rate": 2e-05,
      "loss": 0.4715399742126465,
      "step": 8180
    },
    {
      "epoch": 0.5871809578434184,
      "grad_norm": 1.109758470430622,
      "learning_rate": 2e-05,
      "loss": 0.5211437225341797,
      "step": 8190
    },
    {
      "epoch": 0.5878979065098939,
      "grad_norm": 0.7758838637487026,
      "learning_rate": 2e-05,
      "loss": 0.51683669090271,
      "step": 8200
    },
    {
      "epoch": 0.5886148551763694,
      "grad_norm": 1.0594266817843387,
      "learning_rate": 2e-05,
      "loss": 0.48267555236816406,
      "step": 8210
    },
    {
      "epoch": 0.5893318038428449,
      "grad_norm": 0.7306970780512932,
      "learning_rate": 2e-05,
      "loss": 0.5268644332885742,
      "step": 8220
    },
    {
      "epoch": 0.5900487525093203,
      "grad_norm": 1.2383655706134744,
      "learning_rate": 2e-05,
      "loss": 0.49268856048583987,
      "step": 8230
    },
    {
      "epoch": 0.5907657011757959,
      "grad_norm": 0.7950179779268204,
      "learning_rate": 2e-05,
      "loss": 0.4886177062988281,
      "step": 8240
    },
    {
      "epoch": 0.5914826498422713,
      "grad_norm": 0.7380650942189657,
      "learning_rate": 2e-05,
      "loss": 0.5201387882232666,
      "step": 8250
    },
    {
      "epoch": 0.5921995985087468,
      "grad_norm": 0.8652157959088963,
      "learning_rate": 2e-05,
      "loss": 0.48958797454833985,
      "step": 8260
    },
    {
      "epoch": 0.5929165471752222,
      "grad_norm": 1.0108541293943996,
      "learning_rate": 2e-05,
      "loss": 0.5485910415649414,
      "step": 8270
    },
    {
      "epoch": 0.5936334958416978,
      "grad_norm": 0.8214572891979975,
      "learning_rate": 2e-05,
      "loss": 0.5144498348236084,
      "step": 8280
    },
    {
      "epoch": 0.5943504445081732,
      "grad_norm": 1.0147108851535669,
      "learning_rate": 2e-05,
      "loss": 0.49741644859313966,
      "step": 8290
    },
    {
      "epoch": 0.5950673931746487,
      "grad_norm": 0.7497464034101148,
      "learning_rate": 2e-05,
      "loss": 0.5149343490600586,
      "step": 8300
    },
    {
      "epoch": 0.5957843418411242,
      "grad_norm": 0.6197677926110264,
      "learning_rate": 2e-05,
      "loss": 0.4975454330444336,
      "step": 8310
    },
    {
      "epoch": 0.5965012905075997,
      "grad_norm": 0.7346134834534593,
      "learning_rate": 2e-05,
      "loss": 0.5016090393066406,
      "step": 8320
    },
    {
      "epoch": 0.5972182391740751,
      "grad_norm": 1.490614059015046,
      "learning_rate": 2e-05,
      "loss": 0.505299425125122,
      "step": 8330
    },
    {
      "epoch": 0.5979351878405507,
      "grad_norm": 0.7149067913928806,
      "learning_rate": 2e-05,
      "loss": 0.48895974159240724,
      "step": 8340
    },
    {
      "epoch": 0.5986521365070261,
      "grad_norm": 0.7177667677173021,
      "learning_rate": 2e-05,
      "loss": 0.4885610580444336,
      "step": 8350
    },
    {
      "epoch": 0.5993690851735016,
      "grad_norm": 1.3945278633049218,
      "learning_rate": 2e-05,
      "loss": 0.48483591079711913,
      "step": 8360
    },
    {
      "epoch": 0.600086033839977,
      "grad_norm": 0.8111210486752815,
      "learning_rate": 2e-05,
      "loss": 0.502918815612793,
      "step": 8370
    },
    {
      "epoch": 0.6008029825064526,
      "grad_norm": 0.8366487816550784,
      "learning_rate": 2e-05,
      "loss": 0.4957133769989014,
      "step": 8380
    },
    {
      "epoch": 0.601519931172928,
      "grad_norm": 0.6753849310501518,
      "learning_rate": 2e-05,
      "loss": 0.47420368194580076,
      "step": 8390
    },
    {
      "epoch": 0.6022368798394035,
      "grad_norm": 0.7432497795545511,
      "learning_rate": 2e-05,
      "loss": 0.4869903564453125,
      "step": 8400
    },
    {
      "epoch": 0.602953828505879,
      "grad_norm": 0.686560275081152,
      "learning_rate": 2e-05,
      "loss": 0.48692951202392576,
      "step": 8410
    },
    {
      "epoch": 0.6036707771723545,
      "grad_norm": 0.7145478843184155,
      "learning_rate": 2e-05,
      "loss": 0.5197232723236084,
      "step": 8420
    },
    {
      "epoch": 0.6043877258388299,
      "grad_norm": 1.418760852294346,
      "learning_rate": 2e-05,
      "loss": 0.5240144729614258,
      "step": 8430
    },
    {
      "epoch": 0.6051046745053055,
      "grad_norm": 0.6463211391770027,
      "learning_rate": 2e-05,
      "loss": 0.5151795864105224,
      "step": 8440
    },
    {
      "epoch": 0.6058216231717809,
      "grad_norm": 0.7413145537409702,
      "learning_rate": 2e-05,
      "loss": 0.5101013660430909,
      "step": 8450
    },
    {
      "epoch": 0.6065385718382564,
      "grad_norm": 0.9424617835526052,
      "learning_rate": 2e-05,
      "loss": 0.5276088714599609,
      "step": 8460
    },
    {
      "epoch": 0.6072555205047319,
      "grad_norm": 9.062678565212344,
      "learning_rate": 2e-05,
      "loss": 0.5107160091400147,
      "step": 8470
    },
    {
      "epoch": 0.6079724691712074,
      "grad_norm": 0.7702275264410343,
      "learning_rate": 2e-05,
      "loss": 0.5139708518981934,
      "step": 8480
    },
    {
      "epoch": 0.6086894178376828,
      "grad_norm": 0.9166115083919714,
      "learning_rate": 2e-05,
      "loss": 0.5447901725769043,
      "step": 8490
    },
    {
      "epoch": 0.6094063665041582,
      "grad_norm": 1.7404751625000472,
      "learning_rate": 2e-05,
      "loss": 0.5050385475158692,
      "step": 8500
    },
    {
      "epoch": 0.6101233151706338,
      "grad_norm": 2.00870577186918,
      "learning_rate": 2e-05,
      "loss": 0.48280916213989256,
      "step": 8510
    },
    {
      "epoch": 0.6108402638371092,
      "grad_norm": 1.0731292957874137,
      "learning_rate": 2e-05,
      "loss": 0.4860265254974365,
      "step": 8520
    },
    {
      "epoch": 0.6115572125035847,
      "grad_norm": 0.8458559366202995,
      "learning_rate": 2e-05,
      "loss": 0.4579441547393799,
      "step": 8530
    },
    {
      "epoch": 0.6122741611700602,
      "grad_norm": 1.0262887446258846,
      "learning_rate": 2e-05,
      "loss": 0.5118321895599365,
      "step": 8540
    },
    {
      "epoch": 0.6129911098365357,
      "grad_norm": 2.2058904124921357,
      "learning_rate": 2e-05,
      "loss": 0.5217164516448974,
      "step": 8550
    },
    {
      "epoch": 0.6137080585030111,
      "grad_norm": 1.0639599551203114,
      "learning_rate": 2e-05,
      "loss": 0.4963801860809326,
      "step": 8560
    },
    {
      "epoch": 0.6144250071694867,
