{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3174,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000945179584120983,
      "grad_norm": 1.1085860993315724,
      "learning_rate": 0.0,
      "loss": 1.1217434406280518,
      "step": 1
    },
    {
      "epoch": 0.001890359168241966,
      "grad_norm": 0.8898090002864506,
      "learning_rate": 6.28930817610063e-07,
      "loss": 0.9310062527656555,
      "step": 2
    },
    {
      "epoch": 0.002835538752362949,
      "grad_norm": 0.9810376593389925,
      "learning_rate": 1.257861635220126e-06,
      "loss": 0.914983868598938,
      "step": 3
    },
    {
      "epoch": 0.003780718336483932,
      "grad_norm": 0.8973727790860981,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 0.9063125848770142,
      "step": 4
    },
    {
      "epoch": 0.004725897920604915,
      "grad_norm": 1.0140021717059031,
      "learning_rate": 2.515723270440252e-06,
      "loss": 0.9831501245498657,
      "step": 5
    },
    {
      "epoch": 0.005671077504725898,
      "grad_norm": 0.9067768102710736,
      "learning_rate": 3.1446540880503146e-06,
      "loss": 0.9031134843826294,
      "step": 6
    },
    {
      "epoch": 0.006616257088846881,
      "grad_norm": 0.8887155643237732,
      "learning_rate": 3.7735849056603773e-06,
      "loss": 0.8842957019805908,
      "step": 7
    },
    {
      "epoch": 0.007561436672967864,
      "grad_norm": 0.9118428176416897,
      "learning_rate": 4.40251572327044e-06,
      "loss": 0.9458888173103333,
      "step": 8
    },
    {
      "epoch": 0.008506616257088847,
      "grad_norm": 0.7966050649571385,
      "learning_rate": 5.031446540880504e-06,
      "loss": 0.8998948335647583,
      "step": 9
    },
    {
      "epoch": 0.00945179584120983,
      "grad_norm": 0.9176326445166538,
      "learning_rate": 5.660377358490566e-06,
      "loss": 0.9767179489135742,
      "step": 10
    },
    {
      "epoch": 0.010396975425330813,
      "grad_norm": 0.8910825080664788,
      "learning_rate": 6.289308176100629e-06,
      "loss": 1.0211412906646729,
      "step": 11
    },
    {
      "epoch": 0.011342155009451797,
      "grad_norm": 0.8481001418083726,
      "learning_rate": 6.918238993710692e-06,
      "loss": 0.9189810752868652,
      "step": 12
    },
    {
      "epoch": 0.012287334593572778,
      "grad_norm": 0.8153555461353063,
      "learning_rate": 7.547169811320755e-06,
      "loss": 0.9855996966362,
      "step": 13
    },
    {
      "epoch": 0.013232514177693762,
      "grad_norm": 0.8073720805381847,
      "learning_rate": 8.176100628930818e-06,
      "loss": 0.9197836518287659,
      "step": 14
    },
    {
      "epoch": 0.014177693761814745,
      "grad_norm": 0.7079132356604668,
      "learning_rate": 8.80503144654088e-06,
      "loss": 0.844753086566925,
      "step": 15
    },
    {
      "epoch": 0.015122873345935728,
      "grad_norm": 0.6022799077765889,
      "learning_rate": 9.433962264150944e-06,
      "loss": 0.7627860307693481,
      "step": 16
    },
    {
      "epoch": 0.01606805293005671,
      "grad_norm": 0.6265436736815001,
      "learning_rate": 1.0062893081761008e-05,
      "loss": 0.7974623441696167,
      "step": 17
    },
    {
      "epoch": 0.017013232514177693,
      "grad_norm": 0.6114704777425828,
      "learning_rate": 1.069182389937107e-05,
      "loss": 0.8882802724838257,
      "step": 18
    },
    {
      "epoch": 0.017958412098298678,
      "grad_norm": 0.5982358124346371,
      "learning_rate": 1.1320754716981132e-05,
      "loss": 0.8868510723114014,
      "step": 19
    },
    {
      "epoch": 0.01890359168241966,
      "grad_norm": 0.5596347465468406,
      "learning_rate": 1.1949685534591196e-05,
      "loss": 0.7873749732971191,
      "step": 20
    },
    {
      "epoch": 0.01984877126654064,
      "grad_norm": 0.5234144332787791,
      "learning_rate": 1.2578616352201259e-05,
      "loss": 0.8859477043151855,
      "step": 21
    },
    {
      "epoch": 0.020793950850661626,
      "grad_norm": 0.5239931957243392,
      "learning_rate": 1.320754716981132e-05,
      "loss": 0.7527213096618652,
      "step": 22
    },
    {
      "epoch": 0.021739130434782608,
      "grad_norm": 0.6535541435332252,
      "learning_rate": 1.3836477987421385e-05,
      "loss": 0.8144068717956543,
      "step": 23
    },
    {
      "epoch": 0.022684310018903593,
      "grad_norm": 0.5300247473555324,
      "learning_rate": 1.4465408805031447e-05,
      "loss": 0.7220904231071472,
      "step": 24
    },
    {
      "epoch": 0.023629489603024575,
      "grad_norm": 0.572897026978421,
      "learning_rate": 1.509433962264151e-05,
      "loss": 0.8799376487731934,
      "step": 25
    },
    {
      "epoch": 0.024574669187145556,
      "grad_norm": 0.5803614859410433,
      "learning_rate": 1.572327044025157e-05,
      "loss": 0.8639330863952637,
      "step": 26
    },
    {
      "epoch": 0.02551984877126654,
      "grad_norm": 0.6412303566460493,
      "learning_rate": 1.6352201257861635e-05,
      "loss": 0.6895577907562256,
      "step": 27
    },
    {
      "epoch": 0.026465028355387523,
      "grad_norm": 0.6167467779000204,
      "learning_rate": 1.69811320754717e-05,
      "loss": 0.7397151589393616,
      "step": 28
    },
    {
      "epoch": 0.027410207939508508,
      "grad_norm": 0.5666166918845975,
      "learning_rate": 1.761006289308176e-05,
      "loss": 0.6955139636993408,
      "step": 29
    },
    {
      "epoch": 0.02835538752362949,
      "grad_norm": 0.5055249174176583,
      "learning_rate": 1.8238993710691824e-05,
      "loss": 0.8122433423995972,
      "step": 30
    },
    {
      "epoch": 0.02930056710775047,
      "grad_norm": 0.5205123630745588,
      "learning_rate": 1.8867924528301888e-05,
      "loss": 0.6926913261413574,
      "step": 31
    },
    {
      "epoch": 0.030245746691871456,
      "grad_norm": 0.5981026730778634,
      "learning_rate": 1.9496855345911952e-05,
      "loss": 0.7124322652816772,
      "step": 32
    },
    {
      "epoch": 0.031190926275992438,
      "grad_norm": 0.6094414779566503,
      "learning_rate": 2.0125786163522016e-05,
      "loss": 0.8438947200775146,
      "step": 33
    },
    {
      "epoch": 0.03213610586011342,
      "grad_norm": 0.5715918749766736,
      "learning_rate": 2.0754716981132076e-05,
      "loss": 0.747673749923706,
      "step": 34
    },
    {
      "epoch": 0.0330812854442344,
      "grad_norm": 0.563791683823168,
      "learning_rate": 2.138364779874214e-05,
      "loss": 0.7367045879364014,
      "step": 35
    },
    {
      "epoch": 0.034026465028355386,
      "grad_norm": 0.49751664084795705,
      "learning_rate": 2.2012578616352204e-05,
      "loss": 0.7274491190910339,
      "step": 36
    },
    {
      "epoch": 0.03497164461247637,
      "grad_norm": 0.4808792194223229,
      "learning_rate": 2.2641509433962265e-05,
      "loss": 0.6777164936065674,
      "step": 37
    },
    {
      "epoch": 0.035916824196597356,
      "grad_norm": 0.5006580018724986,
      "learning_rate": 2.327044025157233e-05,
      "loss": 0.6841402649879456,
      "step": 38
    },
    {
      "epoch": 0.036862003780718335,
      "grad_norm": 0.4918656440139497,
      "learning_rate": 2.3899371069182393e-05,
      "loss": 0.7405531406402588,
      "step": 39
    },
    {
      "epoch": 0.03780718336483932,
      "grad_norm": 0.5361183242572455,
      "learning_rate": 2.4528301886792453e-05,
      "loss": 0.6283311247825623,
      "step": 40
    },
    {
      "epoch": 0.038752362948960305,
      "grad_norm": 0.5068420344350174,
      "learning_rate": 2.5157232704402517e-05,
      "loss": 0.7284811735153198,
      "step": 41
    },
    {
      "epoch": 0.03969754253308128,
      "grad_norm": 0.4798717723861369,
      "learning_rate": 2.578616352201258e-05,
      "loss": 0.7012759447097778,
      "step": 42
    },
    {
      "epoch": 0.04064272211720227,
      "grad_norm": 0.48507514164377624,
      "learning_rate": 2.641509433962264e-05,
      "loss": 0.626822829246521,
      "step": 43
    },
    {
      "epoch": 0.04158790170132325,
      "grad_norm": 0.5310833893680008,
      "learning_rate": 2.7044025157232706e-05,
      "loss": 0.7137644290924072,
      "step": 44
    },
    {
      "epoch": 0.04253308128544423,
      "grad_norm": 0.4644576385015965,
      "learning_rate": 2.767295597484277e-05,
      "loss": 0.6593171954154968,
      "step": 45
    },
    {
      "epoch": 0.043478260869565216,
      "grad_norm": 0.46578959509239554,
      "learning_rate": 2.830188679245283e-05,
      "loss": 0.6852736473083496,
      "step": 46
    },
    {
      "epoch": 0.0444234404536862,
      "grad_norm": 0.5534138345731002,
      "learning_rate": 2.8930817610062894e-05,
      "loss": 0.7456097602844238,
      "step": 47
    },
    {
      "epoch": 0.045368620037807186,
      "grad_norm": 0.5306253485227328,
      "learning_rate": 2.9559748427672958e-05,
      "loss": 0.6997179985046387,
      "step": 48
    },
    {
      "epoch": 0.046313799621928164,
      "grad_norm": 0.5012686727310159,
      "learning_rate": 3.018867924528302e-05,
      "loss": 0.806352436542511,
      "step": 49
    },
    {
      "epoch": 0.04725897920604915,
      "grad_norm": 0.448903798997779,
      "learning_rate": 3.081761006289308e-05,
      "loss": 0.5763542652130127,
      "step": 50
    },
    {
      "epoch": 0.048204158790170135,
      "grad_norm": 0.47207946115263494,
      "learning_rate": 3.144654088050314e-05,
      "loss": 0.7140188217163086,
      "step": 51
    },
    {
      "epoch": 0.04914933837429111,
      "grad_norm": 0.4582296725565494,
      "learning_rate": 3.207547169811321e-05,
      "loss": 0.7094287872314453,
      "step": 52
    },
    {
      "epoch": 0.0500945179584121,
      "grad_norm": 0.49855031860601406,
      "learning_rate": 3.270440251572327e-05,
      "loss": 0.6836258769035339,
      "step": 53
    },
    {
      "epoch": 0.05103969754253308,
      "grad_norm": 0.5815134229422777,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.602997899055481,
      "step": 54
    },
    {
      "epoch": 0.05198487712665406,
      "grad_norm": 0.4718952370080357,
      "learning_rate": 3.39622641509434e-05,
      "loss": 0.6382819414138794,
      "step": 55
    },
    {
      "epoch": 0.052930056710775046,
      "grad_norm": 0.4851533525294259,
      "learning_rate": 3.4591194968553456e-05,
      "loss": 0.5560786724090576,
      "step": 56
    },
    {
      "epoch": 0.05387523629489603,
      "grad_norm": 0.5029994194920949,
      "learning_rate": 3.522012578616352e-05,
      "loss": 0.6942906975746155,
      "step": 57
    },
    {
      "epoch": 0.054820415879017016,
      "grad_norm": 0.49296505683019687,
      "learning_rate": 3.5849056603773584e-05,
      "loss": 0.6818789839744568,
      "step": 58
    },
    {
      "epoch": 0.055765595463137994,
      "grad_norm": 0.46781820516662803,
      "learning_rate": 3.647798742138365e-05,
      "loss": 0.62168288230896,
      "step": 59
    },
    {
      "epoch": 0.05671077504725898,
      "grad_norm": 0.633798797268738,
      "learning_rate": 3.710691823899371e-05,
      "loss": 0.713526725769043,
      "step": 60
    },
    {
      "epoch": 0.057655954631379965,
      "grad_norm": 0.5371369409268637,
      "learning_rate": 3.7735849056603776e-05,
      "loss": 0.709291398525238,
      "step": 61
    },
    {
      "epoch": 0.05860113421550094,
      "grad_norm": 0.48428051828149854,
      "learning_rate": 3.836477987421384e-05,
      "loss": 0.7215760350227356,
      "step": 62
    },
    {
      "epoch": 0.05954631379962193,
      "grad_norm": 0.5129719691907396,
      "learning_rate": 3.8993710691823904e-05,
      "loss": 0.6647354364395142,
      "step": 63
    },
    {
      "epoch": 0.06049149338374291,
      "grad_norm": 0.4975652192563672,
      "learning_rate": 3.962264150943397e-05,
      "loss": 0.7110547423362732,
      "step": 64
    },
    {
      "epoch": 0.06143667296786389,
      "grad_norm": 0.4399677906582835,
      "learning_rate": 4.025157232704403e-05,
      "loss": 0.6953577399253845,
      "step": 65
    },
    {
      "epoch": 0.062381852551984876,
      "grad_norm": 0.4831889107700821,
      "learning_rate": 4.088050314465409e-05,
      "loss": 0.736778974533081,
      "step": 66
    },
    {
      "epoch": 0.06332703213610585,
      "grad_norm": 0.5033063292017789,
      "learning_rate": 4.150943396226415e-05,
      "loss": 0.6464102268218994,
      "step": 67
    },
    {
      "epoch": 0.06427221172022685,
      "grad_norm": 0.5166789468408276,
      "learning_rate": 4.213836477987422e-05,
      "loss": 0.7761030197143555,
      "step": 68
    },
    {
      "epoch": 0.06521739130434782,
      "grad_norm": 0.5004027843681723,
      "learning_rate": 4.276729559748428e-05,
      "loss": 0.6757903099060059,
      "step": 69
    },
    {
      "epoch": 0.0661625708884688,
      "grad_norm": 0.5493233137786742,
      "learning_rate": 4.3396226415094345e-05,
      "loss": 0.6686078310012817,
      "step": 70
    },
    {
      "epoch": 0.0671077504725898,
      "grad_norm": 0.47583699186762074,
      "learning_rate": 4.402515723270441e-05,
      "loss": 0.5691801309585571,
      "step": 71
    },
    {
      "epoch": 0.06805293005671077,
      "grad_norm": 0.5512310047763085,
      "learning_rate": 4.4654088050314466e-05,
      "loss": 0.76002037525177,
      "step": 72
    },
    {
      "epoch": 0.06899810964083176,
      "grad_norm": 0.4760648509790648,
      "learning_rate": 4.528301886792453e-05,
      "loss": 0.5306321382522583,
      "step": 73
    },
    {
      "epoch": 0.06994328922495274,
      "grad_norm": 0.5137801429672079,
      "learning_rate": 4.5911949685534594e-05,
      "loss": 0.6710862517356873,
      "step": 74
    },
    {
      "epoch": 0.07088846880907372,
      "grad_norm": 0.5129439779224914,
      "learning_rate": 4.654088050314466e-05,
      "loss": 0.6052921414375305,
      "step": 75
    },
    {
      "epoch": 0.07183364839319471,
      "grad_norm": 0.5278903467995558,
      "learning_rate": 4.716981132075472e-05,
      "loss": 0.7229230403900146,
      "step": 76
    },
    {
      "epoch": 0.07277882797731569,
      "grad_norm": 0.5083869792041931,
      "learning_rate": 4.7798742138364785e-05,
      "loss": 0.6530970931053162,
      "step": 77
    },
    {
      "epoch": 0.07372400756143667,
      "grad_norm": 0.543138994150268,
      "learning_rate": 4.842767295597484e-05,
      "loss": 0.6834779381752014,
      "step": 78
    },
    {
      "epoch": 0.07466918714555766,
      "grad_norm": 0.5514570277381782,
      "learning_rate": 4.9056603773584906e-05,
      "loss": 0.599922776222229,
      "step": 79
    },
    {
      "epoch": 0.07561436672967864,
      "grad_norm": 0.5464732513762638,
      "learning_rate": 4.968553459119497e-05,
      "loss": 0.7290014028549194,
      "step": 80
    },
    {
      "epoch": 0.07655954631379962,
      "grad_norm": 0.5222538537560315,
      "learning_rate": 5.0314465408805034e-05,
      "loss": 0.5838358998298645,
      "step": 81
    },
    {
      "epoch": 0.07750472589792061,
      "grad_norm": 0.4241713902983496,
      "learning_rate": 5.09433962264151e-05,
      "loss": 0.5849518775939941,
      "step": 82
    },
    {
      "epoch": 0.07844990548204159,
      "grad_norm": 0.4642955654089573,
      "learning_rate": 5.157232704402516e-05,
      "loss": 0.6268076300621033,
      "step": 83
    },
    {
      "epoch": 0.07939508506616257,
      "grad_norm": 0.4905336814209204,
      "learning_rate": 5.220125786163522e-05,
      "loss": 0.7151265740394592,
      "step": 84
    },
    {
      "epoch": 0.08034026465028356,
      "grad_norm": 0.4720000377017285,
      "learning_rate": 5.283018867924528e-05,
      "loss": 0.5943088531494141,
      "step": 85
    },
    {
      "epoch": 0.08128544423440454,
      "grad_norm": 0.48031453929898066,
      "learning_rate": 5.345911949685535e-05,
      "loss": 0.6591486930847168,
      "step": 86
    },
    {
      "epoch": 0.08223062381852551,
      "grad_norm": 0.4467614164073167,
      "learning_rate": 5.408805031446541e-05,
      "loss": 0.6576695442199707,
      "step": 87
    },
    {
      "epoch": 0.0831758034026465,
      "grad_norm": 0.5489031946340465,
      "learning_rate": 5.4716981132075475e-05,
      "loss": 0.7433443069458008,
      "step": 88
    },
    {
      "epoch": 0.08412098298676748,
      "grad_norm": 0.5434632819558968,
      "learning_rate": 5.534591194968554e-05,
      "loss": 0.7114048600196838,
      "step": 89
    },
    {
      "epoch": 0.08506616257088846,
      "grad_norm": 0.46726664635183596,
      "learning_rate": 5.5974842767295596e-05,
      "loss": 0.48968639969825745,
      "step": 90
    },
    {
      "epoch": 0.08601134215500945,
      "grad_norm": 0.456378210734111,
      "learning_rate": 5.660377358490566e-05,
      "loss": 0.5947794318199158,
      "step": 91
    },
    {
      "epoch": 0.08695652173913043,
      "grad_norm": 0.40567188703359214,
      "learning_rate": 5.7232704402515724e-05,
      "loss": 0.5247464179992676,
      "step": 92
    },
    {
      "epoch": 0.08790170132325142,
      "grad_norm": 0.5027029077173495,
      "learning_rate": 5.786163522012579e-05,
      "loss": 0.6023780703544617,
      "step": 93
    },
    {
      "epoch": 0.0888468809073724,
      "grad_norm": 0.49533190228458285,
      "learning_rate": 5.849056603773585e-05,
      "loss": 0.609745979309082,
      "step": 94
    },
    {
      "epoch": 0.08979206049149338,
      "grad_norm": 0.4930604277484113,
      "learning_rate": 5.9119496855345916e-05,
      "loss": 0.765373945236206,
      "step": 95
    },
    {
      "epoch": 0.09073724007561437,
      "grad_norm": 0.44523769090920384,
      "learning_rate": 5.974842767295597e-05,
      "loss": 0.706850528717041,
      "step": 96
    },
    {
      "epoch": 0.09168241965973535,
      "grad_norm": 0.4323465815555924,
      "learning_rate": 6.037735849056604e-05,
      "loss": 0.7430319786071777,
      "step": 97
    },
    {
      "epoch": 0.09262759924385633,
      "grad_norm": 0.4697948725319461,
      "learning_rate": 6.10062893081761e-05,
      "loss": 0.7576801180839539,
      "step": 98
    },
    {
      "epoch": 0.09357277882797732,
      "grad_norm": 0.46706222314975315,
      "learning_rate": 6.163522012578616e-05,
      "loss": 0.622502326965332,
      "step": 99
    },
    {
      "epoch": 0.0945179584120983,
      "grad_norm": 0.4425221091432325,
      "learning_rate": 6.226415094339622e-05,
      "loss": 0.5928550958633423,
      "step": 100
    },
    {
      "epoch": 0.09546313799621928,
      "grad_norm": 0.43509539753648596,
      "learning_rate": 6.289308176100629e-05,
      "loss": 0.6601524353027344,
      "step": 101
    },
    {
      "epoch": 0.09640831758034027,
      "grad_norm": 0.5328785273946602,
      "learning_rate": 6.352201257861635e-05,
      "loss": 0.66511070728302,
      "step": 102
    },
    {
      "epoch": 0.09735349716446125,
      "grad_norm": 0.45775850144010954,
      "learning_rate": 6.415094339622641e-05,
      "loss": 0.5305960178375244,
      "step": 103
    },
    {
      "epoch": 0.09829867674858223,
      "grad_norm": 0.40135695817895745,
      "learning_rate": 6.477987421383648e-05,
      "loss": 0.6249047517776489,
      "step": 104
    },
    {
      "epoch": 0.09924385633270322,
      "grad_norm": 0.4813386062315147,
      "learning_rate": 6.540880503144654e-05,
      "loss": 0.8066498637199402,
      "step": 105
    },
    {
      "epoch": 0.1001890359168242,
      "grad_norm": 0.5265093315662366,
      "learning_rate": 6.60377358490566e-05,
      "loss": 0.6620453596115112,
      "step": 106
    },
    {
      "epoch": 0.10113421550094517,
      "grad_norm": 0.4523649208426104,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.6570029258728027,
      "step": 107
    },
    {
      "epoch": 0.10207939508506617,
      "grad_norm": 0.45954113210488406,
      "learning_rate": 6.729559748427673e-05,
      "loss": 0.6286371350288391,
      "step": 108
    },
    {
      "epoch": 0.10302457466918714,
      "grad_norm": 0.4577272643611639,
      "learning_rate": 6.79245283018868e-05,
      "loss": 0.7243556976318359,
      "step": 109
    },
    {
      "epoch": 0.10396975425330812,
      "grad_norm": 0.4128769371536337,
      "learning_rate": 6.855345911949685e-05,
      "loss": 0.5613802671432495,
      "step": 110
    },
    {
      "epoch": 0.10491493383742911,
      "grad_norm": 0.4202728678608408,
      "learning_rate": 6.918238993710691e-05,
      "loss": 0.6425524950027466,
      "step": 111
    },
    {
      "epoch": 0.10586011342155009,
      "grad_norm": 0.45388307887641244,
      "learning_rate": 6.981132075471698e-05,
      "loss": 0.57962965965271,
      "step": 112
    },
    {
      "epoch": 0.10680529300567108,
      "grad_norm": 0.453736477553693,
      "learning_rate": 7.044025157232704e-05,
      "loss": 0.6627553701400757,
      "step": 113
    },
    {
      "epoch": 0.10775047258979206,
      "grad_norm": 0.4603394443100072,
      "learning_rate": 7.10691823899371e-05,
      "loss": 0.73423171043396,
      "step": 114
    },
    {
      "epoch": 0.10869565217391304,
      "grad_norm": 0.41176066863999994,
      "learning_rate": 7.169811320754717e-05,
      "loss": 0.5957436561584473,
      "step": 115
    },
    {
      "epoch": 0.10964083175803403,
      "grad_norm": 0.4417585166261668,
      "learning_rate": 7.232704402515723e-05,
      "loss": 0.6240658164024353,
      "step": 116
    },
    {
      "epoch": 0.11058601134215501,
      "grad_norm": 0.4148139019853128,
      "learning_rate": 7.29559748427673e-05,
      "loss": 0.5719410181045532,
      "step": 117
    },
    {
      "epoch": 0.11153119092627599,
      "grad_norm": 0.42398221715268375,
      "learning_rate": 7.358490566037736e-05,
      "loss": 0.7426532506942749,
      "step": 118
    },
    {
      "epoch": 0.11247637051039698,
      "grad_norm": 0.48673966903017807,
      "learning_rate": 7.421383647798742e-05,
      "loss": 0.6333979964256287,
      "step": 119
    },
    {
      "epoch": 0.11342155009451796,
      "grad_norm": 0.447987774398833,
      "learning_rate": 7.484276729559749e-05,
      "loss": 0.6257369518280029,
      "step": 120
    },
    {
      "epoch": 0.11436672967863894,
      "grad_norm": 0.4439444003696264,
      "learning_rate": 7.547169811320755e-05,
      "loss": 0.7121595740318298,
      "step": 121
    },
    {
      "epoch": 0.11531190926275993,
      "grad_norm": 0.45648800162404896,
      "learning_rate": 7.610062893081762e-05,
      "loss": 0.7031106352806091,
      "step": 122
    },
    {
      "epoch": 0.11625708884688091,
      "grad_norm": 0.42521297254778423,
      "learning_rate": 7.672955974842768e-05,
      "loss": 0.6446982622146606,
      "step": 123
    },
    {
      "epoch": 0.11720226843100189,
      "grad_norm": 0.37400345342962577,
      "learning_rate": 7.735849056603774e-05,
      "loss": 0.5766610503196716,
      "step": 124
    },
    {
      "epoch": 0.11814744801512288,
      "grad_norm": 0.4253608862701931,
      "learning_rate": 7.798742138364781e-05,
      "loss": 0.5567334294319153,
      "step": 125
    },
    {
      "epoch": 0.11909262759924386,
      "grad_norm": 0.45800114452863416,
      "learning_rate": 7.861635220125787e-05,
      "loss": 0.6188783645629883,
      "step": 126
    },
    {
      "epoch": 0.12003780718336483,
      "grad_norm": 0.37595462840184657,
      "learning_rate": 7.924528301886794e-05,
      "loss": 0.602725625038147,
      "step": 127
    },
    {
      "epoch": 0.12098298676748583,
      "grad_norm": 0.42416524287697355,
      "learning_rate": 7.9874213836478e-05,
      "loss": 0.5110478401184082,
      "step": 128
    },
    {
      "epoch": 0.1219281663516068,
      "grad_norm": 0.41788617315567284,
      "learning_rate": 8.050314465408806e-05,
      "loss": 0.6451765298843384,
      "step": 129
    },
    {
      "epoch": 0.12287334593572778,
      "grad_norm": 0.47908049872690617,
      "learning_rate": 8.113207547169813e-05,
      "loss": 0.6317439079284668,
      "step": 130
    },
    {
      "epoch": 0.12381852551984877,
      "grad_norm": 0.41717306433670415,
      "learning_rate": 8.176100628930818e-05,
      "loss": 0.6793529987335205,
      "step": 131
    },
    {
      "epoch": 0.12476370510396975,
      "grad_norm": 0.4833943246396523,
      "learning_rate": 8.238993710691824e-05,
      "loss": 0.5261908769607544,
      "step": 132
    },
    {
      "epoch": 0.12570888468809074,
      "grad_norm": 0.4698186014963808,
      "learning_rate": 8.30188679245283e-05,
      "loss": 0.6723647117614746,
      "step": 133
    },
    {
      "epoch": 0.1266540642722117,
      "grad_norm": 0.46357938035216106,
      "learning_rate": 8.364779874213837e-05,
      "loss": 0.6001070737838745,
      "step": 134
    },
    {
      "epoch": 0.1275992438563327,
      "grad_norm": 0.4423526742173153,
      "learning_rate": 8.427672955974843e-05,
      "loss": 0.5585460066795349,
      "step": 135
    },
    {
      "epoch": 0.1285444234404537,
      "grad_norm": 0.3954931976456201,
      "learning_rate": 8.49056603773585e-05,
      "loss": 0.5827357172966003,
      "step": 136
    },
    {
      "epoch": 0.12948960302457466,
      "grad_norm": 0.3597225381282072,
      "learning_rate": 8.553459119496856e-05,
      "loss": 0.5420295000076294,
      "step": 137
    },
    {
      "epoch": 0.13043478260869565,
      "grad_norm": 0.4079671933255738,
      "learning_rate": 8.616352201257863e-05,
      "loss": 0.681174099445343,
      "step": 138
    },
    {
      "epoch": 0.13137996219281664,
      "grad_norm": 0.4083570565894232,
      "learning_rate": 8.679245283018869e-05,
      "loss": 0.6750494241714478,
      "step": 139
    },
    {
      "epoch": 0.1323251417769376,
      "grad_norm": 0.40858445095379614,
      "learning_rate": 8.742138364779875e-05,
      "loss": 0.6693521738052368,
      "step": 140
    },
    {
      "epoch": 0.1332703213610586,
      "grad_norm": 0.42395439783152244,
      "learning_rate": 8.805031446540882e-05,
      "loss": 0.6262834072113037,
      "step": 141
    },
    {
      "epoch": 0.1342155009451796,
      "grad_norm": 0.418749897509572,
      "learning_rate": 8.867924528301888e-05,
      "loss": 0.6805487871170044,
      "step": 142
    },
    {
      "epoch": 0.13516068052930058,
      "grad_norm": 0.3953284754149619,
      "learning_rate": 8.930817610062893e-05,
      "loss": 0.6563012599945068,
      "step": 143
    },
    {
      "epoch": 0.13610586011342155,
      "grad_norm": 0.39239893313026347,
      "learning_rate": 8.9937106918239e-05,
      "loss": 0.5636981725692749,
      "step": 144
    },
    {
      "epoch": 0.13705103969754254,
      "grad_norm": 0.34394705208906345,
      "learning_rate": 9.056603773584906e-05,
      "loss": 0.5491082668304443,
      "step": 145
    },
    {
      "epoch": 0.13799621928166353,
      "grad_norm": 0.4498589771759585,
      "learning_rate": 9.119496855345912e-05,
      "loss": 0.6469371318817139,
      "step": 146
    },
    {
      "epoch": 0.1389413988657845,
      "grad_norm": 0.39450146198538477,
      "learning_rate": 9.182389937106919e-05,
      "loss": 0.6369979977607727,
      "step": 147
    },
    {
      "epoch": 0.13988657844990549,
      "grad_norm": 0.4252807811943148,
      "learning_rate": 9.245283018867925e-05,
      "loss": 0.6126425862312317,
      "step": 148
    },
    {
      "epoch": 0.14083175803402648,
      "grad_norm": 0.37131166483832917,
      "learning_rate": 9.308176100628931e-05,
      "loss": 0.6006712913513184,
      "step": 149
    },
    {
      "epoch": 0.14177693761814744,
      "grad_norm": 0.3795760064010849,
      "learning_rate": 9.371069182389938e-05,
      "loss": 0.6794224381446838,
      "step": 150
    },
    {
      "epoch": 0.14272211720226843,
      "grad_norm": 0.3865308850490883,
      "learning_rate": 9.433962264150944e-05,
      "loss": 0.6338252425193787,
      "step": 151
    },
    {
      "epoch": 0.14366729678638943,
      "grad_norm": 0.4559942678879102,
      "learning_rate": 9.496855345911951e-05,
      "loss": 0.7483862042427063,
      "step": 152
    },
    {
      "epoch": 0.1446124763705104,
      "grad_norm": 0.40821520152312174,
      "learning_rate": 9.559748427672957e-05,
      "loss": 0.6107313632965088,
      "step": 153
    },
    {
      "epoch": 0.14555765595463138,
      "grad_norm": 0.45292579436330255,
      "learning_rate": 9.622641509433963e-05,
      "loss": 0.6730316877365112,
      "step": 154
    },
    {
      "epoch": 0.14650283553875237,
      "grad_norm": 0.420605567528612,
      "learning_rate": 9.685534591194969e-05,
      "loss": 0.6946786046028137,
      "step": 155
    },
    {
      "epoch": 0.14744801512287334,
      "grad_norm": 0.42346314496182763,
      "learning_rate": 9.748427672955975e-05,
      "loss": 0.6319793462753296,
      "step": 156
    },
    {
      "epoch": 0.14839319470699433,
      "grad_norm": 0.44681952499415445,
      "learning_rate": 9.811320754716981e-05,
      "loss": 0.7184162139892578,
      "step": 157
    },
    {
      "epoch": 0.14933837429111532,
      "grad_norm": 0.4043458195638719,
      "learning_rate": 9.874213836477988e-05,
      "loss": 0.6220781803131104,
      "step": 158
    },
    {
      "epoch": 0.1502835538752363,
      "grad_norm": 0.3580837041974643,
      "learning_rate": 9.937106918238994e-05,
      "loss": 0.6543900966644287,
      "step": 159
    },
    {
      "epoch": 0.15122873345935728,
      "grad_norm": 0.39199698155574575,
      "learning_rate": 0.0001,
      "loss": 0.5602942705154419,
      "step": 160
    },
    {
      "epoch": 0.15217391304347827,
      "grad_norm": 0.43202797717593505,
      "learning_rate": 9.99999728565478e-05,
      "loss": 0.5864697098731995,
      "step": 161
    },
    {
      "epoch": 0.15311909262759923,
      "grad_norm": 0.3607997001164942,
      "learning_rate": 9.999989142622071e-05,
      "loss": 0.721717119216919,
      "step": 162
    },
    {
      "epoch": 0.15406427221172023,
      "grad_norm": 0.33939664328271024,
      "learning_rate": 9.99997557091071e-05,
      "loss": 0.5623525381088257,
      "step": 163
    },
    {
      "epoch": 0.15500945179584122,
      "grad_norm": 0.3943539616273223,
      "learning_rate": 9.999956570535435e-05,
      "loss": 0.6893969774246216,
      "step": 164
    },
    {
      "epoch": 0.15595463137996218,
      "grad_norm": 0.46825085305034275,
      "learning_rate": 9.999932141516873e-05,
      "loss": 0.5889137983322144,
      "step": 165
    },
    {
      "epoch": 0.15689981096408318,
      "grad_norm": 0.3748357318120009,
      "learning_rate": 9.99990228388155e-05,
      "loss": 0.6009525656700134,
      "step": 166
    },
    {
      "epoch": 0.15784499054820417,
      "grad_norm": 0.3966977506932631,
      "learning_rate": 9.999866997661885e-05,
      "loss": 0.5179510712623596,
      "step": 167
    },
    {
      "epoch": 0.15879017013232513,
      "grad_norm": 0.38115052774109986,
      "learning_rate": 9.999826282896184e-05,
      "loss": 0.5693991184234619,
      "step": 168
    },
    {
      "epoch": 0.15973534971644612,
      "grad_norm": 0.4398524953961016,
      "learning_rate": 9.999780139628657e-05,
      "loss": 0.6648521423339844,
      "step": 169
    },
    {
      "epoch": 0.16068052930056712,
      "grad_norm": 0.3843272186069826,
      "learning_rate": 9.999728567909403e-05,
      "loss": 0.5415248274803162,
      "step": 170
    },
    {
      "epoch": 0.16162570888468808,
      "grad_norm": 0.43684399506530147,
      "learning_rate": 9.999671567794414e-05,