      "grad_norm": 1.711251950775745,
      "learning_rate": 2e-05,
      "loss": 0.500682258605957,
      "step": 8570
    },
    {
      "epoch": 0.6151419558359621,
      "grad_norm": 0.763846610920214,
      "learning_rate": 2e-05,
      "loss": 0.5239065170288086,
      "step": 8580
    },
    {
      "epoch": 0.6158589045024376,
      "grad_norm": 7.585107834141741,
      "learning_rate": 2e-05,
      "loss": 0.5197652816772461,
      "step": 8590
    },
    {
      "epoch": 0.6165758531689131,
      "grad_norm": 1.3496379093734356,
      "learning_rate": 2e-05,
      "loss": 0.4987212657928467,
      "step": 8600
    },
    {
      "epoch": 0.6172928018353886,
      "grad_norm": 1.070386412518482,
      "learning_rate": 2e-05,
      "loss": 0.5147864341735839,
      "step": 8610
    },
    {
      "epoch": 0.618009750501864,
      "grad_norm": 0.820355908654114,
      "learning_rate": 2e-05,
      "loss": 0.5028390884399414,
      "step": 8620
    },
    {
      "epoch": 0.6187266991683396,
      "grad_norm": 0.8885160728796373,
      "learning_rate": 2e-05,
      "loss": 0.49216446876525877,
      "step": 8630
    },
    {
      "epoch": 0.619443647834815,
      "grad_norm": 1.897390693286958,
      "learning_rate": 2e-05,
      "loss": 0.5046564102172851,
      "step": 8640
    },
    {
      "epoch": 0.6201605965012905,
      "grad_norm": 1.0777342305284707,
      "learning_rate": 2e-05,
      "loss": 0.49618024826049806,
      "step": 8650
    },
    {
      "epoch": 0.6208775451677659,
      "grad_norm": 0.8131040424398489,
      "learning_rate": 2e-05,
      "loss": 0.5083162784576416,
      "step": 8660
    },
    {
      "epoch": 0.6215944938342415,
      "grad_norm": 1.1974757845151842,
      "learning_rate": 2e-05,
      "loss": 0.5039738655090332,
      "step": 8670
    },
    {
      "epoch": 0.6223114425007169,
      "grad_norm": 0.818604785138252,
      "learning_rate": 2e-05,
      "loss": 0.49582843780517577,
      "step": 8680
    },
    {
      "epoch": 0.6230283911671924,
      "grad_norm": 0.8174752319601576,
      "learning_rate": 2e-05,
      "loss": 0.4840613842010498,
      "step": 8690
    },
    {
      "epoch": 0.6237453398336679,
      "grad_norm": 0.882408548904734,
      "learning_rate": 2e-05,
      "loss": 0.4717586517333984,
      "step": 8700
    },
    {
      "epoch": 0.6244622885001434,
      "grad_norm": 1.169710241584211,
      "learning_rate": 2e-05,
      "loss": 0.48357019424438474,
      "step": 8710
    },
    {
      "epoch": 0.6251792371666188,
      "grad_norm": 0.7565056485330365,
      "learning_rate": 2e-05,
      "loss": 0.5052444458007812,
      "step": 8720
    },
    {
      "epoch": 0.6258961858330944,
      "grad_norm": 0.9194586269865175,
      "learning_rate": 2e-05,
      "loss": 0.493572998046875,
      "step": 8730
    },
    {
      "epoch": 0.6266131344995698,
      "grad_norm": 1.0447343917682692,
      "learning_rate": 2e-05,
      "loss": 0.49687972068786623,
      "step": 8740
    },
    {
      "epoch": 0.6273300831660453,
      "grad_norm": 1.107513612181719,
      "learning_rate": 2e-05,
      "loss": 0.5027109146118164,
      "step": 8750
    },
    {
      "epoch": 0.6280470318325208,
      "grad_norm": 0.7606455276731956,
      "learning_rate": 2e-05,
      "loss": 0.5113688468933105,
      "step": 8760
    },
    {
      "epoch": 0.6287639804989963,
      "grad_norm": 0.8449171972248194,
      "learning_rate": 2e-05,
      "loss": 0.5007295608520508,
      "step": 8770
    },
    {
      "epoch": 0.6294809291654717,
      "grad_norm": 1.200229423364618,
      "learning_rate": 2e-05,
      "loss": 0.4831671714782715,
      "step": 8780
    },
    {
      "epoch": 0.6301978778319473,
      "grad_norm": 0.6957885742266982,
      "learning_rate": 2e-05,
      "loss": 0.48686418533325193,
      "step": 8790
    },
    {
      "epoch": 0.6309148264984227,
      "grad_norm": 0.7396693458301739,
      "learning_rate": 2e-05,
      "loss": 0.488104248046875,
      "step": 8800
    },
    {
      "epoch": 0.6316317751648982,
      "grad_norm": 0.887049417697557,
      "learning_rate": 2e-05,
      "loss": 0.5087513446807861,
      "step": 8810
    },
    {
      "epoch": 0.6323487238313736,
      "grad_norm": 0.8826438189562795,
      "learning_rate": 2e-05,
      "loss": 0.4904178619384766,
      "step": 8820
    },
    {
      "epoch": 0.6330656724978492,
      "grad_norm": 0.9951324795024091,
      "learning_rate": 2e-05,
      "loss": 0.5177725791931153,
      "step": 8830
    },
    {
      "epoch": 0.6337826211643246,
      "grad_norm": 0.7154768196623569,
      "learning_rate": 2e-05,
      "loss": 0.5207562446594238,
      "step": 8840
    },
    {
      "epoch": 0.6344995698308001,
      "grad_norm": 0.8442573342977486,
      "learning_rate": 2e-05,
      "loss": 0.47325563430786133,
      "step": 8850
    },
    {
      "epoch": 0.6352165184972756,
      "grad_norm": 2.2239547261450627,
      "learning_rate": 2e-05,
      "loss": 0.49422545433044435,
      "step": 8860
    },
    {
      "epoch": 0.6359334671637511,
      "grad_norm": 2.7360322232916743,
      "learning_rate": 2e-05,
      "loss": 0.4994163513183594,
      "step": 8870
    },
    {
      "epoch": 0.6366504158302265,
      "grad_norm": 0.7727160943606342,
      "learning_rate": 2e-05,
      "loss": 0.4916572570800781,
      "step": 8880
    },
    {
      "epoch": 0.6373673644967021,
      "grad_norm": 0.8410335417444242,
      "learning_rate": 2e-05,
      "loss": 0.5071803569793701,
      "step": 8890
    },
    {
      "epoch": 0.6380843131631775,
      "grad_norm": 0.890570395140567,
      "learning_rate": 2e-05,
      "loss": 0.4748265266418457,
      "step": 8900
    },
    {
      "epoch": 0.638801261829653,
      "grad_norm": 0.9797339013377107,
      "learning_rate": 2e-05,
      "loss": 0.5049604415893555,
      "step": 8910
    },
    {
      "epoch": 0.6395182104961284,
      "grad_norm": 1.3150370698494158,
      "learning_rate": 2e-05,
      "loss": 0.4946861267089844,
      "step": 8920
    },
    {
      "epoch": 0.640235159162604,
      "grad_norm": 1.5378035056302786,
      "learning_rate": 2e-05,
      "loss": 0.5011652946472168,
      "step": 8930
    },
    {
      "epoch": 0.6409521078290794,
      "grad_norm": 0.9638007086009709,
      "learning_rate": 2e-05,
      "loss": 0.5003650665283204,
      "step": 8940
    },
    {
      "epoch": 0.641669056495555,
      "grad_norm": 1.365172835720884,
      "learning_rate": 2e-05,
      "loss": 0.49186954498291013,