      "loss": 0.6328208446502686,
      "step": 171
    },
    {
      "epoch": 0.16257088846880907,
      "grad_norm": 0.42808409499253286,
      "learning_rate": 9.999609139345579e-05,
      "loss": 0.6938070058822632,
      "step": 172
    },
    {
      "epoch": 0.16351606805293006,
      "grad_norm": 0.3839958483186126,
      "learning_rate": 9.999541282630676e-05,
      "loss": 0.696319043636322,
      "step": 173
    },
    {
      "epoch": 0.16446124763705103,
      "grad_norm": 0.36219493228884453,
      "learning_rate": 9.999467997723381e-05,
      "loss": 0.5908347368240356,
      "step": 174
    },
    {
      "epoch": 0.16540642722117202,
      "grad_norm": 0.45230443823943245,
      "learning_rate": 9.999389284703265e-05,
      "loss": 0.6461359858512878,
      "step": 175
    },
    {
      "epoch": 0.166351606805293,
      "grad_norm": 0.3779402443053707,
      "learning_rate": 9.999305143655787e-05,
      "loss": 0.5622029304504395,
      "step": 176
    },
    {
      "epoch": 0.16729678638941398,
      "grad_norm": 0.3539576811928708,
      "learning_rate": 9.999215574672303e-05,
      "loss": 0.5924274921417236,
      "step": 177
    },
    {
      "epoch": 0.16824196597353497,
      "grad_norm": 0.366219909198924,
      "learning_rate": 9.99912057785006e-05,
      "loss": 0.5554662346839905,
      "step": 178
    },
    {
      "epoch": 0.16918714555765596,
      "grad_norm": 0.36151925897539333,
      "learning_rate": 9.9990201532922e-05,
      "loss": 0.5845496654510498,
      "step": 179
    },
    {
      "epoch": 0.17013232514177692,
      "grad_norm": 0.41139875574097134,
      "learning_rate": 9.99891430110776e-05,
      "loss": 0.5820896625518799,
      "step": 180
    },
    {
      "epoch": 0.17107750472589792,
      "grad_norm": 0.36413707919933913,
      "learning_rate": 9.998803021411667e-05,
      "loss": 0.5294264554977417,
      "step": 181
    },
    {
      "epoch": 0.1720226843100189,
      "grad_norm": 0.37274048618906097,
      "learning_rate": 9.998686314324739e-05,
      "loss": 0.6303973197937012,
      "step": 182
    },
    {
      "epoch": 0.17296786389413987,
      "grad_norm": 0.39041700426115933,
      "learning_rate": 9.99856417997369e-05,
      "loss": 0.639573872089386,
      "step": 183
    },
    {
      "epoch": 0.17391304347826086,
      "grad_norm": 0.4157403933560126,
      "learning_rate": 9.998436618491129e-05,
      "loss": 0.7660428881645203,
      "step": 184
    },
    {
      "epoch": 0.17485822306238186,
      "grad_norm": 0.3712304105218652,
      "learning_rate": 9.998303630015553e-05,
      "loss": 0.6309301853179932,
      "step": 185
    },
    {
      "epoch": 0.17580340264650285,
      "grad_norm": 0.37684799605263825,
      "learning_rate": 9.998165214691351e-05,
      "loss": 0.5865470170974731,
      "step": 186
    },
    {
      "epoch": 0.1767485822306238,
      "grad_norm": 0.3546374748287763,
      "learning_rate": 9.998021372668808e-05,
      "loss": 0.616080641746521,
      "step": 187
    },
    {
      "epoch": 0.1776937618147448,
      "grad_norm": 0.4154770496836085,
      "learning_rate": 9.997872104104097e-05,
      "loss": 0.631364107131958,
      "step": 188
    },
    {
      "epoch": 0.1786389413988658,
      "grad_norm": 0.3938301443968005,
      "learning_rate": 9.997717409159283e-05,
      "loss": 0.6043752431869507,
      "step": 189
    },
    {
      "epoch": 0.17958412098298676,
      "grad_norm": 0.3176530353446335,
      "learning_rate": 9.99755728800233e-05,
      "loss": 0.5220621228218079,
      "step": 190
    },
    {
      "epoch": 0.18052930056710775,
      "grad_norm": 0.3839464825071138,
      "learning_rate": 9.997391740807083e-05,
      "loss": 0.5944043397903442,
      "step": 191
    },
    {
      "epoch": 0.18147448015122875,
      "grad_norm": 0.4248877805827054,
      "learning_rate": 9.997220767753284e-05,
      "loss": 0.6977246999740601,
      "step": 192
    },
    {
      "epoch": 0.1824196597353497,
      "grad_norm": 0.3745237663269758,
      "learning_rate": 9.997044369026562e-05,
      "loss": 0.6076537370681763,
      "step": 193
    },
    {
      "epoch": 0.1833648393194707,
      "grad_norm": 0.3878536209882931,
      "learning_rate": 9.996862544818446e-05,
      "loss": 0.6019546985626221,
      "step": 194
    },
    {
      "epoch": 0.1843100189035917,
      "grad_norm": 0.5063435930426046,
      "learning_rate": 9.996675295326346e-05,
      "loss": 0.5709128975868225,
      "step": 195
    },
    {
      "epoch": 0.18525519848771266,
      "grad_norm": 0.38849834094273794,
      "learning_rate": 9.996482620753565e-05,
      "loss": 0.6035029888153076,
      "step": 196
    },
    {
      "epoch": 0.18620037807183365,
      "grad_norm": 0.3729992743825264,
      "learning_rate": 9.996284521309299e-05,
      "loss": 0.600641131401062,
      "step": 197
    },
    {
      "epoch": 0.18714555765595464,
      "grad_norm": 0.3726217830606135,
      "learning_rate": 9.99608099720863e-05,
      "loss": 0.6091938018798828,
      "step": 198
    },
    {
      "epoch": 0.1880907372400756,
      "grad_norm": 0.32840797917165926,
      "learning_rate": 9.995872048672534e-05,
      "loss": 0.5011407136917114,
      "step": 199
    },
    {
      "epoch": 0.1890359168241966,
      "grad_norm": 0.3621026494421897,
      "learning_rate": 9.995657675927874e-05,
      "loss": 0.6356720328330994,
      "step": 200
    },
    {
      "epoch": 0.1899810964083176,
      "grad_norm": 0.4008405438952466,
      "learning_rate": 9.995437879207401e-05,
      "loss": 0.5381849408149719,
      "step": 201
    },
    {
      "epoch": 0.19092627599243855,
      "grad_norm": 0.38102082724774466,
      "learning_rate": 9.99521265874976e-05,
      "loss": 0.5922316312789917,
      "step": 202
    },
    {
      "epoch": 0.19187145557655955,
      "grad_norm": 0.36571753672968704,
      "learning_rate": 9.994982014799477e-05,
      "loss": 0.5625612139701843,
      "step": 203
    },
    {
      "epoch": 0.19281663516068054,
      "grad_norm": 0.37563756360283856,
      "learning_rate": 9.994745947606974e-05,
      "loss": 0.5532487630844116,
      "step": 204
    },
    {
      "epoch": 0.1937618147448015,
      "grad_norm": 0.39349128501050157,
      "learning_rate": 9.994504457428558e-05,
      "loss": 0.5845286846160889,
      "step": 205
    },
    {
      "epoch": 0.1947069943289225,
      "grad_norm": 0.41126573330516353,
      "learning_rate": 9.994257544526422e-05,
      "loss": 0.6356018781661987,
      "step": 206
    },
    {
      "epoch": 0.1956521739130435,
      "grad_norm": 0.39832019948239455,
      "learning_rate": 9.994005209168652e-05,
      "loss": 0.5959442853927612,
      "step": 207
    },
    {
      "epoch": 0.19659735349716445,
      "grad_norm": 0.3819922654790408,
      "learning_rate": 9.993747451629216e-05,
      "loss": 0.5553781390190125,
      "step": 208
    },
    {
      "epoch": 0.19754253308128544,
      "grad_norm": 0.41362853870636995,
      "learning_rate": 9.99348427218797e-05,
      "loss": 0.5777321457862854,
      "step": 209
    },
    {
      "epoch": 0.19848771266540643,
      "grad_norm": 0.39991785469445645,
      "learning_rate": 9.993215671130662e-05,
      "loss": 0.6610274314880371,
      "step": 210
    },
    {
      "epoch": 0.1994328922495274,
      "grad_norm": 0.3526985653381435,
      "learning_rate": 9.992941648748918e-05,
      "loss": 0.5998183488845825,
      "step": 211
    },
    {
      "epoch": 0.2003780718336484,
      "grad_norm": 0.3556182689008548,
      "learning_rate": 9.992662205340257e-05,
      "loss": 0.5697258114814758,
      "step": 212
    },
    {
      "epoch": 0.20132325141776938,
      "grad_norm": 0.37616827943723485,
      "learning_rate": 9.992377341208078e-05,
      "loss": 0.6659120321273804,
      "step": 213
    },
    {
      "epoch": 0.20226843100189035,
      "grad_norm": 0.3639565844935246,
      "learning_rate": 9.992087056661677e-05,
      "loss": 0.6644794940948486,
      "step": 214
    },
    {
      "epoch": 0.20321361058601134,
      "grad_norm": 0.3810247570549747,
      "learning_rate": 9.991791352016217e-05,
      "loss": 0.5549978017807007,
      "step": 215
    },
    {
      "epoch": 0.20415879017013233,
      "grad_norm": 0.37631908505449996,
      "learning_rate": 9.991490227592763e-05,
      "loss": 0.611125111579895,
      "step": 216
    },
    {
      "epoch": 0.2051039697542533,
      "grad_norm": 0.4158652471378538,
      "learning_rate": 9.991183683718253e-05,
      "loss": 0.6398652791976929,
      "step": 217
    },
    {
      "epoch": 0.2060491493383743,
      "grad_norm": 0.3280641530513564,
      "learning_rate": 9.990871720725515e-05,
      "loss": 0.5115941166877747,
      "step": 218
    },
    {
      "epoch": 0.20699432892249528,
      "grad_norm": 0.37709204782478495,
      "learning_rate": 9.99055433895326e-05,
      "loss": 0.5907667279243469,
      "step": 219
    },
    {
      "epoch": 0.20793950850661624,
      "grad_norm": 0.3471149645107163,
      "learning_rate": 9.99023153874608e-05,
      "loss": 0.6614482998847961,
      "step": 220
    },
    {
      "epoch": 0.20888468809073724,
      "grad_norm": 0.3989710480740663,
      "learning_rate": 9.989903320454453e-05,
      "loss": 0.6948390007019043,
      "step": 221
    },
    {
      "epoch": 0.20982986767485823,
      "grad_norm": 0.37613000955093745,
      "learning_rate": 9.989569684434737e-05,
      "loss": 0.6537297964096069,
      "step": 222
    },
    {
      "epoch": 0.2107750472589792,
      "grad_norm": 0.3359218718933988,
      "learning_rate": 9.989230631049171e-05,
      "loss": 0.5869542360305786,
      "step": 223
    },
    {
      "epoch": 0.21172022684310018,
      "grad_norm": 0.3298714428175111,
      "learning_rate": 9.988886160665883e-05,
      "loss": 0.5069009065628052,
      "step": 224
    },
    {
      "epoch": 0.21266540642722118,
      "grad_norm": 0.357431582001639,
      "learning_rate": 9.988536273658876e-05,
      "loss": 0.6099156737327576,
      "step": 225
    },
    {
      "epoch": 0.21361058601134217,
      "grad_norm": 0.3729419199699211,
      "learning_rate": 9.988180970408032e-05,
      "loss": 0.5534413456916809,
      "step": 226
    },
    {
      "epoch": 0.21455576559546313,
      "grad_norm": 0.37132103072563216,
      "learning_rate": 9.987820251299122e-05,
      "loss": 0.7166958451271057,
      "step": 227
    },
    {
      "epoch": 0.21550094517958412,
      "grad_norm": 0.4397707152017941,
      "learning_rate": 9.98745411672379e-05,
      "loss": 0.6465284824371338,
      "step": 228
    },
    {
      "epoch": 0.21644612476370512,
      "grad_norm": 0.3644027807388181,
      "learning_rate": 9.987082567079563e-05,
      "loss": 0.5845780968666077,
      "step": 229
    },
    {
      "epoch": 0.21739130434782608,
      "grad_norm": 0.37980755747978723,
      "learning_rate": 9.986705602769847e-05,
      "loss": 0.5225163698196411,
      "step": 230
    },
    {
      "epoch": 0.21833648393194707,
      "grad_norm": 0.4094763074486131,
      "learning_rate": 9.986323224203926e-05,
      "loss": 0.6527047157287598,
      "step": 231
    },
    {
      "epoch": 0.21928166351606806,
      "grad_norm": 0.4117176755248091,
      "learning_rate": 9.985935431796962e-05,
      "loss": 0.7079921960830688,
      "step": 232
    },
    {
      "epoch": 0.22022684310018903,
      "grad_norm": 0.3456231030521,
      "learning_rate": 9.985542225969998e-05,
      "loss": 0.5491409301757812,
      "step": 233
    },
    {
      "epoch": 0.22117202268431002,
      "grad_norm": 0.35277355107513503,
      "learning_rate": 9.985143607149951e-05,
      "loss": 0.5956839323043823,
      "step": 234
    },
    {
      "epoch": 0.222117202268431,
      "grad_norm": 0.3886944547772192,
      "learning_rate": 9.984739575769618e-05,
      "loss": 0.6215295791625977,
      "step": 235
    },
    {
      "epoch": 0.22306238185255198,
      "grad_norm": 0.3432969344771943,
      "learning_rate": 9.98433013226767e-05,
      "loss": 0.5300068259239197,
      "step": 236
    },
    {
      "epoch": 0.22400756143667297,
      "grad_norm": 0.36095244555765243,
      "learning_rate": 9.983915277088656e-05,
      "loss": 0.5080676674842834,
      "step": 237
    },
    {
      "epoch": 0.22495274102079396,
      "grad_norm": 0.4344079024065252,
      "learning_rate": 9.983495010683e-05,
      "loss": 0.6555147171020508,
      "step": 238
    },
    {
      "epoch": 0.22589792060491493,
      "grad_norm": 0.39338602422967317,
      "learning_rate": 9.983069333507e-05,
      "loss": 0.619961142539978,
      "step": 239
    },
    {
      "epoch": 0.22684310018903592,
      "grad_norm": 0.3362023895400619,
      "learning_rate": 9.982638246022831e-05,
      "loss": 0.5172266364097595,
      "step": 240
    },
    {
      "epoch": 0.2277882797731569,
      "grad_norm": 0.3925919641825858,
      "learning_rate": 9.982201748698542e-05,
      "loss": 0.5848421454429626,
      "step": 241
    },
    {
      "epoch": 0.22873345935727787,
      "grad_norm": 0.4226943262565782,
      "learning_rate": 9.981759842008053e-05,
      "loss": 0.5889077186584473,
      "step": 242
    },
    {
      "epoch": 0.22967863894139887,
      "grad_norm": 0.3583251027614852,
      "learning_rate": 9.98131252643116e-05,
      "loss": 0.5855844020843506,
      "step": 243
    },
    {
      "epoch": 0.23062381852551986,
      "grad_norm": 0.399922649782656,
      "learning_rate": 9.980859802453533e-05,
      "loss": 0.622330367565155,
      "step": 244
    },
    {
      "epoch": 0.23156899810964082,
      "grad_norm": 0.39751223618593723,
      "learning_rate": 9.980401670566706e-05,
      "loss": 0.5470833778381348,
      "step": 245
    },
    {
      "epoch": 0.23251417769376181,
      "grad_norm": 0.39536463519501225,
      "learning_rate": 9.979938131268095e-05,
      "loss": 0.5675852298736572,
      "step": 246
    },
    {
      "epoch": 0.2334593572778828,
      "grad_norm": 0.44039632511182975,
      "learning_rate": 9.97946918506098e-05,
      "loss": 0.6296347379684448,
      "step": 247
    },
    {
      "epoch": 0.23440453686200377,
      "grad_norm": 0.35167232940028387,
      "learning_rate": 9.978994832454513e-05,
      "loss": 0.6166523694992065,
      "step": 248
    },
    {
      "epoch": 0.23534971644612476,
      "grad_norm": 0.39101792599897217,
      "learning_rate": 9.978515073963721e-05,
      "loss": 0.5897223353385925,
      "step": 249
    },
    {
      "epoch": 0.23629489603024575,
      "grad_norm": 0.40279439468158373,
      "learning_rate": 9.978029910109491e-05,
      "loss": 0.6157960891723633,
      "step": 250
    },
    {
      "epoch": 0.23724007561436672,
      "grad_norm": 0.381927992512547,
      "learning_rate": 9.977539341418587e-05,
      "loss": 0.5962035655975342,
      "step": 251
    },
    {
      "epoch": 0.2381852551984877,
      "grad_norm": 0.4007638074844483,
      "learning_rate": 9.977043368423635e-05,
      "loss": 0.5617565512657166,
      "step": 252
    },
    {
      "epoch": 0.2391304347826087,
      "grad_norm": 0.3733020200988004,
      "learning_rate": 9.976541991663135e-05,
      "loss": 0.5673946142196655,
      "step": 253
    },
    {
      "epoch": 0.24007561436672967,
      "grad_norm": 0.421355991384445,
      "learning_rate": 9.976035211681451e-05,
      "loss": 0.6756607294082642,
      "step": 254
    },
    {
      "epoch": 0.24102079395085066,
      "grad_norm": 0.3264577905723408,
      "learning_rate": 9.975523029028811e-05,
      "loss": 0.5740541219711304,
      "step": 255
    },
    {
      "epoch": 0.24196597353497165,
      "grad_norm": 0.36766330257949453,
      "learning_rate": 9.975005444261312e-05,
      "loss": 0.5955086350440979,
      "step": 256
    },
    {