      "step": 8950
    },
    {
      "epoch": 0.6423860051620304,
      "grad_norm": 0.7842687495413343,
      "learning_rate": 2e-05,
      "loss": 0.48912272453308103,
      "step": 8960
    },
    {
      "epoch": 0.6431029538285059,
      "grad_norm": 1.0191128678107302,
      "learning_rate": 2e-05,
      "loss": 0.49554147720336916,
      "step": 8970
    },
    {
      "epoch": 0.6438199024949813,
      "grad_norm": 0.9802454839559522,
      "learning_rate": 2e-05,
      "loss": 0.5205283164978027,
      "step": 8980
    },
    {
      "epoch": 0.6445368511614569,
      "grad_norm": 1.2588678641199018,
      "learning_rate": 2e-05,
      "loss": 0.5060201644897461,
      "step": 8990
    },
    {
      "epoch": 0.6452537998279323,
      "grad_norm": 2.4090358360514794,
      "learning_rate": 2e-05,
      "loss": 0.5261709213256835,
      "step": 9000
    },
    {
      "epoch": 0.6459707484944078,
      "grad_norm": 0.8892155423832242,
      "learning_rate": 2e-05,
      "loss": 0.48573665618896483,
      "step": 9010
    },
    {
      "epoch": 0.6466876971608833,
      "grad_norm": 1.0931335422370814,
      "learning_rate": 2e-05,
      "loss": 0.527579927444458,
      "step": 9020
    },
    {
      "epoch": 0.6474046458273588,
      "grad_norm": 1.0895758362687853,
      "learning_rate": 2e-05,
      "loss": 0.4798113822937012,
      "step": 9030
    },
    {
      "epoch": 0.6481215944938342,
      "grad_norm": 1.7059448807242572,
      "learning_rate": 2e-05,
      "loss": 0.4998103141784668,
      "step": 9040
    },
    {
      "epoch": 0.6488385431603098,
      "grad_norm": 7.7748782164247965,
      "learning_rate": 2e-05,
      "loss": 0.4720963478088379,
      "step": 9050
    },
    {
      "epoch": 0.6495554918267852,
      "grad_norm": 1.3937869974329926,
      "learning_rate": 2e-05,
      "loss": 0.5179917335510253,
      "step": 9060
    },
    {
      "epoch": 0.6502724404932607,
      "grad_norm": 1.149190461927993,
      "learning_rate": 2e-05,
      "loss": 0.5157068729400635,
      "step": 9070
    },
    {
      "epoch": 0.6509893891597361,
      "grad_norm": 1.0872772824222885,
      "learning_rate": 2e-05,
      "loss": 0.5022348880767822,
      "step": 9080
    },
    {
      "epoch": 0.6517063378262117,
      "grad_norm": 0.8016294258037026,
      "learning_rate": 2e-05,
      "loss": 0.49106802940368655,
      "step": 9090
    },
    {
      "epoch": 0.6524232864926871,
      "grad_norm": 0.9076529718875517,
      "learning_rate": 2e-05,
      "loss": 0.5109787940979004,
      "step": 9100
    },
    {
      "epoch": 0.6531402351591626,
      "grad_norm": 0.8455446691599119,
      "learning_rate": 2e-05,
      "loss": 0.5054721832275391,
      "step": 9110
    },
    {
      "epoch": 0.6538571838256381,
      "grad_norm": 0.8434195356022148,
      "learning_rate": 2e-05,
      "loss": 0.4999844074249268,
      "step": 9120
    },
    {
      "epoch": 0.6545741324921136,
      "grad_norm": 0.7008928600794765,
      "learning_rate": 2e-05,
      "loss": 0.47991719245910647,
      "step": 9130
    },
    {
      "epoch": 0.655291081158589,
      "grad_norm": 1.1486442056733266,
      "learning_rate": 2e-05,
      "loss": 0.5143932819366455,
      "step": 9140
    },
    {
      "epoch": 0.6560080298250646,
      "grad_norm": 0.860059061858592,
      "learning_rate": 2e-05,
      "loss": 0.4804642677307129,
      "step": 9150
    },
    {
      "epoch": 0.65672497849154,
      "grad_norm": 0.8418092609306933,
      "learning_rate": 2e-05,
      "loss": 0.49195475578308107,
      "step": 9160
    },
    {
      "epoch": 0.6574419271580155,
      "grad_norm": 1.0221179546525605,
      "learning_rate": 2e-05,
      "loss": 0.4720583915710449,
      "step": 9170
    },
    {
      "epoch": 0.658158875824491,
      "grad_norm": 1.0446602231224167,
      "learning_rate": 2e-05,
      "loss": 0.5044754028320313,
      "step": 9180
    },
    {
      "epoch": 0.6588758244909665,
      "grad_norm": 1.2473455468865258,
      "learning_rate": 2e-05,
      "loss": 0.490444278717041,
      "step": 9190
    },
    {
      "epoch": 0.6595927731574419,
      "grad_norm": 0.714508378048821,
      "learning_rate": 2e-05,
      "loss": 0.5201548099517822,
      "step": 9200
    },
    {
      "epoch": 0.6603097218239175,
      "grad_norm": 0.8890330733117219,
      "learning_rate": 2e-05,
      "loss": 0.4897912502288818,
      "step": 9210
    },
    {
      "epoch": 0.6610266704903929,
      "grad_norm": 0.6916428366970486,
      "learning_rate": 2e-05,
      "loss": 0.4978433609008789,
      "step": 9220
    },
    {
      "epoch": 0.6617436191568684,
      "grad_norm": 1.1854272996231594,
      "learning_rate": 2e-05,
      "loss": 0.47353010177612304,
      "step": 9230
    },
    {
      "epoch": 0.6624605678233438,
      "grad_norm": 1.3386818723902099,
      "learning_rate": 2e-05,
      "loss": 0.5273138999938964,
      "step": 9240
    },
    {
      "epoch": 0.6631775164898194,
      "grad_norm": 1.0219430620358145,
      "learning_rate": 2e-05,
      "loss": 0.49330949783325195,
      "step": 9250
    },
    {
      "epoch": 0.6638944651562948,
      "grad_norm": 0.695965898822817,
      "learning_rate": 2e-05,
      "loss": 0.4677430152893066,
      "step": 9260
    },
    {
      "epoch": 0.6646114138227703,
      "grad_norm": 0.8872328571810546,
      "learning_rate": 2e-05,
      "loss": 0.4958826541900635,
      "step": 9270
    },
    {
      "epoch": 0.6653283624892458,
      "grad_norm": 0.895036771012448,
      "learning_rate": 2e-05,
      "loss": 0.5159953594207763,
      "step": 9280
    },
    {
      "epoch": 0.6660453111557213,
      "grad_norm": 0.9083165279088139,
      "learning_rate": 2e-05,
      "loss": 0.49268455505371095,
      "step": 9290
    },
    {
      "epoch": 0.6667622598221967,
      "grad_norm": 0.7220901519371575,
      "learning_rate": 2e-05,
      "loss": 0.49859933853149413,
      "step": 9300
    },
    {
      "epoch": 0.6674792084886723,
      "grad_norm": 0.795291400984382,
      "learning_rate": 2e-05,
      "loss": 0.4952845573425293,
      "step": 9310
    },
    {
      "epoch": 0.6681961571551477,
      "grad_norm": 3.046293126999376,
      "learning_rate": 2e-05,
      "loss": 0.5038476467132569,
      "step": 9320
    },
    {
      "epoch": 0.6689131058216232,
      "grad_norm": 0.690093663278623,
      "learning_rate": 2e-05,
      "loss": 0.4823905944824219,
      "step": 9330
    },
    {
      "epoch": 0.6696300544880986,