      "epoch": 0.24291115311909262,
      "grad_norm": 0.4200110649327692,
      "learning_rate": 9.974482457940917e-05,
      "loss": 0.6896380186080933,
      "step": 257
    },
    {
      "epoch": 0.2438563327032136,
      "grad_norm": 0.3753844581967504,
      "learning_rate": 9.973954070635449e-05,
      "loss": 0.6333515048027039,
      "step": 258
    },
    {
      "epoch": 0.2448015122873346,
      "grad_norm": 0.40839466208505243,
      "learning_rate": 9.973420282918601e-05,
      "loss": 0.5618309378623962,
      "step": 259
    },
    {
      "epoch": 0.24574669187145556,
      "grad_norm": 0.3881727481712547,
      "learning_rate": 9.972881095369926e-05,
      "loss": 0.6062426567077637,
      "step": 260
    },
    {
      "epoch": 0.24669187145557656,
      "grad_norm": 0.37710900196223307,
      "learning_rate": 9.97233650857484e-05,
      "loss": 0.6826722621917725,
      "step": 261
    },
    {
      "epoch": 0.24763705103969755,
      "grad_norm": 0.397625617907197,
      "learning_rate": 9.971786523124622e-05,
      "loss": 0.5476322770118713,
      "step": 262
    },
    {
      "epoch": 0.2485822306238185,
      "grad_norm": 0.34398231215911634,
      "learning_rate": 9.971231139616413e-05,
      "loss": 0.5735257267951965,
      "step": 263
    },
    {
      "epoch": 0.2495274102079395,
      "grad_norm": 0.32905544181290036,
      "learning_rate": 9.970670358653212e-05,
      "loss": 0.5083062648773193,
      "step": 264
    },
    {
      "epoch": 0.2504725897920605,
      "grad_norm": 9.424595187865636,
      "learning_rate": 9.97010418084388e-05,
      "loss": 0.7101013660430908,
      "step": 265
    },
    {
      "epoch": 0.2514177693761815,
      "grad_norm": 0.3401105084008698,
      "learning_rate": 9.969532606803141e-05,
      "loss": 0.5444250106811523,
      "step": 266
    },
    {
      "epoch": 0.2523629489603025,
      "grad_norm": 0.34142733858892826,
      "learning_rate": 9.968955637151573e-05,
      "loss": 0.6701530814170837,
      "step": 267
    },
    {
      "epoch": 0.2533081285444234,
      "grad_norm": 0.3556290877254379,
      "learning_rate": 9.968373272515612e-05,
      "loss": 0.5110030174255371,
      "step": 268
    },
    {
      "epoch": 0.2542533081285444,
      "grad_norm": 0.35188359393429336,
      "learning_rate": 9.967785513527557e-05,
      "loss": 0.6104820966720581,
      "step": 269
    },
    {
      "epoch": 0.2551984877126654,
      "grad_norm": 0.34085252273372774,
      "learning_rate": 9.967192360825557e-05,
      "loss": 0.6367886662483215,
      "step": 270
    },
    {
      "epoch": 0.2561436672967864,
      "grad_norm": 0.433681567378444,
      "learning_rate": 9.966593815053622e-05,
      "loss": 0.5972983837127686,
      "step": 271
    },
    {
      "epoch": 0.2570888468809074,
      "grad_norm": 0.4305375387596,
      "learning_rate": 9.965989876861616e-05,
      "loss": 0.6534222364425659,
      "step": 272
    },
    {
      "epoch": 0.2580340264650284,
      "grad_norm": 0.35059558227484766,
      "learning_rate": 9.96538054690526e-05,
      "loss": 0.6173360347747803,
      "step": 273
    },
    {
      "epoch": 0.2589792060491493,
      "grad_norm": 0.35337731989731325,
      "learning_rate": 9.964765825846122e-05,
      "loss": 0.5213968753814697,
      "step": 274
    },
    {
      "epoch": 0.2599243856332703,
      "grad_norm": 0.3200451681879418,
      "learning_rate": 9.964145714351631e-05,
      "loss": 0.6102221012115479,
      "step": 275
    },
    {
      "epoch": 0.2608695652173913,
      "grad_norm": 0.3472408624907474,
      "learning_rate": 9.963520213095065e-05,
      "loss": 0.5472086668014526,
      "step": 276
    },
    {
      "epoch": 0.2618147448015123,
      "grad_norm": 0.30994670319955053,
      "learning_rate": 9.962889322755555e-05,
      "loss": 0.5413646697998047,
      "step": 277
    },
    {
      "epoch": 0.2627599243856333,
      "grad_norm": 0.3808539462781725,
      "learning_rate": 9.962253044018082e-05,
      "loss": 0.6621538400650024,
      "step": 278
    },
    {
      "epoch": 0.2637051039697543,
      "grad_norm": 0.40312414704006944,
      "learning_rate": 9.961611377573478e-05,
      "loss": 0.6600912809371948,
      "step": 279
    },
    {
      "epoch": 0.2646502835538752,
      "grad_norm": 0.3287896492556474,
      "learning_rate": 9.960964324118426e-05,
      "loss": 0.5004773736000061,
      "step": 280
    },
    {
      "epoch": 0.2655954631379962,
      "grad_norm": 0.37192250347721567,
      "learning_rate": 9.960311884355456e-05,
      "loss": 0.5551917552947998,
      "step": 281
    },
    {
      "epoch": 0.2665406427221172,
      "grad_norm": 0.3948406201939902,
      "learning_rate": 9.959654058992947e-05,
      "loss": 0.6206574440002441,
      "step": 282
    },
    {
      "epoch": 0.2674858223062382,
      "grad_norm": 0.3798823300610819,
      "learning_rate": 9.958990848745123e-05,
      "loss": 0.6511093974113464,
      "step": 283
    },
    {
      "epoch": 0.2684310018903592,
      "grad_norm": 0.38983932754135925,
      "learning_rate": 9.958322254332058e-05,
      "loss": 0.6653543710708618,
      "step": 284
    },
    {
      "epoch": 0.26937618147448017,
      "grad_norm": 0.3317963348540956,
      "learning_rate": 9.95764827647967e-05,
      "loss": 0.5058095455169678,
      "step": 285
    },
    {
      "epoch": 0.27032136105860116,
      "grad_norm": 0.36933580773672814,
      "learning_rate": 9.956968915919725e-05,
      "loss": 0.6194404363632202,
      "step": 286
    },
    {
      "epoch": 0.2712665406427221,
      "grad_norm": 0.3402625269522336,
      "learning_rate": 9.956284173389826e-05,
      "loss": 0.5294819474220276,
      "step": 287
    },
    {
      "epoch": 0.2722117202268431,
      "grad_norm": 0.36511631386030957,
      "learning_rate": 9.955594049633427e-05,
      "loss": 0.5984237194061279,
      "step": 288
    },
    {
      "epoch": 0.2731568998109641,
      "grad_norm": 0.35892918257570355,
      "learning_rate": 9.954898545399822e-05,
      "loss": 0.6415587663650513,
      "step": 289
    },
    {
      "epoch": 0.2741020793950851,
      "grad_norm": 0.3480339919866449,
      "learning_rate": 9.954197661444147e-05,
      "loss": 0.5517644882202148,
      "step": 290
    },
    {
      "epoch": 0.27504725897920607,
      "grad_norm": 0.3624564487453845,
      "learning_rate": 9.953491398527375e-05,
      "loss": 0.5582977533340454,
      "step": 291
    },
    {
      "epoch": 0.27599243856332706,
      "grad_norm": 0.41492441870315366,
      "learning_rate": 9.952779757416325e-05,
      "loss": 0.503410816192627,
      "step": 292
    },
    {
      "epoch": 0.276937618147448,
      "grad_norm": 0.3681745419744239,
      "learning_rate": 9.952062738883652e-05,
      "loss": 0.5839820504188538,
      "step": 293
    },
    {
      "epoch": 0.277882797731569,
      "grad_norm": 0.35755658149571834,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.5866888761520386,
      "step": 294
    },
    {
      "epoch": 0.27882797731569,
      "grad_norm": 0.3432996635684639,
      "learning_rate": 9.950612572673255e-05,
      "loss": 0.5801386833190918,
      "step": 295
    },
    {
      "epoch": 0.27977315689981097,
      "grad_norm": 0.38687608246853017,
      "learning_rate": 9.949879426570032e-05,
      "loss": 0.5618935227394104,
      "step": 296
    },
    {
      "epoch": 0.28071833648393196,
      "grad_norm": 0.3796153897010662,
      "learning_rate": 9.949140906194183e-05,
      "loss": 0.5681856870651245,
      "step": 297
    },
    {
      "epoch": 0.28166351606805295,
      "grad_norm": 0.3381825343824669,
      "learning_rate": 9.948397012347554e-05,
      "loss": 0.65195631980896,
      "step": 298
    },
    {
      "epoch": 0.2826086956521739,
      "grad_norm": 0.32643737768420383,
      "learning_rate": 9.947647745837814e-05,
      "loss": 0.5328071117401123,
      "step": 299
    },
    {
      "epoch": 0.2835538752362949,
      "grad_norm": 0.3348628395685986,
      "learning_rate": 9.946893107478473e-05,
      "loss": 0.575130820274353,
      "step": 300
    },
    {
      "epoch": 0.2844990548204159,
      "grad_norm": 0.4236724252628029,
      "learning_rate": 9.946133098088868e-05,
      "loss": 0.7859106063842773,
      "step": 301
    },
    {
      "epoch": 0.28544423440453687,
      "grad_norm": 0.42326761315174544,
      "learning_rate": 9.945367718494173e-05,
      "loss": 0.7460644245147705,
      "step": 302
    },
    {
      "epoch": 0.28638941398865786,
      "grad_norm": 0.3491028935950423,
      "learning_rate": 9.944596969525388e-05,
      "loss": 0.6303318738937378,
      "step": 303
    },
    {
      "epoch": 0.28733459357277885,
      "grad_norm": 0.3653600148403433,
      "learning_rate": 9.943820852019344e-05,
      "loss": 0.5800546407699585,
      "step": 304
    },
    {
      "epoch": 0.2882797731568998,
      "grad_norm": 0.34349438150235,
      "learning_rate": 9.943039366818704e-05,
      "loss": 0.5446357131004333,
      "step": 305
    },
    {
      "epoch": 0.2892249527410208,
      "grad_norm": 0.3425626603920122,
      "learning_rate": 9.942252514771954e-05,
      "loss": 0.5568522214889526,
      "step": 306
    },
    {
      "epoch": 0.29017013232514177,
      "grad_norm": 0.38821643012408213,
      "learning_rate": 9.94146029673341e-05,
      "loss": 0.5952762365341187,
      "step": 307
    },
    {
      "epoch": 0.29111531190926276,
      "grad_norm": 0.38885030154486183,
      "learning_rate": 9.940662713563211e-05,
      "loss": 0.5902369022369385,
      "step": 308
    },
    {
      "epoch": 0.29206049149338376,
      "grad_norm": 0.3736232797463254,
      "learning_rate": 9.939859766127327e-05,
      "loss": 0.594128429889679,
      "step": 309
    },
    {
      "epoch": 0.29300567107750475,
      "grad_norm": 0.37609543171267457,
      "learning_rate": 9.939051455297547e-05,
      "loss": 0.6968319416046143,
      "step": 310
    },
    {
      "epoch": 0.2939508506616257,
      "grad_norm": 0.3504934479508785,
      "learning_rate": 9.938237781951486e-05,
      "loss": 0.5792875289916992,
      "step": 311
    },
    {
      "epoch": 0.2948960302457467,
      "grad_norm": 0.36896053764711073,
      "learning_rate": 9.937418746972578e-05,
      "loss": 0.6154212355613708,
      "step": 312
    },
    {
      "epoch": 0.29584120982986767,
      "grad_norm": 0.3804345264058515,
      "learning_rate": 9.936594351250082e-05,
      "loss": 0.6264108419418335,
      "step": 313
    },
    {
      "epoch": 0.29678638941398866,
      "grad_norm": 0.33787302966258387,
      "learning_rate": 9.935764595679075e-05,
      "loss": 0.5943938493728638,
      "step": 314
    },
    {
      "epoch": 0.29773156899810965,
      "grad_norm": 0.36928472062158213,
      "learning_rate": 9.934929481160455e-05,
      "loss": 0.5380959510803223,
      "step": 315
    },
    {
      "epoch": 0.29867674858223064,
      "grad_norm": 0.3155197055140746,
      "learning_rate": 9.934089008600937e-05,
      "loss": 0.5340145826339722,
      "step": 316
    },
    {
      "epoch": 0.2996219281663516,
      "grad_norm": 0.3673002930453005,
      "learning_rate": 9.933243178913054e-05,
      "loss": 0.6554127931594849,
      "step": 317
    },
    {
      "epoch": 0.3005671077504726,
      "grad_norm": 0.3529002264213382,
      "learning_rate": 9.932391993015157e-05,
      "loss": 0.6091997027397156,
      "step": 318
    },
    {
      "epoch": 0.30151228733459357,
      "grad_norm": 0.373032734689169,
      "learning_rate": 9.931535451831409e-05,
      "loss": 0.5740947127342224,
      "step": 319
    },
    {
      "epoch": 0.30245746691871456,
      "grad_norm": 0.38480743544769985,
      "learning_rate": 9.93067355629179e-05,
      "loss": 0.6777826547622681,
      "step": 320
    },
    {
      "epoch": 0.30340264650283555,
      "grad_norm": 0.3806604147430199,
      "learning_rate": 9.929806307332092e-05,
      "loss": 0.6820289492607117,
      "step": 321
    },
    {
      "epoch": 0.30434782608695654,
      "grad_norm": 0.3610366520703887,
      "learning_rate": 9.928933705893924e-05,
      "loss": 0.6171772480010986,
      "step": 322
    },
    {
      "epoch": 0.3052930056710775,
      "grad_norm": 0.39144320252546855,
      "learning_rate": 9.928055752924697e-05,
      "loss": 0.6238487958908081,
      "step": 323
    },
    {
      "epoch": 0.30623818525519847,
      "grad_norm": 0.428104926895194,
      "learning_rate": 9.927172449377643e-05,
      "loss": 0.5898630619049072,
      "step": 324
    },
    {
      "epoch": 0.30718336483931946,
      "grad_norm": 0.3591278029119663,
      "learning_rate": 9.926283796211795e-05,
      "loss": 0.646600604057312,
      "step": 325
    },
    {
      "epoch": 0.30812854442344045,
      "grad_norm": 0.38811598686283805,
      "learning_rate": 9.925389794391999e-05,
      "loss": 0.6077823638916016,
      "step": 326
    },
    {
      "epoch": 0.30907372400756145,
      "grad_norm": 0.3926395818915623,
      "learning_rate": 9.924490444888907e-05,
      "loss": 0.629550576210022,
      "step": 327
    },
    {
      "epoch": 0.31001890359168244,
      "grad_norm": 0.37896894807227066,
      "learning_rate": 9.923585748678977e-05,
      "loss": 0.6644381284713745,
      "step": 328
    },
    {
      "epoch": 0.31096408317580343,
      "grad_norm": 0.39204177951205377,
      "learning_rate": 9.922675706744469e-05,
      "loss": 0.6122768521308899,
      "step": 329
    },
    {
      "epoch": 0.31190926275992437,
      "grad_norm": 0.38963091384073717,
      "learning_rate": 9.921760320073456e-05,
      "loss": 0.6454260945320129,
      "step": 330
    },
    {
      "epoch": 0.31285444234404536,
      "grad_norm": 0.374094003646931,
      "learning_rate": 9.920839589659803e-05,
      "loss": 0.5508807301521301,
      "step": 331
    },
    {
      "epoch": 0.31379962192816635,
      "grad_norm": 0.4165813384088629,
      "learning_rate": 9.919913516503186e-05,
      "loss": 0.7433034181594849,
      "step": 332
    },
    {
      "epoch": 0.31474480151228734,
      "grad_norm": 0.28890588014240937,
      "learning_rate": 9.918982101609074e-05,
      "loss": 0.5257115364074707,
      "step": 333
    },
    {
      "epoch": 0.31568998109640833,
      "grad_norm": 0.36540508519616766,
      "learning_rate": 9.918045345988742e-05,
      "loss": 0.6040310859680176,
      "step": 334
    },
    {
      "epoch": 0.3166351606805293,
      "grad_norm": 0.4083172902256693,
      "learning_rate": 9.917103250659262e-05,
      "loss": 0.6121755838394165,
      "step": 335
    },
    {
      "epoch": 0.31758034026465026,
      "grad_norm": 0.34348211819475744,
      "learning_rate": 9.916155816643501e-05,
      "loss": 0.5129197835922241,
      "step": 336
    },
    {
      "epoch": 0.31852551984877125,
      "grad_norm": 0.3839905022935845,
      "learning_rate": 9.915203044970125e-05,
      "loss": 0.5164848566055298,
      "step": 337
    },
    {
      "epoch": 0.31947069943289225,
      "grad_norm": 0.3730311444607243,
      "learning_rate": 9.914244936673594e-05,
      "loss": 0.6429909467697144,
      "step": 338
    },
    {
      "epoch": 0.32041587901701324,
      "grad_norm": 0.35972385550092023,
      "learning_rate": 9.913281492794164e-05,
      "loss": 0.5795838832855225,
      "step": 339
    },
    {
      "epoch": 0.32136105860113423,
      "grad_norm": 0.3514435205603149,
      "learning_rate": 9.91231271437788e-05,
      "loss": 0.6254268288612366,
      "step": 340
    },
    {
      "epoch": 0.3223062381852552,
      "grad_norm": 0.3858375635714363,
      "learning_rate": 9.911338602476585e-05,
      "loss": 0.6220166683197021,
      "step": 341
    },
    {
      "epoch": 0.32325141776937616,
      "grad_norm": 0.3655775139227761,
      "learning_rate": 9.910359158147908e-05,
      "loss": 0.5682015419006348,
      "step": 342
    },
    {
      "epoch": 0.32419659735349715,
      "grad_norm": 0.3591142250557085,
      "learning_rate": 9.909374382455269e-05,
      "loss": 0.6139419078826904,
      "step": 343
    },
    {
      "epoch": 0.32514177693761814,
      "grad_norm": 0.31786856385643114,
      "learning_rate": 9.908384276467875e-05,
      "loss": 0.5583867430686951,
      "step": 344
    },
    {
      "epoch": 0.32608695652173914,
      "grad_norm": 0.38144391001497396,
      "learning_rate": 9.907388841260723e-05,
      "loss": 0.515856921672821,
      "step": 345
    },
    {
      "epoch": 0.3270321361058601,
      "grad_norm": 0.3661213144576043,
      "learning_rate": 9.906388077914597e-05,
      "loss": 0.6757944822311401,
      "step": 346
    },
    {
      "epoch": 0.3279773156899811,
      "grad_norm": 0.3145154039629811,
      "learning_rate": 9.905381987516059e-05,
      "loss": 0.6330128908157349,
      "step": 347
    },
    {
      "epoch": 0.32892249527410206,
      "grad_norm": 0.3557810882324332,
      "learning_rate": 9.904370571157465e-05,
      "loss": 0.5805292129516602,
      "step": 348
    },
    {
      "epoch": 0.32986767485822305,
      "grad_norm": 0.33568973831983695,
      "learning_rate": 9.903353829936943e-05,
      "loss": 0.6091195344924927,
      "step": 349
    },
    {
      "epoch": 0.33081285444234404,
      "grad_norm": 0.3493887487413567,
      "learning_rate": 9.902331764958413e-05,
      "loss": 0.6094391942024231,
      "step": 350
    },
    {
      "epoch": 0.33175803402646503,
      "grad_norm": 0.36597702535132526,
      "learning_rate": 9.901304377331566e-05,
      "loss": 0.5626646280288696,
      "step": 351
    },
    {
      "epoch": 0.332703213610586,
      "grad_norm": 0.38035803936946083,
      "learning_rate": 9.900271668171878e-05,
      "loss": 0.566908597946167,
      "step": 352
    },
    {
      "epoch": 0.333648393194707,
      "grad_norm": 0.3691857563827574,
      "learning_rate": 9.8992336386006e-05,
      "loss": 0.6257275342941284,
      "step": 353
    },
    {
      "epoch": 0.33459357277882795,
      "grad_norm": 0.4286564333622215,
      "learning_rate": 9.89819028974476e-05,
      "loss": 0.6058813333511353,
      "step": 354
    },
    {
      "epoch": 0.33553875236294894,
      "grad_norm": 0.37566350959513656,
      "learning_rate": 9.89714162273716e-05,
      "loss": 0.6171496510505676,
      "step": 355
    },
    {
      "epoch": 0.33648393194706994,
      "grad_norm": 0.33486381322698244,
      "learning_rate": 9.896087638716381e-05,
      "loss": 0.5557661056518555,
      "step": 356
    },
    {
      "epoch": 0.33742911153119093,
      "grad_norm": 0.4358388983625812,
      "learning_rate": 9.895028338826773e-05,
      "loss": 0.6131457686424255,
      "step": 357
    },
    {
      "epoch": 0.3383742911153119,
      "grad_norm": 0.35106343011423297,
      "learning_rate": 9.893963724218455e-05,
      "loss": 0.5648720264434814,
      "step": 358
    },
    {
      "epoch": 0.3393194706994329,
      "grad_norm": 0.36353074722508305,
      "learning_rate": 9.892893796047323e-05,
      "loss": 0.6617948412895203,
      "step": 359
    },
    {
      "epoch": 0.34026465028355385,
      "grad_norm": 0.3393501859964104,
      "learning_rate": 9.891818555475037e-05,
      "loss": 0.6558154821395874,
      "step": 360
    },
    {
      "epoch": 0.34120982986767484,
      "grad_norm": 0.35644585014803015,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.5623412728309631,
      "step": 361
    },
    {
      "epoch": 0.34215500945179583,
      "grad_norm": 0.35227060717182374,
      "learning_rate": 9.889652141802492e-05,
      "loss": 0.6555649042129517,
      "step": 362
    },
    {
      "epoch": 0.3431001890359168,
      "grad_norm": 0.3524599219616938,
      "learning_rate": 9.88856097105439e-05,
      "loss": 0.6357150077819824,
      "step": 363
    },
    {
      "epoch": 0.3440453686200378,
      "grad_norm": 0.38448440421139757,
      "learning_rate": 9.887464492609446e-05,
      "loss": 0.6395916938781738,
      "step": 364
    },
    {
      "epoch": 0.3449905482041588,
      "grad_norm": 0.3218475961843421,
      "learning_rate": 9.886362707658152e-05,
      "loss": 0.45486485958099365,
      "step": 365
    },
    {
      "epoch": 0.34593572778827975,
      "grad_norm": 0.3572936983243142,
      "learning_rate": 9.885255617396755e-05,
      "loss": 0.5903646945953369,
      "step": 366
    },
    {
      "epoch": 0.34688090737240074,
      "grad_norm": 0.3063487980007302,
      "learning_rate": 9.884143223027266e-05,
      "loss": 0.4782397747039795,
      "step": 367
    },
    {
      "epoch": 0.34782608695652173,
      "grad_norm": 0.35546757543178437,
      "learning_rate": 9.883025525757457e-05,
      "loss": 0.5813631415367126,
      "step": 368
    },
    {
      "epoch": 0.3487712665406427,
      "grad_norm": 0.39739616946337075,
      "learning_rate": 9.881902526800847e-05,
      "loss": 0.6888124346733093,
      "step": 369
    },
    {
      "epoch": 0.3497164461247637,
      "grad_norm": 0.3646496045608875,
      "learning_rate": 9.880774227376726e-05,
      "loss": 0.6178686618804932,
      "step": 370
    },
    {
      "epoch": 0.3506616257088847,
      "grad_norm": 0.3698796838859708,
      "learning_rate": 9.87964062871013e-05,
      "loss": 0.6231975555419922,
      "step": 371
    },
    {
      "epoch": 0.3516068052930057,
      "grad_norm": 0.3771716862334163,
      "learning_rate": 9.878501732031848e-05,
      "loss": 0.6054439544677734,
      "step": 372
    },
    {
      "epoch": 0.35255198487712663,
      "grad_norm": 0.36499141231293375,
      "learning_rate": 9.877357538578425e-05,
      "loss": 0.5802357792854309,
      "step": 373
    },
    {
      "epoch": 0.3534971644612476,
      "grad_norm": 0.35648409592895586,
      "learning_rate": 9.876208049592153e-05,
      "loss": 0.584117591381073,
      "step": 374
    },
    {
      "epoch": 0.3544423440453686,
      "grad_norm": 0.32284685033851673,
      "learning_rate": 9.87505326632108e-05,
      "loss": 0.5701320171356201,
      "step": 375
    },
    {
      "epoch": 0.3553875236294896,
      "grad_norm": 0.3494539779254885,
      "learning_rate": 9.873893190018995e-05,
      "loss": 0.6342921257019043,
      "step": 376
    },
    {
      "epoch": 0.3563327032136106,
      "grad_norm": 0.36142391796353157,
      "learning_rate": 9.87272782194544e-05,
      "loss": 0.5499509572982788,
      "step": 377
    },
    {
      "epoch": 0.3572778827977316,
      "grad_norm": 0.3578468347536742,
      "learning_rate": 9.871557163365697e-05,
      "loss": 0.6342394948005676,
      "step": 378
    },
    {
      "epoch": 0.35822306238185253,
      "grad_norm": 0.34320590837142906,
      "learning_rate": 9.870381215550795e-05,
      "loss": 0.5861274003982544,
      "step": 379
    },
    {
      "epoch": 0.3591682419659735,
      "grad_norm": 0.34336401091240937,
      "learning_rate": 9.869199979777505e-05,
      "loss": 0.512317419052124,
      "step": 380
    },
    {
      "epoch": 0.3601134215500945,
      "grad_norm": 0.432030614931658,
      "learning_rate": 9.868013457328344e-05,
      "loss": 0.6610612869262695,
      "step": 381
    },
    {
      "epoch": 0.3610586011342155,
      "grad_norm": 0.3237818459365279,
      "learning_rate": 9.866821649491558e-05,
      "loss": 0.5680691599845886,
      "step": 382
    },
    {
      "epoch": 0.3620037807183365,
      "grad_norm": 0.3386618282190292,
      "learning_rate": 9.865624557561141e-05,
      "loss": 0.5289667844772339,
      "step": 383
    },
    {
      "epoch": 0.3629489603024575,
      "grad_norm": 0.37382170833135686,
      "learning_rate": 9.864422182836823e-05,
      "loss": 0.5868505239486694,
      "step": 384
    },
    {
      "epoch": 0.3638941398865784,
      "grad_norm": 0.35307846959339234,
      "learning_rate": 9.863214526624065e-05,
      "loss": 0.5574005246162415,
      "step": 385
    },
    {
      "epoch": 0.3648393194706994,
      "grad_norm": 0.39824692758284724,
      "learning_rate": 9.862001590234068e-05,
      "loss": 0.7115128040313721,
      "step": 386
    },
    {
      "epoch": 0.3657844990548204,
      "grad_norm": 0.36939780468089484,
      "learning_rate": 9.860783374983762e-05,
      "loss": 0.6597063541412354,
      "step": 387
    },
    {
      "epoch": 0.3667296786389414,
      "grad_norm": 0.40885405972516575,
      "learning_rate": 9.85955988219581e-05,
      "loss": 0.5782377123832703,
      "step": 388
    },
    {
      "epoch": 0.3676748582230624,
      "grad_norm": 0.39871301157574135,
      "learning_rate": 9.858331113198605e-05,
      "loss": 0.6167107820510864,
      "step": 389
    },
    {
      "epoch": 0.3686200378071834,
      "grad_norm": 0.3482967047481056,
      "learning_rate": 9.857097069326267e-05,
      "loss": 0.6152725219726562,
      "step": 390
    },
    {
      "epoch": 0.3695652173913043,
      "grad_norm": 0.371559320370247,
      "learning_rate": 9.855857751918645e-05,
      "loss": 0.6725733280181885,
      "step": 391
    },
    {
      "epoch": 0.3705103969754253,
      "grad_norm": 0.38903897863538167,
      "learning_rate": 9.854613162321313e-05,
      "loss": 0.7273510098457336,
      "step": 392
    },
    {
      "epoch": 0.3714555765595463,
      "grad_norm": 0.37479387467797665,
      "learning_rate": 9.85336330188557e-05,
      "loss": 0.5785548686981201,
      "step": 393
    },
    {
      "epoch": 0.3724007561436673,
      "grad_norm": 0.35419602815315465,
      "learning_rate": 9.852108171968436e-05,
      "loss": 0.5789729356765747,
      "step": 394
    },
    {
      "epoch": 0.3733459357277883,
      "grad_norm": 0.37971066925915886,
      "learning_rate": 9.850847773932656e-05,
      "loss": 0.6129900217056274,
      "step": 395
    },
    {
      "epoch": 0.3742911153119093,
      "grad_norm": 0.39790828252705013,
      "learning_rate": 9.84958210914669e-05,
      "loss": 0.6265106201171875,
      "step": 396
    },
    {
      "epoch": 0.3752362948960302,
      "grad_norm": 0.3610679655132291,
      "learning_rate": 9.848311178984718e-05,
      "loss": 0.633583128452301,
      "step": 397
    },
    {
      "epoch": 0.3761814744801512,
      "grad_norm": 0.3334317159532274,
      "learning_rate": 9.847034984826639e-05,
      "loss": 0.6356780529022217,
      "step": 398
    },
    {
      "epoch": 0.3771266540642722,
      "grad_norm": 0.34606761614769926,
      "learning_rate": 9.845753528058064e-05,
      "loss": 0.6111500859260559,
      "step": 399
    },
    {
      "epoch": 0.3780718336483932,
      "grad_norm": 0.3399523124217761,
      "learning_rate": 9.844466810070319e-05,
      "loss": 0.6542142629623413,
      "step": 400
    },
    {
      "epoch": 0.3790170132325142,
      "grad_norm": 0.31718360389700573,
      "learning_rate": 9.843174832260445e-05,
      "loss": 0.572300910949707,
      "step": 401
    },
    {
      "epoch": 0.3799621928166352,
      "grad_norm": 0.33402316533721305,
      "learning_rate": 9.84187759603119e-05,
      "loss": 0.5900975465774536,
      "step": 402
    },
    {
      "epoch": 0.3809073724007561,
      "grad_norm": 0.3494324070973723,
      "learning_rate": 9.840575102791013e-05,
      "loss": 0.5570684671401978,
      "step": 403
    },
    {
      "epoch": 0.3818525519848771,
      "grad_norm": 0.3072183809869307,
      "learning_rate": 9.839267353954081e-05,
      "loss": 0.509558379650116,
      "step": 404
    },
    {
      "epoch": 0.3827977315689981,
      "grad_norm": 0.31877657334359766,
      "learning_rate": 9.837954350940266e-05,
      "loss": 0.5568068027496338,
      "step": 405
    },
    {
      "epoch": 0.3837429111531191,
      "grad_norm": 0.3367483144097971,
      "learning_rate": 9.836636095175147e-05,
      "loss": 0.6476598978042603,
      "step": 406
    },
    {
      "epoch": 0.3846880907372401,
      "grad_norm": 0.3755623891095991,
      "learning_rate": 9.835312588090001e-05,
      "loss": 0.6324979066848755,
      "step": 407
    },
    {
      "epoch": 0.3856332703213611,
      "grad_norm": 0.31811672821365444,
      "learning_rate": 9.833983831121814e-05,
      "loss": 0.5544037818908691,
      "step": 408
    },
    {
      "epoch": 0.38657844990548207,
      "grad_norm": 0.3452412866713756,
      "learning_rate": 9.832649825713265e-05,
      "loss": 0.591183066368103,
      "step": 409
    },
    {
      "epoch": 0.387523629489603,
      "grad_norm": 0.3463156746899519,
      "learning_rate": 9.831310573312736e-05,
      "loss": 0.5394643545150757,
      "step": 410
    },
    {
      "epoch": 0.388468809073724,
      "grad_norm": 0.32470972143628735,
      "learning_rate": 9.829966075374305e-05,
      "loss": 0.5327455997467041,
      "step": 411
    },
    {
      "epoch": 0.389413988657845,
      "grad_norm": 0.3360598611609789,
      "learning_rate": 9.828616333357743e-05,
      "loss": 0.6194008588790894,
      "step": 412
    },
    {
      "epoch": 0.390359168241966,
      "grad_norm": 0.3359493974720307,
      "learning_rate": 9.827261348728518e-05,
      "loss": 0.674022912979126,
      "step": 413
    },
    {
      "epoch": 0.391304347826087,
      "grad_norm": 0.3196386806720542,
      "learning_rate": 9.825901122957786e-05,
      "loss": 0.5294101238250732,
      "step": 414
    },
    {
      "epoch": 0.39224952741020797,
      "grad_norm": 0.330592141678956,
      "learning_rate": 9.824535657522398e-05,
      "loss": 0.6000341176986694,
      "step": 415
    },
    {
      "epoch": 0.3931947069943289,
      "grad_norm": 0.35768227806019726,
      "learning_rate": 9.823164953904893e-05,
      "loss": 0.5129495859146118,
      "step": 416
    },
    {
      "epoch": 0.3941398865784499,
      "grad_norm": 0.3394307031822883,
      "learning_rate": 9.821789013593492e-05,
      "loss": 0.5186187028884888,
      "step": 417
    },
    {
      "epoch": 0.3950850661625709,
      "grad_norm": 0.35323455732575304,
      "learning_rate": 9.820407838082109e-05,
      "loss": 0.6206973791122437,
      "step": 418
    },
    {
      "epoch": 0.3960302457466919,
      "grad_norm": 0.34624193621601557,
      "learning_rate": 9.819021428870338e-05,
      "loss": 0.48554354906082153,
      "step": 419
    },
    {
      "epoch": 0.39697542533081287,
      "grad_norm": 0.3594856916856893,
      "learning_rate": 9.817629787463456e-05,
      "loss": 0.6015715599060059,
      "step": 420
    },
    {
      "epoch": 0.39792060491493386,
      "grad_norm": 0.36716377311572573,
      "learning_rate": 9.816232915372423e-05,
      "loss": 0.6753523945808411,
      "step": 421
    },
    {
      "epoch": 0.3988657844990548,
      "grad_norm": 0.3495428621324048,
      "learning_rate": 9.814830814113874e-05,
      "loss": 0.5887548923492432,
      "step": 422
    },
    {
      "epoch": 0.3998109640831758,
      "grad_norm": 0.3258476421671743,
      "learning_rate": 9.813423485210122e-05,
      "loss": 0.5979602336883545,
      "step": 423
    },
    {
      "epoch": 0.4007561436672968,
      "grad_norm": 0.36248830782168895,
      "learning_rate": 9.812010930189162e-05,
      "loss": 0.7491916418075562,
      "step": 424
    },
    {
      "epoch": 0.4017013232514178,
      "grad_norm": 0.37522220192032635,
      "learning_rate": 9.810593150584658e-05,
      "loss": 0.5701152682304382,
      "step": 425
    },
    {
      "epoch": 0.40264650283553877,
      "grad_norm": 0.3120756953364481,
      "learning_rate": 9.809170147935944e-05,
      "loss": 0.5853552222251892,
      "step": 426
    },
    {
      "epoch": 0.40359168241965976,
      "grad_norm": 0.36072977660850114,
      "learning_rate": 9.80774192378803e-05,
      "loss": 0.5467804670333862,
      "step": 427
    },
    {
      "epoch": 0.4045368620037807,