      "grad_norm": 0.8471317432352331,
      "learning_rate": 2e-05,
      "loss": 0.523288631439209,
      "step": 9340
    },
    {
      "epoch": 0.6703470031545742,
      "grad_norm": 0.6386446113635933,
      "learning_rate": 2e-05,
      "loss": 0.4919003486633301,
      "step": 9350
    },
    {
      "epoch": 0.6710639518210496,
      "grad_norm": 0.807904901173998,
      "learning_rate": 2e-05,
      "loss": 0.5034087657928467,
      "step": 9360
    },
    {
      "epoch": 0.6717809004875251,
      "grad_norm": 0.8722297314282209,
      "learning_rate": 2e-05,
      "loss": 0.5000992774963379,
      "step": 9370
    },
    {
      "epoch": 0.6724978491540006,
      "grad_norm": 11.804948547619725,
      "learning_rate": 2e-05,
      "loss": 0.5087643146514893,
      "step": 9380
    },
    {
      "epoch": 0.6732147978204761,
      "grad_norm": 0.9423488878502342,
      "learning_rate": 2e-05,
      "loss": 0.49074244499206543,
      "step": 9390
    },
    {
      "epoch": 0.6739317464869515,
      "grad_norm": 0.6867857702598473,
      "learning_rate": 2e-05,
      "loss": 0.5236787796020508,
      "step": 9400
    },
    {
      "epoch": 0.6746486951534271,
      "grad_norm": 1.2096078756440447,
      "learning_rate": 2e-05,
      "loss": 0.5020617485046387,
      "step": 9410
    },
    {
      "epoch": 0.6753656438199025,
      "grad_norm": 0.8687443575980973,
      "learning_rate": 2e-05,
      "loss": 0.5068084716796875,
      "step": 9420
    },
    {
      "epoch": 0.6760825924863779,
      "grad_norm": 0.6909705144285024,
      "learning_rate": 2e-05,
      "loss": 0.4963220119476318,
      "step": 9430
    },
    {
      "epoch": 0.6767995411528535,
      "grad_norm": 0.812075700996762,
      "learning_rate": 2e-05,
      "loss": 0.5334953308105469,
      "step": 9440
    },
    {
      "epoch": 0.6775164898193289,
      "grad_norm": 0.9104081001088079,
      "learning_rate": 2e-05,
      "loss": 0.4929928779602051,
      "step": 9450
    },
    {
      "epoch": 0.6782334384858044,
      "grad_norm": 0.7619128781720905,
      "learning_rate": 2e-05,
      "loss": 0.5211712837219238,
      "step": 9460
    },
    {
      "epoch": 0.6789503871522798,
      "grad_norm": 0.8597479562946452,
      "learning_rate": 2e-05,
      "loss": 0.49042582511901855,
      "step": 9470
    },
    {
      "epoch": 0.6796673358187554,
      "grad_norm": 0.7400557178947533,
      "learning_rate": 2e-05,
      "loss": 0.49613513946533205,
      "step": 9480
    },
    {
      "epoch": 0.6803842844852308,
      "grad_norm": 0.7914742678073611,
      "learning_rate": 2e-05,
      "loss": 0.48298206329345705,
      "step": 9490
    },
    {
      "epoch": 0.6811012331517063,
      "grad_norm": 0.7570464909905467,
      "learning_rate": 2e-05,
      "loss": 0.49598379135131837,
      "step": 9500
    },
    {
      "epoch": 0.6818181818181818,
      "grad_norm": 0.9931138668843512,
      "learning_rate": 2e-05,
      "loss": 0.4933030128479004,
      "step": 9510
    },
    {
      "epoch": 0.6825351304846573,
      "grad_norm": 1.201395239994635,
      "learning_rate": 2e-05,
      "loss": 0.48967432975769043,
      "step": 9520
    },
    {
      "epoch": 0.6832520791511327,
      "grad_norm": 1.3970914605162736,
      "learning_rate": 2e-05,
      "loss": 0.4740629196166992,
      "step": 9530
    },
    {
      "epoch": 0.6839690278176083,
      "grad_norm": 1.178185378367238,
      "learning_rate": 2e-05,
      "loss": 0.5007808685302735,
      "step": 9540
    },
    {
      "epoch": 0.6846859764840837,
      "grad_norm": 0.8826379601772388,
      "learning_rate": 2e-05,
      "loss": 0.5071878910064698,
      "step": 9550
    },
    {
      "epoch": 0.6854029251505592,
      "grad_norm": 1.1509219865032543,
      "learning_rate": 2e-05,
      "loss": 0.4780189037322998,
      "step": 9560
    },
    {
      "epoch": 0.6861198738170347,
      "grad_norm": 0.969538926435925,
      "learning_rate": 2e-05,
      "loss": 0.5183696269989013,
      "step": 9570
    },
    {
      "epoch": 0.6868368224835102,
      "grad_norm": 0.8528965122419571,
      "learning_rate": 2e-05,
      "loss": 0.5109705924987793,
      "step": 9580
    },
    {
      "epoch": 0.6875537711499856,
      "grad_norm": 0.889401221272539,
      "learning_rate": 2e-05,
      "loss": 0.5026504516601562,
      "step": 9590
    },
    {
      "epoch": 0.6882707198164612,
      "grad_norm": 2.1197491654191127,
      "learning_rate": 2e-05,
      "loss": 0.5354850769042969,
      "step": 9600
    },
    {
      "epoch": 0.6889876684829366,
      "grad_norm": 2.0389473066545163,
      "learning_rate": 2e-05,
      "loss": 0.5193264007568359,
      "step": 9610
    },
    {
      "epoch": 0.6897046171494121,
      "grad_norm": 0.7087199148858011,
      "learning_rate": 2e-05,
      "loss": 0.5029548168182373,
      "step": 9620
    },
    {
      "epoch": 0.6904215658158875,
      "grad_norm": 1.2127995431863314,
      "learning_rate": 2e-05,
      "loss": 0.4868501663208008,
      "step": 9630
    },
    {
      "epoch": 0.6911385144823631,
      "grad_norm": 0.8747761743994781,
      "learning_rate": 2e-05,
      "loss": 0.49190487861633303,
      "step": 9640
    },
    {
      "epoch": 0.6918554631488385,
      "grad_norm": 0.756640821659776,
      "learning_rate": 2e-05,
      "loss": 0.5217726707458497,
      "step": 9650
    },
    {
      "epoch": 0.692572411815314,
      "grad_norm": 0.7004311754139898,
      "learning_rate": 2e-05,
      "loss": 0.4647576332092285,
      "step": 9660
    },
    {
      "epoch": 0.6932893604817895,
      "grad_norm": 0.8359671610360565,
      "learning_rate": 2e-05,
      "loss": 0.5082999229431152,
      "step": 9670
    },
    {
      "epoch": 0.694006309148265,
      "grad_norm": 0.7558403379293819,
      "learning_rate": 2e-05,
      "loss": 0.49845199584960936,
      "step": 9680
    },
    {
      "epoch": 0.6947232578147404,
      "grad_norm": 0.7304501351866326,
      "learning_rate": 2e-05,
      "loss": 0.4656517028808594,
      "step": 9690
    },
    {
      "epoch": 0.695440206481216,
      "grad_norm": 24.03149315088498,
      "learning_rate": 2e-05,
      "loss": 0.5014355659484864,
      "step": 9700
    },
    {
      "epoch": 0.6961571551476914,
      "grad_norm": 0.894979927757599,
      "learning_rate": 2e-05,
      "loss": 0.46665067672729493,
      "step": 9710
    },
    {
      "epoch": 0.6968741038141669,
      "grad_norm": 0.7242946417432715,
      "learning_rate": 2e-05,
      "loss": 0.5111838340759277,
      "step": 9720