      "grad_norm": 0.38863649241121384,
      "learning_rate": 9.806308479691595e-05,
      "loss": 0.6084837913513184,
      "step": 428
    },
    {
      "epoch": 0.4054820415879017,
      "grad_norm": 0.477955926047549,
      "learning_rate": 9.804869817202981e-05,
      "loss": 0.748458981513977,
      "step": 429
    },
    {
      "epoch": 0.4064272211720227,
      "grad_norm": 0.4338540898193796,
      "learning_rate": 9.8034259378842e-05,
      "loss": 0.6354202032089233,
      "step": 430
    },
    {
      "epoch": 0.40737240075614367,
      "grad_norm": 0.40469160056086845,
      "learning_rate": 9.80197684330293e-05,
      "loss": 0.6297599077224731,
      "step": 431
    },
    {
      "epoch": 0.40831758034026466,
      "grad_norm": 0.4159581554371078,
      "learning_rate": 9.800522535032502e-05,
      "loss": 0.5999263525009155,
      "step": 432
    },
    {
      "epoch": 0.40926275992438566,
      "grad_norm": 0.3919869283310248,
      "learning_rate": 9.799063014651918e-05,
      "loss": 0.5557018518447876,
      "step": 433
    },
    {
      "epoch": 0.4102079395085066,
      "grad_norm": 0.38525525598322025,
      "learning_rate": 9.797598283745833e-05,
      "loss": 0.6159899234771729,
      "step": 434
    },
    {
      "epoch": 0.4111531190926276,
      "grad_norm": 0.31171191270194887,
      "learning_rate": 9.796128343904562e-05,
      "loss": 0.5955788493156433,
      "step": 435
    },
    {
      "epoch": 0.4120982986767486,
      "grad_norm": 0.32870083331947686,
      "learning_rate": 9.794653196724077e-05,
      "loss": 0.530913770198822,
      "step": 436
    },
    {
      "epoch": 0.41304347826086957,
      "grad_norm": 0.38288296647042336,
      "learning_rate": 9.793172843805996e-05,
      "loss": 0.5832242965698242,
      "step": 437
    },
    {
      "epoch": 0.41398865784499056,
      "grad_norm": 0.3591020937462836,
      "learning_rate": 9.791687286757599e-05,
      "loss": 0.6388247013092041,
      "step": 438
    },
    {
      "epoch": 0.41493383742911155,
      "grad_norm": 0.3307458225010303,
      "learning_rate": 9.790196527191811e-05,
      "loss": 0.5696872472763062,
      "step": 439
    },
    {
      "epoch": 0.4158790170132325,
      "grad_norm": 0.34064668928632474,
      "learning_rate": 9.788700566727205e-05,
      "loss": 0.579971194267273,
      "step": 440
    },
    {
      "epoch": 0.4168241965973535,
      "grad_norm": 0.39129788897066325,
      "learning_rate": 9.787199406988004e-05,
      "loss": 0.6262000799179077,
      "step": 441
    },
    {
      "epoch": 0.41776937618147447,
      "grad_norm": 0.3608919997567068,
      "learning_rate": 9.785693049604073e-05,
      "loss": 0.6142866611480713,
      "step": 442
    },
    {
      "epoch": 0.41871455576559546,
      "grad_norm": 0.3644881040657021,
      "learning_rate": 9.784181496210922e-05,
      "loss": 0.588630199432373,
      "step": 443
    },
    {
      "epoch": 0.41965973534971646,
      "grad_norm": 0.3929402658840271,
      "learning_rate": 9.782664748449703e-05,
      "loss": 0.6270040273666382,
      "step": 444
    },
    {
      "epoch": 0.42060491493383745,
      "grad_norm": 0.3793182401494153,
      "learning_rate": 9.781142807967205e-05,
      "loss": 0.6046649217605591,
      "step": 445
    },
    {
      "epoch": 0.4215500945179584,
      "grad_norm": 0.34725672409824543,
      "learning_rate": 9.779615676415858e-05,
      "loss": 0.5773067474365234,
      "step": 446
    },
    {
      "epoch": 0.4224952741020794,
      "grad_norm": 0.31575529566389754,
      "learning_rate": 9.778083355453726e-05,
      "loss": 0.48710280656814575,
      "step": 447
    },
    {
      "epoch": 0.42344045368620037,
      "grad_norm": 0.3939449670432832,
      "learning_rate": 9.776545846744509e-05,
      "loss": 0.5672383308410645,
      "step": 448
    },
    {
      "epoch": 0.42438563327032136,
      "grad_norm": 0.3334834851985572,
      "learning_rate": 9.77500315195754e-05,
      "loss": 0.5756024718284607,
      "step": 449
    },
    {
      "epoch": 0.42533081285444235,
      "grad_norm": 0.33376684098998277,
      "learning_rate": 9.773455272767779e-05,
      "loss": 0.5812773704528809,
      "step": 450
    },
    {
      "epoch": 0.42627599243856334,
      "grad_norm": 0.3792398337515658,
      "learning_rate": 9.771902210855819e-05,
      "loss": 0.6274706125259399,
      "step": 451
    },
    {
      "epoch": 0.42722117202268434,
      "grad_norm": 0.3286092352092712,
      "learning_rate": 9.770343967907877e-05,
      "loss": 0.5544033646583557,
      "step": 452
    },
    {
      "epoch": 0.4281663516068053,
      "grad_norm": 0.3586082032841436,
      "learning_rate": 9.768780545615799e-05,
      "loss": 0.5674200654029846,
      "step": 453
    },
    {
      "epoch": 0.42911153119092627,
      "grad_norm": 0.34491171671492094,
      "learning_rate": 9.767211945677051e-05,
      "loss": 0.5779148936271667,
      "step": 454
    },
    {
      "epoch": 0.43005671077504726,
      "grad_norm": 0.383349065889785,
      "learning_rate": 9.765638169794719e-05,
      "loss": 0.6760722398757935,
      "step": 455
    },
    {
      "epoch": 0.43100189035916825,
      "grad_norm": 0.40087552967732526,
      "learning_rate": 9.764059219677517e-05,
      "loss": 0.658582329750061,
      "step": 456
    },
    {
      "epoch": 0.43194706994328924,
      "grad_norm": 0.3581193161900853,
      "learning_rate": 9.762475097039767e-05,
      "loss": 0.6885819435119629,
      "step": 457
    },
    {
      "epoch": 0.43289224952741023,
      "grad_norm": 0.3154059673775601,
      "learning_rate": 9.760885803601411e-05,
      "loss": 0.505074143409729,
      "step": 458
    },
    {
      "epoch": 0.43383742911153117,
      "grad_norm": 0.38050355419121673,
      "learning_rate": 9.759291341088009e-05,
      "loss": 0.5545379519462585,
      "step": 459
    },
    {
      "epoch": 0.43478260869565216,
      "grad_norm": 0.36182023562569504,
      "learning_rate": 9.757691711230727e-05,
      "loss": 0.6113598346710205,
      "step": 460
    },
    {
      "epoch": 0.43572778827977315,
      "grad_norm": 0.3649984146520518,
      "learning_rate": 9.756086915766345e-05,
      "loss": 0.5881245136260986,
      "step": 461
    },
    {
      "epoch": 0.43667296786389415,
      "grad_norm": 0.3885580186683344,
      "learning_rate": 9.754476956437251e-05,
      "loss": 0.6278244256973267,
      "step": 462
    },
    {
      "epoch": 0.43761814744801514,
      "grad_norm": 0.35733859006734925,
      "learning_rate": 9.752861834991435e-05,
      "loss": 0.5627791881561279,
      "step": 463
    },
    {
      "epoch": 0.43856332703213613,
      "grad_norm": 0.3479846546198601,
      "learning_rate": 9.751241553182502e-05,
      "loss": 0.5778782963752747,
      "step": 464
    },
    {
      "epoch": 0.43950850661625707,
      "grad_norm": 0.33583331138555983,
      "learning_rate": 9.74961611276965e-05,
      "loss": 0.5732190608978271,
      "step": 465
    },
    {
      "epoch": 0.44045368620037806,
      "grad_norm": 0.358209332175393,
      "learning_rate": 9.747985515517683e-05,
      "loss": 0.6072621941566467,
      "step": 466
    },
    {
      "epoch": 0.44139886578449905,
      "grad_norm": 0.3522183210448868,
      "learning_rate": 9.746349763197003e-05,
      "loss": 0.5420138835906982,
      "step": 467
    },
    {
      "epoch": 0.44234404536862004,
      "grad_norm": 0.3751596345414983,
      "learning_rate": 9.744708857583607e-05,
      "loss": 0.48897841572761536,
      "step": 468
    },
    {
      "epoch": 0.44328922495274103,
      "grad_norm": 0.38950214039849407,
      "learning_rate": 9.743062800459088e-05,
      "loss": 0.5895791053771973,
      "step": 469
    },
    {
      "epoch": 0.444234404536862,
      "grad_norm": 0.29968939501611547,
      "learning_rate": 9.741411593610635e-05,
      "loss": 0.5509815812110901,
      "step": 470
    },
    {
      "epoch": 0.44517958412098296,
      "grad_norm": 0.3712287103033739,
      "learning_rate": 9.739755238831027e-05,
      "loss": 0.6269895434379578,
      "step": 471
    },
    {
      "epoch": 0.44612476370510395,
      "grad_norm": 0.36877708372992385,
      "learning_rate": 9.738093737918628e-05,
      "loss": 0.5657069087028503,
      "step": 472
    },
    {
      "epoch": 0.44706994328922495,
      "grad_norm": 0.3275272959192288,
      "learning_rate": 9.736427092677396e-05,
      "loss": 0.5635449290275574,
      "step": 473
    },
    {
      "epoch": 0.44801512287334594,
      "grad_norm": 0.3259321350161926,
      "learning_rate": 9.734755304916868e-05,
      "loss": 0.5164295434951782,
      "step": 474
    },
    {
      "epoch": 0.44896030245746693,
      "grad_norm": 0.3339575279175625,
      "learning_rate": 9.733078376452171e-05,
      "loss": 0.5160853266716003,
      "step": 475
    },
    {
      "epoch": 0.4499054820415879,
      "grad_norm": 0.32320190276058913,
      "learning_rate": 9.731396309104009e-05,
      "loss": 0.5191076993942261,
      "step": 476
    },
    {
      "epoch": 0.45085066162570886,
      "grad_norm": 0.31425773936984747,
      "learning_rate": 9.729709104698665e-05,
      "loss": 0.5844600200653076,
      "step": 477
    },
    {
      "epoch": 0.45179584120982985,
      "grad_norm": 0.35055379744641035,
      "learning_rate": 9.728016765068004e-05,
      "loss": 0.5537489652633667,
      "step": 478
    },
    {
      "epoch": 0.45274102079395084,
      "grad_norm": 0.33041862784122755,
      "learning_rate": 9.726319292049462e-05,
      "loss": 0.5904551148414612,
      "step": 479
    },
    {
      "epoch": 0.45368620037807184,
      "grad_norm": 0.34447081525758505,
      "learning_rate": 9.724616687486048e-05,
      "loss": 0.574586808681488,
      "step": 480
    },
    {
      "epoch": 0.4546313799621928,
      "grad_norm": 0.3786265574961823,
      "learning_rate": 9.722908953226348e-05,
      "loss": 0.5798699855804443,
      "step": 481
    },
    {
      "epoch": 0.4555765595463138,
      "grad_norm": 0.37518153416430694,
      "learning_rate": 9.721196091124512e-05,
      "loss": 0.5909737944602966,
      "step": 482
    },
    {
      "epoch": 0.45652173913043476,
      "grad_norm": 0.3308899892080598,
      "learning_rate": 9.719478103040263e-05,
      "loss": 0.5291968584060669,
      "step": 483
    },
    {
      "epoch": 0.45746691871455575,
      "grad_norm": 0.3435471107709723,
      "learning_rate": 9.717754990838881e-05,
      "loss": 0.574386477470398,
      "step": 484
    },
    {
      "epoch": 0.45841209829867674,
      "grad_norm": 0.31808115483827876,
      "learning_rate": 9.716026756391217e-05,
      "loss": 0.5616765022277832,
      "step": 485
    },
    {
      "epoch": 0.45935727788279773,
      "grad_norm": 0.3148705055448909,
      "learning_rate": 9.714293401573682e-05,
      "loss": 0.4873449504375458,
      "step": 486
    },
    {
      "epoch": 0.4603024574669187,
      "grad_norm": 0.3911068695814253,
      "learning_rate": 9.712554928268245e-05,
      "loss": 0.6485486030578613,
      "step": 487
    },
    {
      "epoch": 0.4612476370510397,
      "grad_norm": 0.3305739241664989,
      "learning_rate": 9.710811338362431e-05,
      "loss": 0.5135831236839294,
      "step": 488
    },
    {
      "epoch": 0.4621928166351607,
      "grad_norm": 0.36555326690847567,
      "learning_rate": 9.709062633749324e-05,
      "loss": 0.6002957820892334,
      "step": 489
    },
    {
      "epoch": 0.46313799621928164,
      "grad_norm": 0.3734879598464776,
      "learning_rate": 9.707308816327557e-05,
      "loss": 0.5919216871261597,
      "step": 490
    },
    {
      "epoch": 0.46408317580340264,
      "grad_norm": 0.3686967636871946,
      "learning_rate": 9.705549888001319e-05,
      "loss": 0.5563586950302124,
      "step": 491
    },
    {
      "epoch": 0.46502835538752363,
      "grad_norm": 0.35020570342525614,
      "learning_rate": 9.703785850680342e-05,
      "loss": 0.6002446413040161,
      "step": 492
    },
    {
      "epoch": 0.4659735349716446,
      "grad_norm": 0.3585378408506028,
      "learning_rate": 9.702016706279913e-05,
      "loss": 0.6708824634552002,
      "step": 493
    },
    {
      "epoch": 0.4669187145557656,
      "grad_norm": 0.34382091476098525,
      "learning_rate": 9.700242456720856e-05,
      "loss": 0.5627458095550537,
      "step": 494
    },
    {
      "epoch": 0.4678638941398866,
      "grad_norm": 0.3530877370584276,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.6047898530960083,
      "step": 495
    },
    {
      "epoch": 0.46880907372400754,
      "grad_norm": 0.3289285125823899,
      "learning_rate": 9.696678649837883e-05,
      "loss": 0.6046621799468994,
      "step": 496
    },
    {
      "epoch": 0.46975425330812853,
      "grad_norm": 0.3340383429752398,
      "learning_rate": 9.694889096383328e-05,
      "loss": 0.6380482912063599,
      "step": 497
    },
    {
      "epoch": 0.4706994328922495,
      "grad_norm": 0.31904368543419637,
      "learning_rate": 9.693094445508865e-05,
      "loss": 0.5593453645706177,
      "step": 498
    },
    {
      "epoch": 0.4716446124763705,
      "grad_norm": 0.3787859066347041,
      "learning_rate": 9.691294699163012e-05,
      "loss": 0.6537929773330688,
      "step": 499
    },
    {
      "epoch": 0.4725897920604915,
      "grad_norm": 0.3525624546096739,
      "learning_rate": 9.689489859299823e-05,
      "loss": 0.6304800510406494,
      "step": 500
    },
    {
      "epoch": 0.4735349716446125,
      "grad_norm": 0.31912301298704165,
      "learning_rate": 9.687679927878884e-05,
      "loss": 0.5721961855888367,
      "step": 501
    },
    {
      "epoch": 0.47448015122873344,
      "grad_norm": 0.3796163512311206,
      "learning_rate": 9.685864906865303e-05,
      "loss": 0.7018606662750244,
      "step": 502
    },
    {
      "epoch": 0.47542533081285443,
      "grad_norm": 0.3456657824998754,
      "learning_rate": 9.684044798229719e-05,
      "loss": 0.6203368306159973,
      "step": 503
    },
    {
      "epoch": 0.4763705103969754,
      "grad_norm": 0.3626858035594562,
      "learning_rate": 9.682219603948292e-05,
      "loss": 0.56705641746521,
      "step": 504
    },
    {
      "epoch": 0.4773156899810964,
      "grad_norm": 0.339627648369534,
      "learning_rate": 9.680389326002708e-05,
      "loss": 0.6183629631996155,
      "step": 505
    },
    {
      "epoch": 0.4782608695652174,
      "grad_norm": 0.3258951658321143,
      "learning_rate": 9.678553966380167e-05,
      "loss": 0.5586029291152954,
      "step": 506
    },
    {
      "epoch": 0.4792060491493384,
      "grad_norm": 0.34377669527721555,
      "learning_rate": 9.67671352707339e-05,
      "loss": 0.5724195241928101,
      "step": 507
    },
    {
      "epoch": 0.48015122873345933,
      "grad_norm": 0.32114209856999104,
      "learning_rate": 9.67486801008061e-05,
      "loss": 0.5595238208770752,
      "step": 508
    },
    {
      "epoch": 0.4810964083175803,
      "grad_norm": 0.3376553205808138,
      "learning_rate": 9.673017417405577e-05,
      "loss": 0.5395722389221191,
      "step": 509
    },
    {
      "epoch": 0.4820415879017013,
      "grad_norm": 0.3380819154296847,
      "learning_rate": 9.671161751057551e-05,
      "loss": 0.6015375852584839,
      "step": 510
    },
    {
      "epoch": 0.4829867674858223,
      "grad_norm": 0.3772083309084413,
      "learning_rate": 9.669301013051297e-05,
      "loss": 0.6917191743850708,
      "step": 511
    },
    {
      "epoch": 0.4839319470699433,
      "grad_norm": 0.3343569926790589,
      "learning_rate": 9.66743520540709e-05,
      "loss": 0.5636065006256104,
      "step": 512
    },
    {
      "epoch": 0.4848771266540643,
      "grad_norm": 0.35375609742179137,
      "learning_rate": 9.665564330150709e-05,
      "loss": 0.48130911588668823,
      "step": 513
    },
    {
      "epoch": 0.48582230623818523,
      "grad_norm": 0.3547983640005102,
      "learning_rate": 9.663688389313434e-05,
      "loss": 0.5427068471908569,
      "step": 514
    },
    {
      "epoch": 0.4867674858223062,
      "grad_norm": 0.34806596558132885,
      "learning_rate": 9.661807384932047e-05,
      "loss": 0.5928956270217896,
      "step": 515
    },
    {
      "epoch": 0.4877126654064272,
      "grad_norm": 0.3350525145658891,
      "learning_rate": 9.659921319048824e-05,
      "loss": 0.49638015031814575,
      "step": 516
    },
    {
      "epoch": 0.4886578449905482,
      "grad_norm": 0.35524612811158646,
      "learning_rate": 9.658030193711538e-05,
      "loss": 0.6205455660820007,
      "step": 517
    },
    {
      "epoch": 0.4896030245746692,
      "grad_norm": 0.3590136432080948,
      "learning_rate": 9.656134010973459e-05,
      "loss": 0.5530794858932495,
      "step": 518
    },
    {
      "epoch": 0.4905482041587902,
      "grad_norm": 0.3209917623141349,
      "learning_rate": 9.654232772893341e-05,
      "loss": 0.5171040296554565,
      "step": 519
    },
    {
      "epoch": 0.4914933837429111,
      "grad_norm": 0.40880763134710085,
      "learning_rate": 9.652326481535435e-05,
      "loss": 0.7739261984825134,
      "step": 520
    },
    {
      "epoch": 0.4924385633270321,
      "grad_norm": 0.36135692644000694,
      "learning_rate": 9.650415138969468e-05,
      "loss": 0.6454090476036072,
      "step": 521
    },
    {
      "epoch": 0.4933837429111531,
      "grad_norm": 0.40391924894371,
      "learning_rate": 9.648498747270665e-05,
      "loss": 0.6356430053710938,
      "step": 522
    },
    {
      "epoch": 0.4943289224952741,
      "grad_norm": 0.3492751287358038,
      "learning_rate": 9.646577308519718e-05,
      "loss": 0.5559333562850952,
      "step": 523
    },
    {
      "epoch": 0.4952741020793951,
      "grad_norm": 0.3694064927375815,
      "learning_rate": 9.644650824802813e-05,
      "loss": 0.6886277198791504,
      "step": 524
    },
    {
      "epoch": 0.4962192816635161,
      "grad_norm": 0.3563589004331441,
      "learning_rate": 9.642719298211602e-05,
      "loss": 0.6175848245620728,
      "step": 525
    },
    {
      "epoch": 0.497164461247637,
      "grad_norm": 0.3783214217213387,
      "learning_rate": 9.640782730843218e-05,
      "loss": 0.7010941505432129,
      "step": 526
    },
    {
      "epoch": 0.498109640831758,
      "grad_norm": 0.3364809483871497,
      "learning_rate": 9.638841124800268e-05,
      "loss": 0.6438390016555786,
      "step": 527
    },
    {
      "epoch": 0.499054820415879,
      "grad_norm": 0.31374612552907843,
      "learning_rate": 9.636894482190823e-05,
      "loss": 0.517719566822052,
      "step": 528
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3867210710572921,
      "learning_rate": 9.634942805128433e-05,
      "loss": 0.7359694242477417,
      "step": 529
    },
    {
      "epoch": 0.500945179584121,
      "grad_norm": 0.3243293661644752,
      "learning_rate": 9.632986095732107e-05,
      "loss": 0.5258830785751343,
      "step": 530
    },
    {
      "epoch": 0.501890359168242,
      "grad_norm": 0.3098103599748442,
      "learning_rate": 9.631024356126312e-05,
      "loss": 0.49457982182502747,
      "step": 531
    },
    {
      "epoch": 0.502835538752363,
      "grad_norm": 0.28932727673173547,
      "learning_rate": 9.629057588440993e-05,
      "loss": 0.4880021810531616,
      "step": 532
    },
    {
      "epoch": 0.503780718336484,
      "grad_norm": 0.35803461540095116,
      "learning_rate": 9.627085794811538e-05,
      "loss": 0.6137503385543823,
      "step": 533
    },
    {
      "epoch": 0.504725897920605,
      "grad_norm": 0.42431621449815615,
      "learning_rate": 9.625108977378801e-05,
      "loss": 0.653312087059021,
      "step": 534
    },
    {
      "epoch": 0.505671077504726,
      "grad_norm": 0.4290183213498434,
      "learning_rate": 9.623127138289087e-05,
      "loss": 0.6189536452293396,
      "step": 535
    },
    {
      "epoch": 0.5066162570888468,
      "grad_norm": 0.33795152579613597,
      "learning_rate": 9.621140279694156e-05,
      "loss": 0.5407998561859131,
      "step": 536
    },
    {
      "epoch": 0.5075614366729678,
      "grad_norm": 0.35456472325364025,
      "learning_rate": 9.619148403751214e-05,
      "loss": 0.6311345100402832,
      "step": 537
    },
    {
      "epoch": 0.5085066162570888,
      "grad_norm": 0.400320862569392,
      "learning_rate": 9.617151512622917e-05,
      "loss": 0.7493088245391846,
      "step": 538
    },
    {
      "epoch": 0.5094517958412098,
      "grad_norm": 0.36999215317321393,
      "learning_rate": 9.615149608477368e-05,
      "loss": 0.6966419219970703,
      "step": 539
    },
    {
      "epoch": 0.5103969754253308,
      "grad_norm": 0.31908068133642614,
      "learning_rate": 9.613142693488106e-05,
      "loss": 0.5421736240386963,
      "step": 540
    },
    {
      "epoch": 0.5113421550094518,
      "grad_norm": 0.37747440494615603,
      "learning_rate": 9.61113076983412e-05,
      "loss": 0.686873197555542,
      "step": 541
    },
    {
      "epoch": 0.5122873345935728,
      "grad_norm": 0.40096059953256225,
      "learning_rate": 9.609113839699829e-05,
      "loss": 0.6354721188545227,
      "step": 542
    },
    {
      "epoch": 0.5132325141776938,
      "grad_norm": 0.3648136106598335,
      "learning_rate": 9.607091905275092e-05,
      "loss": 0.6070059537887573,
      "step": 543
    },
    {
      "epoch": 0.5141776937618148,
      "grad_norm": 0.3005709493583339,
      "learning_rate": 9.6050649687552e-05,
      "loss": 0.4656637907028198,
      "step": 544
    },
    {
      "epoch": 0.5151228733459358,
      "grad_norm": 0.32411923106244955,
      "learning_rate": 9.603033032340875e-05,
      "loss": 0.5220184326171875,
      "step": 545
    },
    {
      "epoch": 0.5160680529300568,
      "grad_norm": 0.514046104942516,
      "learning_rate": 9.600996098238267e-05,
      "loss": 0.5665009021759033,
      "step": 546
    },
    {
      "epoch": 0.5170132325141777,
      "grad_norm": 0.38812835947195495,
      "learning_rate": 9.598954168658955e-05,
      "loss": 0.6098254919052124,
      "step": 547
    },
    {
      "epoch": 0.5179584120982986,
      "grad_norm": 0.34068565873016377,
      "learning_rate": 9.59690724581994e-05,
      "loss": 0.650014340877533,
      "step": 548
    },
    {
      "epoch": 0.5189035916824196,
      "grad_norm": 0.35747211071697316,
      "learning_rate": 9.59485533194364e-05,
      "loss": 0.6101299524307251,
      "step": 549
    },
    {
      "epoch": 0.5198487712665406,
      "grad_norm": 0.35602586029998406,
      "learning_rate": 9.5927984292579e-05,
      "loss": 0.6005699038505554,
      "step": 550
    },
    {
      "epoch": 0.5207939508506616,
      "grad_norm": 0.3597788275672169,
      "learning_rate": 9.590736539995975e-05,
      "loss": 0.5947712659835815,
      "step": 551
    },
    {
      "epoch": 0.5217391304347826,
      "grad_norm": 0.3569216685351984,
      "learning_rate": 9.588669666396539e-05,
      "loss": 0.6251085996627808,
      "step": 552
    },
    {
      "epoch": 0.5226843100189036,
      "grad_norm": 0.36821437766234677,
      "learning_rate": 9.586597810703675e-05,
      "loss": 0.6210182905197144,
      "step": 553
    },
    {
      "epoch": 0.5236294896030246,
      "grad_norm": 0.3531042948769986,
      "learning_rate": 9.584520975166875e-05,
      "loss": 0.6443819999694824,
      "step": 554
    },
    {
      "epoch": 0.5245746691871456,
      "grad_norm": 0.33016411944232027,
      "learning_rate": 9.582439162041037e-05,
      "loss": 0.5661275386810303,
      "step": 555
    },
    {
      "epoch": 0.5255198487712666,
      "grad_norm": 0.3361702479245698,
      "learning_rate": 9.580352373586467e-05,
      "loss": 0.6192845702171326,
      "step": 556
    },
    {
      "epoch": 0.5264650283553876,
      "grad_norm": 0.34181660580505413,
      "learning_rate": 9.578260612068871e-05,
      "loss": 0.6251072883605957,
      "step": 557
    },
    {
      "epoch": 0.5274102079395085,
      "grad_norm": 0.31595612570182763,
      "learning_rate": 9.576163879759352e-05,
      "loss": 0.491088330745697,
      "step": 558
    },
    {
      "epoch": 0.5283553875236295,
      "grad_norm": 0.37757369001627694,
      "learning_rate": 9.574062178934415e-05,
      "loss": 0.6067293882369995,
      "step": 559
    },
    {
      "epoch": 0.5293005671077504,
      "grad_norm": 0.33090158986313667,
      "learning_rate": 9.571955511875954e-05,
      "loss": 0.5791563987731934,
      "step": 560
    },
    {
      "epoch": 0.5302457466918714,
      "grad_norm": 0.38325379763202677,
      "learning_rate": 9.569843880871257e-05,
      "loss": 0.5640578866004944,
      "step": 561
    },
    {
      "epoch": 0.5311909262759924,
      "grad_norm": 0.35560900400058576,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.5823667049407959,
      "step": 562
    },
    {
      "epoch": 0.5321361058601134,
      "grad_norm": 0.3730719508603438,
      "learning_rate": 9.565605736199262e-05,
      "loss": 0.6011694073677063,
      "step": 563
    },
    {
      "epoch": 0.5330812854442344,
      "grad_norm": 0.35287462313606965,
      "learning_rate": 9.563479227133479e-05,
      "loss": 0.5545784831047058,
      "step": 564
    },
    {
      "epoch": 0.5340264650283554,
      "grad_norm": 0.3444007986289177,
      "learning_rate": 9.561347763324484e-05,
      "loss": 0.542917013168335,
      "step": 565
    },
    {
      "epoch": 0.5349716446124764,
      "grad_norm": 0.38442813504875356,
      "learning_rate": 9.559211347086494e-05,
      "loss": 0.6275271773338318,
      "step": 566
    },
    {
      "epoch": 0.5359168241965974,
      "grad_norm": 0.37323056950687505,
      "learning_rate": 9.557069980739093e-05,
      "loss": 0.5985451936721802,
      "step": 567
    },
    {
      "epoch": 0.5368620037807184,
      "grad_norm": 0.38409460497464154,
      "learning_rate": 9.554923666607244e-05,
      "loss": 0.6015529632568359,
      "step": 568
    },
    {
      "epoch": 0.5378071833648393,
      "grad_norm": 0.344569897655607,
      "learning_rate": 9.552772407021288e-05,
      "loss": 0.5818746089935303,
      "step": 569
    },
    {
      "epoch": 0.5387523629489603,
      "grad_norm": 0.3274116536941209,
      "learning_rate": 9.550616204316922e-05,
      "loss": 0.5112401247024536,
      "step": 570
    },
    {
      "epoch": 0.5396975425330813,
      "grad_norm": 0.3065890488304919,
      "learning_rate": 9.548455060835222e-05,
      "loss": 0.5772692561149597,
      "step": 571
    },
    {
      "epoch": 0.5406427221172023,
      "grad_norm": 0.36108908394498707,
      "learning_rate": 9.546288978922622e-05,
      "loss": 0.5105010271072388,
      "step": 572
    },
    {
      "epoch": 0.5415879017013232,
      "grad_norm": 0.33463896906828033,
      "learning_rate": 9.544117960930919e-05,
      "loss": 0.5190365314483643,
      "step": 573
    },
    {
      "epoch": 0.5425330812854442,
      "grad_norm": 0.3640014681130308,
      "learning_rate": 9.541942009217273e-05,
      "loss": 0.6215477585792542,
      "step": 574
    },
    {
      "epoch": 0.5434782608695652,
      "grad_norm": 0.34097296800392585,
      "learning_rate": 9.539761126144193e-05,
      "loss": 0.5854123830795288,
      "step": 575
    },
    {
      "epoch": 0.5444234404536862,
      "grad_norm": 0.3424813899978309,
      "learning_rate": 9.537575314079552e-05,
      "loss": 0.5901289582252502,
      "step": 576
    },
    {
      "epoch": 0.5453686200378072,
      "grad_norm": 0.33412948987251384,
      "learning_rate": 9.535384575396565e-05,
      "loss": 0.5419120192527771,
      "step": 577
    },
    {
      "epoch": 0.5463137996219282,
      "grad_norm": 0.3340371979307214,
      "learning_rate": 9.533188912473803e-05,
      "loss": 0.5135220289230347,
      "step": 578
    },
    {
      "epoch": 0.5472589792060492,
      "grad_norm": 0.3726118418215802,
      "learning_rate": 9.530988327695179e-05,
      "loss": 0.5953744649887085,
      "step": 579
    },
    {
      "epoch": 0.5482041587901701,
      "grad_norm": 0.3547713142003794,
      "learning_rate": 9.528782823449954e-05,
      "loss": 0.6391772031784058,
      "step": 580
    },
    {
      "epoch": 0.5491493383742911,
      "grad_norm": 0.3588119295619241,
      "learning_rate": 9.526572402132726e-05,
      "loss": 0.666323184967041,
      "step": 581
    },
    {
      "epoch": 0.5500945179584121,
      "grad_norm": 0.28457982218417427,
      "learning_rate": 9.524357066143435e-05,
      "loss": 0.4853088855743408,
      "step": 582
    },
    {
      "epoch": 0.5510396975425331,
      "grad_norm": 0.35470178622346293,
      "learning_rate": 9.522136817887353e-05,
      "loss": 0.5992828607559204,
      "step": 583
    },
    {
      "epoch": 0.5519848771266541,
      "grad_norm": 0.3555512789142821,
      "learning_rate": 9.519911659775091e-05,
      "loss": 0.5438647270202637,
      "step": 584
    },
    {
      "epoch": 0.552930056710775,
      "grad_norm": 0.3809688210606957,
      "learning_rate": 9.517681594222589e-05,
      "loss": 0.6907604336738586,
      "step": 585
    },
    {
      "epoch": 0.553875236294896,
      "grad_norm": 0.3400734609307528,
      "learning_rate": 9.51544662365111e-05,
      "loss": 0.6333542466163635,
      "step": 586
    },
    {
      "epoch": 0.554820415879017,
      "grad_norm": 0.3295436426769136,
      "learning_rate": 9.51320675048725e-05,
      "loss": 0.5705955028533936,
      "step": 587
    },
    {
      "epoch": 0.555765595463138,
      "grad_norm": 0.3637305036688789,
      "learning_rate": 9.510961977162923e-05,
      "loss": 0.5501980781555176,
      "step": 588
    },
    {
      "epoch": 0.556710775047259,
      "grad_norm": 0.4176582775238026,
      "learning_rate": 9.508712306115365e-05,
      "loss": 0.6093301773071289,
      "step": 589
    },
    {
      "epoch": 0.55765595463138,
      "grad_norm": 0.29732700807325046,
      "learning_rate": 9.506457739787132e-05,
      "loss": 0.4791608452796936,
      "step": 590
    },
    {
      "epoch": 0.558601134215501,
      "grad_norm": 0.38522443337470635,
      "learning_rate": 9.504198280626087e-05,
      "loss": 0.7227703332901001,
      "step": 591
    },
    {
      "epoch": 0.5595463137996219,
      "grad_norm": 0.37280169108510036,
      "learning_rate": 9.501933931085416e-05,
      "loss": 0.5999171733856201,
      "step": 592
    },
    {
      "epoch": 0.5604914933837429,
      "grad_norm": 0.46254628018603305,
      "learning_rate": 9.499664693623608e-05,
      "loss": 0.5809277296066284,
      "step": 593
    },
    {
      "epoch": 0.5614366729678639,
      "grad_norm": 0.3319678882418436,
      "learning_rate": 9.49739057070446e-05,
      "loss": 0.5718780755996704,
      "step": 594
    },
    {
      "epoch": 0.5623818525519849,
      "grad_norm": 0.32406746236905465,
      "learning_rate": 9.495111564797074e-05,
      "loss": 0.5858637094497681,
      "step": 595
    },
    {
      "epoch": 0.5633270321361059,
      "grad_norm": 0.38554246205343085,
      "learning_rate": 9.492827678375853e-05,
      "loss": 0.6227236986160278,
      "step": 596
    },
    {
      "epoch": 0.5642722117202268,
      "grad_norm": 0.399172099373056,
      "learning_rate": 9.490538913920502e-05,
      "loss": 0.5894288420677185,
      "step": 597
    },
    {
      "epoch": 0.5652173913043478,
      "grad_norm": 0.3487567207322285,
      "learning_rate": 9.488245273916017e-05,
      "loss": 0.5886117219924927,
      "step": 598
    },
    {
      "epoch": 0.5661625708884688,
      "grad_norm": 0.30908222453048345,