    },
    {
      "epoch": 0.6975910524806423,
      "grad_norm": 0.6633311648377395,
      "learning_rate": 2e-05,
      "loss": 0.4743668079376221,
      "step": 9730
    },
    {
      "epoch": 0.6983080011471179,
      "grad_norm": 0.5833520807806973,
      "learning_rate": 2e-05,
      "loss": 0.4821101188659668,
      "step": 9740
    },
    {
      "epoch": 0.6990249498135933,
      "grad_norm": 0.6141845468059245,
      "learning_rate": 2e-05,
      "loss": 0.5141980171203613,
      "step": 9750
    },
    {
      "epoch": 0.6997418984800688,
      "grad_norm": 0.6825740833301499,
      "learning_rate": 2e-05,
      "loss": 0.5126357555389405,
      "step": 9760
    },
    {
      "epoch": 0.7004588471465443,
      "grad_norm": 0.6399014157764704,
      "learning_rate": 2e-05,
      "loss": 0.5052045822143555,
      "step": 9770
    },
    {
      "epoch": 0.7011757958130198,
      "grad_norm": 0.7814481672864668,
      "learning_rate": 2e-05,
      "loss": 0.5228930950164795,
      "step": 9780
    },
    {
      "epoch": 0.7018927444794952,
      "grad_norm": 0.6266218193344435,
      "learning_rate": 2e-05,
      "loss": 0.5060283660888671,
      "step": 9790
    },
    {
      "epoch": 0.7026096931459708,
      "grad_norm": 0.6934335474734691,
      "learning_rate": 2e-05,
      "loss": 0.48061318397521974,
      "step": 9800
    },
    {
      "epoch": 0.7033266418124462,
      "grad_norm": 0.7315483550459627,
      "learning_rate": 2e-05,
      "loss": 0.5156671524047851,
      "step": 9810
    },
    {
      "epoch": 0.7040435904789217,
      "grad_norm": 1.0805275448764364,
      "learning_rate": 2e-05,
      "loss": 0.5048374176025391,
      "step": 9820
    },
    {
      "epoch": 0.7047605391453972,
      "grad_norm": 0.7328614388172059,
      "learning_rate": 2e-05,
      "loss": 0.5022332668304443,
      "step": 9830
    },
    {
      "epoch": 0.7054774878118727,
      "grad_norm": 0.7039906500725398,
      "learning_rate": 2e-05,
      "loss": 0.4835071563720703,
      "step": 9840
    },
    {
      "epoch": 0.7061944364783481,
      "grad_norm": 0.6430569732054849,
      "learning_rate": 2e-05,
      "loss": 0.48531227111816405,
      "step": 9850
    },
    {
      "epoch": 0.7069113851448237,
      "grad_norm": 0.5651836085389784,
      "learning_rate": 2e-05,
      "loss": 0.5054697036743164,
      "step": 9860
    },
    {
      "epoch": 0.7076283338112991,
      "grad_norm": 0.6487720429634459,
      "learning_rate": 2e-05,
      "loss": 0.5119773864746093,
      "step": 9870
    },
    {
      "epoch": 0.7083452824777746,
      "grad_norm": 0.7313700471998886,
      "learning_rate": 2e-05,
      "loss": 0.48406591415405276,
      "step": 9880
    },
    {
      "epoch": 0.70906223114425,
      "grad_norm": 0.72210961662195,
      "learning_rate": 2e-05,
      "loss": 0.5028663635253906,
      "step": 9890
    },
    {
      "epoch": 0.7097791798107256,
      "grad_norm": 0.6874036994943528,
      "learning_rate": 2e-05,
      "loss": 0.48190922737121583,
      "step": 9900
    },
    {
      "epoch": 0.710496128477201,
      "grad_norm": 0.7869764815983818,
      "learning_rate": 2e-05,
      "loss": 0.49747462272644044,
      "step": 9910
    },
    {
      "epoch": 0.7112130771436765,
      "grad_norm": 0.764037395459954,
      "learning_rate": 2e-05,
      "loss": 0.48811841011047363,
      "step": 9920
    },
    {
      "epoch": 0.711930025810152,
      "grad_norm": 0.6868225505539057,
      "learning_rate": 2e-05,
      "loss": 0.4627346992492676,
      "step": 9930
    },
    {
      "epoch": 0.7126469744766275,
      "grad_norm": 0.7674574505534661,
      "learning_rate": 2e-05,
      "loss": 0.5016355991363526,
      "step": 9940
    },
    {
      "epoch": 0.7133639231431029,
      "grad_norm": 0.705362111274322,
      "learning_rate": 2e-05,
      "loss": 0.5030617713928223,
      "step": 9950
    },
    {
      "epoch": 0.7140808718095785,
      "grad_norm": 0.662919914065567,
      "learning_rate": 2e-05,
      "loss": 0.4625560760498047,
      "step": 9960
    },
    {
      "epoch": 0.7147978204760539,
      "grad_norm": 0.848578382497129,
      "learning_rate": 2e-05,
      "loss": 0.5092275619506836,
      "step": 9970
    },
    {
      "epoch": 0.7155147691425294,
      "grad_norm": 0.7293780477737621,
      "learning_rate": 2e-05,
      "loss": 0.4932692527770996,
      "step": 9980
    },
    {
      "epoch": 0.7162317178090049,
      "grad_norm": 0.8739080989995757,
      "learning_rate": 2e-05,
      "loss": 0.483767032623291,
      "step": 9990
    },
    {
      "epoch": 0.7169486664754804,
      "grad_norm": 0.7533238348660772,
      "learning_rate": 2e-05,
      "loss": 0.5219897270202637,
      "step": 10000
    },
    {
      "epoch": 0.7176656151419558,
      "grad_norm": 0.754240153626359,
      "learning_rate": 2e-05,
      "loss": 0.477223539352417,
      "step": 10010
    },
    {
      "epoch": 0.7183825638084314,
      "grad_norm": 0.7975962622409445,
      "learning_rate": 2e-05,
      "loss": 0.478654670715332,
      "step": 10020
    },
    {
      "epoch": 0.7190995124749068,
      "grad_norm": 0.8468379186132933,
      "learning_rate": 2e-05,
      "loss": 0.4989358425140381,
      "step": 10030
    },
    {
      "epoch": 0.7198164611413823,
      "grad_norm": 0.7140610416479808,
      "learning_rate": 2e-05,
      "loss": 0.49674625396728517,
      "step": 10040
    },
    {
      "epoch": 0.7205334098078577,
      "grad_norm": 0.7131911919481713,
      "learning_rate": 2e-05,
      "loss": 0.4896728992462158,
      "step": 10050
    },
    {
      "epoch": 0.7212503584743333,
      "grad_norm": 0.7910847002012612,
      "learning_rate": 2e-05,
      "loss": 0.4795466423034668,
      "step": 10060
    },
    {
      "epoch": 0.7219673071408087,
      "grad_norm": 1.307502652724149,
      "learning_rate": 2e-05,
      "loss": 0.4704431533813477,
      "step": 10070
    },
    {
      "epoch": 0.7226842558072842,
      "grad_norm": 0.6238628259720567,
      "learning_rate": 2e-05,
      "loss": 0.48863825798034666,
      "step": 10080
    },
    {
      "epoch": 0.7234012044737597,
      "grad_norm": 0.7379933591603842,
      "learning_rate": 2e-05,
      "loss": 0.5257764339447022,
      "step": 10090
    },
    {
      "epoch": 0.7241181531402352,
      "grad_norm": 0.7581822070095932,
      "learning_rate": 2e-05,
      "loss": 0.5008316993713379,
      "step": 10100
    },
    {
      "epoch": 0.7248351018067106,
      "grad_norm": 0.8321522784954003,
      "learning_rate": 2e-05,
      "loss": 0.48516130447387695,
      "step": 10110
    },
    {
      "epoch": 0.7255520504731862,
      "grad_norm": 0.7592589229785698,
      "learning_rate": 2e-05,
      "loss": 0.45975580215454104,
      "step": 10120
    },
    {
      "epoch": 0.7262689991396616,
      "grad_norm": 0.9725900994744937,
      "learning_rate": 2e-05,
      "loss": 0.48896303176879885,
      "step": 10130
    },
    {
      "epoch": 0.7269859478061371,
      "grad_norm": 0.8525630330475137,
      "learning_rate": 2e-05,
      "loss": 0.4968308925628662,
      "step": 10140
    },
    {
      "epoch": 0.7277028964726125,
      "grad_norm": 0.7582392287905331,
      "learning_rate": 2e-05,
      "loss": 0.48056783676147463,
      "step": 10150
    },
    {
      "epoch": 0.7284198451390881,
      "grad_norm": 0.8633040303642623,
      "learning_rate": 2e-05,
      "loss": 0.5002838134765625,
      "step": 10160
    },
    {
      "epoch": 0.7291367938055635,
      "grad_norm": 0.7205327591260188,
      "learning_rate": 2e-05,
      "loss": 0.5016871929168701,
      "step": 10170
    },
    {
      "epoch": 0.729853742472039,
      "grad_norm": 0.8151849144076816,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.4889083504676819,
      "step": 10180
    },
    {
      "epoch": 0.7305706911385145,
      "grad_norm": 1.0837232730398263,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.4761191844940186,
      "step": 10190
    },
    {
      "epoch": 0.73128763980499,
      "grad_norm": 0.6901268365040416,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.4868502616882324,
      "step": 10200
    },
    {
      "epoch": 0.7320045884714654,
      "grad_norm": 0.7264452878982376,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.4938946723937988,
      "step": 10210
    },
    {
      "epoch": 0.732721537137941,
      "grad_norm": 0.7420473537057809,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.4748692989349365,
      "step": 10220
    },
    {
      "epoch": 0.7334384858044164,
      "grad_norm": 0.7922888249826106,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.49942903518676757,
      "step": 10230
    },
    {
      "epoch": 0.7341554344708919,
      "grad_norm": 0.856964190672544,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.479322624206543,
      "step": 10240
    },
    {
      "epoch": 0.7348723831373674,
      "grad_norm": 0.6878200144716781,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.48067541122436525,
      "step": 10250
    },
    {
      "epoch": 0.7355893318038429,
      "grad_norm": 0.7381058379538875,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.5030919075012207,
      "step": 10260
    },
    {
      "epoch": 0.7363062804703183,
      "grad_norm": 0.6945571518297972,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.4917293071746826,
      "step": 10270
    },
    {
      "epoch": 0.7370232291367939,
      "grad_norm": 0.812097089226055,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.4911785125732422,
      "step": 10280
    },
    {
      "epoch": 0.7377401778032693,
      "grad_norm": 0.7685012057133969,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.496003532409668,
      "step": 10290
    },
    {
      "epoch": 0.7384571264697448,
      "grad_norm": 0.9814810178084871,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.46823463439941404,
      "step": 10300
    },
    {
      "epoch": 0.7391740751362202,
      "grad_norm": 0.7576311818470874,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.4909675598144531,
      "step": 10310
    },
    {
      "epoch": 0.7398910238026958,
      "grad_norm": 0.6850619350320722,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.46691274642944336,
      "step": 10320
    },
    {
      "epoch": 0.7406079724691712,
      "grad_norm": 1.019024914462093,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.46620969772338866,
      "step": 10330
    },
    {
      "epoch": 0.7413249211356467,
      "grad_norm": 0.702183137610601,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.5248452663421631,
      "step": 10340
    },
    {
      "epoch": 0.7420418698021222,
      "grad_norm": 0.6475872839029189,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.4619590759277344,
      "step": 10350
    },
    {
      "epoch": 0.7427588184685977,
      "grad_norm": 3.191447230312936,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.4940286636352539,
      "step": 10360
    },
    {
      "epoch": 0.7434757671350731,
      "grad_norm": 0.9511854630285578,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.45314865112304686,
      "step": 10370
    },
    {
      "epoch": 0.7441927158015486,
      "grad_norm": 0.7109780129392402,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.5028180599212646,
      "step": 10380
    },
    {
      "epoch": 0.7449096644680241,
      "grad_norm": 0.6591676412530446,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.4685412883758545,
      "step": 10390
    },
    {
      "epoch": 0.7456266131344995,
      "grad_norm": 0.759783553033096,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.49547834396362306,
      "step": 10400
    },
    {
      "epoch": 0.746343561800975,
      "grad_norm": 0.7018786240753214,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.494583797454834,
      "step": 10410
    },
    {
      "epoch": 0.7470605104674505,
      "grad_norm": 0.723065614660499,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.5040961742401123,
      "step": 10420
    },
    {
      "epoch": 0.747777459133926,
      "grad_norm": 0.7032870535394683,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.4740753173828125,
      "step": 10430
    },
    {
      "epoch": 0.7484944078004014,
      "grad_norm": 1.2841607230335341,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.512285852432251,
      "step": 10440
    },
    {
      "epoch": 0.749211356466877,
      "grad_norm": 0.7807845011508054,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.48694562911987305,
      "step": 10450
    },
    {
      "epoch": 0.7499283051333524,
      "grad_norm": 0.7606810487163044,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.49996390342712405,
      "step": 10460
    },
    {
      "epoch": 0.7506452537998279,
      "grad_norm": 0.7622686122478394,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.4776145458221436,
      "step": 10470
    },
    {
      "epoch": 0.7513622024663034,
      "grad_norm": 0.6954491953290093,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.4932582855224609,
      "step": 10480
    },
    {
      "epoch": 0.7520791511327789,
      "grad_norm": 0.7399540246449979,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.47369837760925293,
      "step": 10490
    },
    {
      "epoch": 0.7527960997992543,
      "grad_norm": 0.6770137998458474,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.47462615966796873,
      "step": 10500
    },
    {
      "epoch": 0.7535130484657299,
      "grad_norm": 0.9255980613928344,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.5019224166870118,
      "step": 10510
    },
    {
      "epoch": 0.7542299971322053,
      "grad_norm": 0.7531431059575958,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.5205343246459961,
      "step": 10520
    },
    {
      "epoch": 0.7549469457986808,
      "grad_norm": 0.586182891778256,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.49133005142211916,
      "step": 10530
    },
    {
      "epoch": 0.7556638944651562,
      "grad_norm": 3.291701767327292,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.47762632369995117,
      "step": 10540
    },
    {
      "epoch": 0.7563808431316318,
      "grad_norm": 0.8673428267723197,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.4955639362335205,
      "step": 10550
    },
    {
      "epoch": 0.7570977917981072,
      "grad_norm": 0.9500255268383605,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.4923381328582764,
      "step": 10560
    },
    {
      "epoch": 0.7578147404645827,
      "grad_norm": 0.6269004899662114,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.4933166980743408,
      "step": 10570
    },
    {
      "epoch": 0.7585316891310582,
      "grad_norm": 0.7722263798128499,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.4761941432952881,
      "step": 10580
    },
    {
      "epoch": 0.7592486377975337,
      "grad_norm": 0.7463004296885002,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.5023458957672119,
      "step": 10590
    },
    {
      "epoch": 0.7599655864640091,
      "grad_norm": 0.7999327009779091,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.46838808059692383,
      "step": 10600
    },
    {
      "epoch": 0.7606825351304847,
      "grad_norm": 0.6060343851642188,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.5032796859741211,
      "step": 10610
    },
    {
      "epoch": 0.7613994837969601,
      "grad_norm": 1.0011030017673936,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.5114131927490234,
      "step": 10620
    },
    {
      "epoch": 0.7621164324634356,
      "grad_norm": 0.7831809124891721,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.5045238494873047,
      "step": 10630
    },
    {
      "epoch": 0.7628333811299111,
      "grad_norm": 0.7356844956982318,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.45974035263061525,
      "step": 10640
    },
    {
      "epoch": 0.7635503297963866,
      "grad_norm": 0.607749082329079,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.4569259166717529,
      "step": 10650
    },
    {
      "epoch": 0.764267278462862,
      "grad_norm": 0.7306609960137558,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.47473344802856443,
      "step": 10660
    },
    {
      "epoch": 0.7649842271293376,
      "grad_norm": 0.9751693553208026,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.48780837059021,
      "step": 10670
    },
    {
      "epoch": 0.765701175795813,
      "grad_norm": 0.6359030320483138,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.4973846435546875,
      "step": 10680
    },
    {
      "epoch": 0.7664181244622885,
      "grad_norm": 0.6674059844620233,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.478622579574585,
      "step": 10690
    },
    {
      "epoch": 0.767135073128764,
      "grad_norm": 0.7942008277904752,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.48732309341430663,
      "step": 10700
    },
    {
      "epoch": 0.7678520217952395,
      "grad_norm": 0.6863212344067879,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.47802019119262695,
      "step": 10710
    },
    {
      "epoch": 0.7685689704617149,
      "grad_norm": 0.8434228161306174,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.47385478019714355,
      "step": 10720
    },
    {
      "epoch": 0.7692859191281904,
      "grad_norm": 0.695711319872533,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.4827431678771973,
      "step": 10730
    },
    {
      "epoch": 0.7700028677946659,
      "grad_norm": 0.6791105047324587,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.5031193256378174,
      "step": 10740
    },
    {
      "epoch": 0.7707198164611414,
      "grad_norm": 0.595541556095443,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.46905031204223635,
      "step": 10750
    },
    {
      "epoch": 0.7714367651276168,
      "grad_norm": 0.7833695973830083,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.4958200931549072,
      "step": 10760
    },
    {
      "epoch": 0.7721537137940924,
      "grad_norm": 0.667610434666589,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.4627577781677246,
      "step": 10770
    },
    {
      "epoch": 0.7728706624605678,
      "grad_norm": 0.5596295619195768,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.4542607307434082,
      "step": 10780
    },
    {
      "epoch": 0.7735876111270433,
      "grad_norm": 0.7038772273803294,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.49508209228515626,
      "step": 10790
    },
    {
      "epoch": 0.7743045597935188,
      "grad_norm": 0.703737271062325,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.4785890579223633,
      "step": 10800
    },
    {
      "epoch": 0.7750215084599943,
      "grad_norm": 0.608097935296564,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.4562793254852295,
      "step": 10810
    },
    {
      "epoch": 0.7757384571264697,
      "grad_norm": 0.6006005378987244,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.4895011425018311,
      "step": 10820
    },
    {
      "epoch": 0.7764554057929453,
      "grad_norm": 0.6937883185247019,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.47237362861633303,
      "step": 10830
    },
    {
      "epoch": 0.7771723544594207,
      "grad_norm": 0.6210774722441138,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.4567268848419189,
      "step": 10840
    },
    {
      "epoch": 0.7778893031258962,
      "grad_norm": 0.6810156767623091,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.48558897972106935,
      "step": 10850
    },
    {
      "epoch": 0.7786062517923716,
      "grad_norm": 0.7904654452609708,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.470566463470459,
      "step": 10860
    },
    {
      "epoch": 0.7793232004588472,
      "grad_norm": 0.5768254142581318,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.4643099308013916,
      "step": 10870
    },
    {
      "epoch": 0.7800401491253226,
      "grad_norm": 0.740630070864531,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.46635141372680666,
      "step": 10880
    },
    {
      "epoch": 0.7807570977917981,
      "grad_norm": 0.6551406331212967,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.46950159072875974,
      "step": 10890
    },
    {
      "epoch": 0.7814740464582736,
      "grad_norm": 0.6983054597558608,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.4567814826965332,
      "step": 10900
    },
    {
      "epoch": 0.7821909951247491,
      "grad_norm": 0.7203113622997114,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.469309663772583,
      "step": 10910
    },
    {
      "epoch": 0.7829079437912245,
      "grad_norm": 0.8038608023498618,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.4774013519287109,
      "step": 10920
    },
    {
      "epoch": 0.7836248924577001,
      "grad_norm": 1.0705108484442925,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.48203001022338865,
      "step": 10930
    },
    {
      "epoch": 0.7843418411241755,
      "grad_norm": 0.6058443623080825,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.4646466732025146,
      "step": 10940
    },
    {
      "epoch": 0.785058789790651,
      "grad_norm": 1.245137347305341,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.4687192440032959,
      "step": 10950
    },
    {
      "epoch": 0.7857757384571264,
      "grad_norm": 0.7514941061984217,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.4809783935546875,
      "step": 10960
    },
    {
      "epoch": 0.786492687123602,
      "grad_norm": 0.6306686614625188,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.48053684234619143,
      "step": 10970
    },
    {
      "epoch": 0.7872096357900774,
      "grad_norm": 0.7492894311597125,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.4670445442199707,
      "step": 10980
    },
    {
      "epoch": 0.787926584456553,
      "grad_norm": 1.0821723158392906,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.46114187240600585,
      "step": 10990
    },
    {
      "epoch": 0.7886435331230284,
      "grad_norm": 0.7002484186740843,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.47629270553588865,
      "step": 11000
    },
    {
      "epoch": 0.7893604817895039,
      "grad_norm": 0.695259201748693,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.4607882499694824,
      "step": 11010
    },
    {
      "epoch": 0.7900774304559793,
      "grad_norm": 0.7015011399590464,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.4691750526428223,
      "step": 11020
    },
    {
      "epoch": 0.7907943791224549,
      "grad_norm": 0.9965548676790333,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.4423958778381348,
      "step": 11030
    },
    {
      "epoch": 0.7915113277889303,
      "grad_norm": 0.7727396317395708,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.4394947052001953,
      "step": 11040
    },
    {
      "epoch": 0.7922282764554058,
      "grad_norm": 0.8484967123843982,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.4559922218322754,
      "step": 11050
    },
    {
      "epoch": 0.7929452251218813,
      "grad_norm": 0.6237540923787039,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.4563802719116211,
      "step": 11060
    },
    {
      "epoch": 0.7936621737883568,
      "grad_norm": 0.6742124257148528,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.46270270347595216,
      "step": 11070
    },
    {
      "epoch": 0.7943791224548322,
      "grad_norm": 0.6770308270099322,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.4586421012878418,
      "step": 11080
    },
    {
      "epoch": 0.7950960711213078,
      "grad_norm": 0.6697933892846906,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.44456186294555666,
      "step": 11090
    },
    {
      "epoch": 0.7958130197877832,
      "grad_norm": 0.7085954136958633,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.4758300304412842,
      "step": 11100
    },
    {
      "epoch": 0.7965299684542587,
      "grad_norm": 0.6328619433883974,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.4657938003540039,
      "step": 11110
    },
    {
      "epoch": 0.7972469171207341,
      "grad_norm": 0.5828635917959476,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.46340627670288087,
      "step": 11120
    },
    {
      "epoch": 0.7979638657872097,
      "grad_norm": 0.6417990123561402,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.4868314266204834,
      "step": 11130
    },
    {
      "epoch": 0.7986808144536851,
      "grad_norm": 0.7683151688133414,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.44542875289916994,
      "step": 11140
    },
    {
      "epoch": 0.7993977631201606,
      "grad_norm": 0.6591165707987792,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.4233969211578369,
      "step": 11150
    },
    {
      "epoch": 0.8001147117866361,
      "grad_norm": 0.6648151051374402,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.47496833801269533,
      "step": 11160
    },
    {
      "epoch": 0.8008316604531116,
      "grad_norm": 0.6255108441270076,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.46144442558288573,
      "step": 11170
    },
    {
      "epoch": 0.801548609119587,
      "grad_norm": 0.6380747420603533,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.473084831237793,
      "step": 11180
    },
    {
      "epoch": 0.8022655577860626,
      "grad_norm": 0.7566211278372678,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.45789737701416017,
      "step": 11190
    },
    {
      "epoch": 0.8024806423860051,
      "step": 11193,
      "total_flos": 5300110976090112.0,
      "train_loss": 0.043386135878367385,
      "train_runtime": 14858.9991,
      "train_samples_per_second": 9.039,
      "train_steps_per_second": 0.753
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5300110976090112.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}