      "learning_rate": 9.485946760852691e-05,
      "loss": 0.5927820205688477,
      "step": 599
    },
    {
      "epoch": 0.5671077504725898,
      "grad_norm": 0.38012522500065316,
      "learning_rate": 9.483643377226107e-05,
      "loss": 0.5610758066177368,
      "step": 600
    },
    {
      "epoch": 0.5680529300567108,
      "grad_norm": 0.40027297740477596,
      "learning_rate": 9.481335125537138e-05,
      "loss": 0.6011504530906677,
      "step": 601
    },
    {
      "epoch": 0.5689981096408318,
      "grad_norm": 0.3274691684542873,
      "learning_rate": 9.479022008291938e-05,
      "loss": 0.5774383544921875,
      "step": 602
    },
    {
      "epoch": 0.5699432892249527,
      "grad_norm": 0.34043879710802294,
      "learning_rate": 9.47670402800195e-05,
      "loss": 0.528027355670929,
      "step": 603
    },
    {
      "epoch": 0.5708884688090737,
      "grad_norm": 0.38119791648362067,
      "learning_rate": 9.474381187183889e-05,
      "loss": 0.6564128994941711,
      "step": 604
    },
    {
      "epoch": 0.5718336483931947,
      "grad_norm": 0.35184191348707256,
      "learning_rate": 9.472053488359757e-05,
      "loss": 0.593501091003418,
      "step": 605
    },
    {
      "epoch": 0.5727788279773157,
      "grad_norm": 0.3554075029491632,
      "learning_rate": 9.46972093405682e-05,
      "loss": 0.58607417345047,
      "step": 606
    },
    {
      "epoch": 0.5737240075614367,
      "grad_norm": 0.384298254049937,
      "learning_rate": 9.467383526807626e-05,
      "loss": 0.6306958794593811,
      "step": 607
    },
    {
      "epoch": 0.5746691871455577,
      "grad_norm": 0.38382868932958325,
      "learning_rate": 9.465041269149984e-05,
      "loss": 0.5934082269668579,
      "step": 608
    },
    {
      "epoch": 0.5756143667296786,
      "grad_norm": 0.37560099697308774,
      "learning_rate": 9.462694163626972e-05,
      "loss": 0.6247125267982483,
      "step": 609
    },
    {
      "epoch": 0.5765595463137996,
      "grad_norm": 0.33429640387664883,
      "learning_rate": 9.460342212786932e-05,
      "loss": 0.623967170715332,
      "step": 610
    },
    {
      "epoch": 0.5775047258979206,
      "grad_norm": 0.34131800524957295,
      "learning_rate": 9.45798541918347e-05,
      "loss": 0.604785680770874,
      "step": 611
    },
    {
      "epoch": 0.5784499054820416,
      "grad_norm": 0.31949577107209387,
      "learning_rate": 9.455623785375443e-05,
      "loss": 0.5873017311096191,
      "step": 612
    },
    {
      "epoch": 0.5793950850661626,
      "grad_norm": 0.3932197687952289,
      "learning_rate": 9.453257313926968e-05,
      "loss": 0.6980948448181152,
      "step": 613
    },
    {
      "epoch": 0.5803402646502835,
      "grad_norm": 0.3283551144313492,
      "learning_rate": 9.450886007407412e-05,
      "loss": 0.5343049764633179,
      "step": 614
    },
    {
      "epoch": 0.5812854442344045,
      "grad_norm": 0.3246780877730677,
      "learning_rate": 9.448509868391395e-05,
      "loss": 0.6023126840591431,
      "step": 615
    },
    {
      "epoch": 0.5822306238185255,
      "grad_norm": 0.3352909584888492,
      "learning_rate": 9.44612889945878e-05,
      "loss": 0.6279209852218628,
      "step": 616
    },
    {
      "epoch": 0.5831758034026465,
      "grad_norm": 0.33947011528406584,
      "learning_rate": 9.443743103194675e-05,
      "loss": 0.5828680992126465,
      "step": 617
    },
    {
      "epoch": 0.5841209829867675,
      "grad_norm": 0.40071256645141534,
      "learning_rate": 9.441352482189434e-05,
      "loss": 0.6106035709381104,
      "step": 618
    },
    {
      "epoch": 0.5850661625708885,
      "grad_norm": 0.352047939829225,
      "learning_rate": 9.43895703903864e-05,
      "loss": 0.6217541098594666,
      "step": 619
    },
    {
      "epoch": 0.5860113421550095,
      "grad_norm": 0.5308321770692396,
      "learning_rate": 9.43655677634312e-05,
      "loss": 0.6909875869750977,
      "step": 620
    },
    {
      "epoch": 0.5869565217391305,
      "grad_norm": 0.354062945234155,
      "learning_rate": 9.434151696708928e-05,
      "loss": 0.6033484935760498,
      "step": 621
    },
    {
      "epoch": 0.5879017013232514,
      "grad_norm": 0.3503745019801434,
      "learning_rate": 9.431741802747354e-05,
      "loss": 0.5865163803100586,
      "step": 622
    },
    {
      "epoch": 0.5888468809073724,
      "grad_norm": 0.35343873022763805,
      "learning_rate": 9.42932709707491e-05,
      "loss": 0.620151937007904,
      "step": 623
    },
    {
      "epoch": 0.5897920604914934,
      "grad_norm": 0.3365388965979525,
      "learning_rate": 9.426907582313334e-05,
      "loss": 0.5141041874885559,
      "step": 624
    },
    {
      "epoch": 0.5907372400756143,
      "grad_norm": 0.34921608398392756,
      "learning_rate": 9.424483261089584e-05,
      "loss": 0.5464718341827393,
      "step": 625
    },
    {
      "epoch": 0.5916824196597353,
      "grad_norm": 0.3323305575846991,
      "learning_rate": 9.42205413603584e-05,
      "loss": 0.6408066749572754,
      "step": 626
    },
    {
      "epoch": 0.5926275992438563,
      "grad_norm": 0.3737561928519843,
      "learning_rate": 9.419620209789494e-05,
      "loss": 0.6575788259506226,
      "step": 627
    },
    {
      "epoch": 0.5935727788279773,
      "grad_norm": 0.3440687338557296,
      "learning_rate": 9.417181484993154e-05,
      "loss": 0.6014034152030945,
      "step": 628
    },
    {
      "epoch": 0.5945179584120983,
      "grad_norm": 0.35377703463928206,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.6205920577049255,
      "step": 629
    },
    {
      "epoch": 0.5954631379962193,
      "grad_norm": 0.3763141062842993,
      "learning_rate": 9.412289650346961e-05,
      "loss": 0.556941568851471,
      "step": 630
    },
    {
      "epoch": 0.5964083175803403,
      "grad_norm": 0.341270162384919,
      "learning_rate": 9.40983654580836e-05,
      "loss": 0.54781574010849,
      "step": 631
    },
    {
      "epoch": 0.5973534971644613,
      "grad_norm": 0.3184446699457364,
      "learning_rate": 9.407378653342261e-05,
      "loss": 0.6188017129898071,
      "step": 632
    },
    {
      "epoch": 0.5982986767485823,
      "grad_norm": 0.3660836251714076,
      "learning_rate": 9.404915975617291e-05,
      "loss": 0.608108639717102,
      "step": 633
    },
    {
      "epoch": 0.5992438563327032,
      "grad_norm": 0.31815802371983315,
      "learning_rate": 9.402448515307274e-05,
      "loss": 0.5093142986297607,
      "step": 634
    },
    {
      "epoch": 0.6001890359168242,
      "grad_norm": 0.32179120051140386,
      "learning_rate": 9.399976275091223e-05,
      "loss": 0.626184344291687,
      "step": 635
    },
    {
      "epoch": 0.6011342155009451,
      "grad_norm": 0.3878103072026682,
      "learning_rate": 9.397499257653346e-05,
      "loss": 0.6551644206047058,
      "step": 636
    },
    {
      "epoch": 0.6020793950850661,
      "grad_norm": 0.3671347423485299,
      "learning_rate": 9.395017465683036e-05,
      "loss": 0.5784835815429688,
      "step": 637
    },
    {
      "epoch": 0.6030245746691871,
      "grad_norm": 0.3267006048343398,
      "learning_rate": 9.392530901874865e-05,
      "loss": 0.6027266383171082,
      "step": 638
    },
    {
      "epoch": 0.6039697542533081,
      "grad_norm": 0.32816328498876846,
      "learning_rate": 9.390039568928595e-05,
      "loss": 0.4921568036079407,
      "step": 639
    },
    {
      "epoch": 0.6049149338374291,
      "grad_norm": 0.334465709271374,
      "learning_rate": 9.387543469549156e-05,
      "loss": 0.5156058073043823,
      "step": 640
    },
    {
      "epoch": 0.6058601134215501,
      "grad_norm": 0.3478233988883086,
      "learning_rate": 9.385042606446663e-05,
      "loss": 0.6492546796798706,
      "step": 641
    },
    {
      "epoch": 0.6068052930056711,
      "grad_norm": 0.3588846769541788,
      "learning_rate": 9.382536982336395e-05,
      "loss": 0.6217995882034302,
      "step": 642
    },
    {
      "epoch": 0.6077504725897921,
      "grad_norm": 0.3407179554331574,
      "learning_rate": 9.380026599938805e-05,
      "loss": 0.5474449992179871,
      "step": 643
    },
    {
      "epoch": 0.6086956521739131,
      "grad_norm": 0.3584260042191281,
      "learning_rate": 9.377511461979509e-05,
      "loss": 0.5686377286911011,
      "step": 644
    },
    {
      "epoch": 0.6096408317580341,
      "grad_norm": 0.36643666450603224,
      "learning_rate": 9.374991571189291e-05,
      "loss": 0.6288442015647888,
      "step": 645
    },
    {
      "epoch": 0.610586011342155,
      "grad_norm": 0.30765758561928624,
      "learning_rate": 9.372466930304091e-05,
      "loss": 0.498823881149292,
      "step": 646
    },
    {
      "epoch": 0.611531190926276,
      "grad_norm": 0.3590658351304313,
      "learning_rate": 9.369937542065006e-05,
      "loss": 0.595363974571228,
      "step": 647
    },
    {
      "epoch": 0.6124763705103969,
      "grad_norm": 0.32730052814575067,
      "learning_rate": 9.367403409218292e-05,
      "loss": 0.5174237489700317,
      "step": 648
    },
    {
      "epoch": 0.6134215500945179,
      "grad_norm": 0.3646424480934091,
      "learning_rate": 9.364864534515352e-05,
      "loss": 0.6761890649795532,
      "step": 649
    },
    {
      "epoch": 0.6143667296786389,
      "grad_norm": 0.34811150726445345,
      "learning_rate": 9.362320920712739e-05,
      "loss": 0.5193027257919312,
      "step": 650
    },
    {
      "epoch": 0.6153119092627599,
      "grad_norm": 0.3565546928560269,
      "learning_rate": 9.359772570572154e-05,
      "loss": 0.7110329270362854,
      "step": 651
    },
    {
      "epoch": 0.6162570888468809,
      "grad_norm": 0.3081808950435204,
      "learning_rate": 9.357219486860433e-05,
      "loss": 0.5110676884651184,
      "step": 652
    },
    {
      "epoch": 0.6172022684310019,
      "grad_norm": 0.3431486153439944,
      "learning_rate": 9.354661672349558e-05,
      "loss": 0.5645447969436646,
      "step": 653
    },
    {
      "epoch": 0.6181474480151229,
      "grad_norm": 0.34031313111310074,
      "learning_rate": 9.352099129816649e-05,
      "loss": 0.5899103879928589,
      "step": 654
    },
    {
      "epoch": 0.6190926275992439,
      "grad_norm": 0.30597439128799625,
      "learning_rate": 9.349531862043952e-05,
      "loss": 0.4822605550289154,
      "step": 655
    },
    {
      "epoch": 0.6200378071833649,
      "grad_norm": 0.29145961709971296,
      "learning_rate": 9.34695987181885e-05,
      "loss": 0.4319380521774292,
      "step": 656
    },
    {
      "epoch": 0.6209829867674859,
      "grad_norm": 0.35196219765405035,
      "learning_rate": 9.344383161933848e-05,
      "loss": 0.5885263085365295,
      "step": 657
    },
    {
      "epoch": 0.6219281663516069,
      "grad_norm": 0.35395685446237213,
      "learning_rate": 9.34180173518658e-05,
      "loss": 0.579255223274231,
      "step": 658
    },
    {
      "epoch": 0.6228733459357277,
      "grad_norm": 0.3704577808762285,
      "learning_rate": 9.339215594379797e-05,
      "loss": 0.6619397401809692,
      "step": 659
    },
    {
      "epoch": 0.6238185255198487,
      "grad_norm": 0.40016248026523266,
      "learning_rate": 9.336624742321375e-05,
      "loss": 0.6550012826919556,
      "step": 660
    },
    {
      "epoch": 0.6247637051039697,
      "grad_norm": 0.31275560590613,
      "learning_rate": 9.334029181824299e-05,
      "loss": 0.5490162968635559,
      "step": 661
    },
    {
      "epoch": 0.6257088846880907,
      "grad_norm": 0.3459758853400644,
      "learning_rate": 9.331428915706665e-05,
      "loss": 0.5493158102035522,
      "step": 662
    },
    {
      "epoch": 0.6266540642722117,
      "grad_norm": 0.33714641969021375,
      "learning_rate": 9.328823946791686e-05,
      "loss": 0.6405284404754639,
      "step": 663
    },
    {
      "epoch": 0.6275992438563327,
      "grad_norm": 0.36229601521650023,
      "learning_rate": 9.32621427790767e-05,
      "loss": 0.5724446773529053,
      "step": 664
    },
    {
      "epoch": 0.6285444234404537,
      "grad_norm": 0.33840140745283953,
      "learning_rate": 9.323599911888037e-05,
      "loss": 0.5553156137466431,
      "step": 665
    },
    {
      "epoch": 0.6294896030245747,
      "grad_norm": 0.30542465625259885,
      "learning_rate": 9.320980851571306e-05,
      "loss": 0.5569818019866943,
      "step": 666
    },
    {
      "epoch": 0.6304347826086957,
      "grad_norm": 0.35756080246830463,
      "learning_rate": 9.318357099801087e-05,
      "loss": 0.6541997790336609,
      "step": 667
    },
    {
      "epoch": 0.6313799621928167,
      "grad_norm": 0.33065690072587445,
      "learning_rate": 9.315728659426089e-05,
      "loss": 0.5847055912017822,
      "step": 668
    },
    {
      "epoch": 0.6323251417769377,
      "grad_norm": 0.36900862908180365,
      "learning_rate": 9.313095533300109e-05,
      "loss": 0.5488909482955933,
      "step": 669
    },
    {
      "epoch": 0.6332703213610587,
      "grad_norm": 0.3330156837329539,
      "learning_rate": 9.310457724282034e-05,
      "loss": 0.6253513693809509,
      "step": 670
    },
    {
      "epoch": 0.6342155009451795,
      "grad_norm": 0.30610059485334395,
      "learning_rate": 9.30781523523583e-05,
      "loss": 0.5927833914756775,
      "step": 671
    },
    {
      "epoch": 0.6351606805293005,
      "grad_norm": 0.37768246928448473,
      "learning_rate": 9.305168069030553e-05,
      "loss": 0.5785365104675293,
      "step": 672
    },
    {
      "epoch": 0.6361058601134215,
      "grad_norm": 0.32620597322950745,
      "learning_rate": 9.302516228540327e-05,
      "loss": 0.5923622846603394,
      "step": 673
    },
    {
      "epoch": 0.6370510396975425,
      "grad_norm": 0.33375777824892716,
      "learning_rate": 9.29985971664436e-05,
      "loss": 0.6320258975028992,
      "step": 674
    },
    {
      "epoch": 0.6379962192816635,
      "grad_norm": 0.3153654993241819,
      "learning_rate": 9.297198536226928e-05,
      "loss": 0.546387255191803,
      "step": 675
    },
    {
      "epoch": 0.6389413988657845,
      "grad_norm": 0.3059284295281895,
      "learning_rate": 9.294532690177373e-05,
      "loss": 0.5509023666381836,
      "step": 676
    },
    {
      "epoch": 0.6398865784499055,
      "grad_norm": 0.31826804482055737,
      "learning_rate": 9.291862181390108e-05,
      "loss": 0.5870904922485352,
      "step": 677
    },
    {
      "epoch": 0.6408317580340265,
      "grad_norm": 0.3103716382006234,
      "learning_rate": 9.289187012764606e-05,
      "loss": 0.564508318901062,
      "step": 678
    },
    {
      "epoch": 0.6417769376181475,
      "grad_norm": 0.3494324136947132,
      "learning_rate": 9.286507187205399e-05,
      "loss": 0.5810478925704956,
      "step": 679
    },
    {
      "epoch": 0.6427221172022685,
      "grad_norm": 0.31545698732514743,
      "learning_rate": 9.283822707622075e-05,
      "loss": 0.6332958936691284,
      "step": 680
    },
    {
      "epoch": 0.6436672967863895,
      "grad_norm": 0.3136475209395856,
      "learning_rate": 9.281133576929277e-05,
      "loss": 0.5125893950462341,
      "step": 681
    },
    {
      "epoch": 0.6446124763705104,
      "grad_norm": 0.3655513485111703,
      "learning_rate": 9.278439798046697e-05,
      "loss": 0.5988951325416565,
      "step": 682
    },
    {
      "epoch": 0.6455576559546313,
      "grad_norm": 0.3527516003502201,
      "learning_rate": 9.275741373899072e-05,
      "loss": 0.6034035682678223,
      "step": 683
    },
    {
      "epoch": 0.6465028355387523,
      "grad_norm": 0.3162521053455099,
      "learning_rate": 9.273038307416183e-05,
      "loss": 0.5958161950111389,
      "step": 684
    },
    {
      "epoch": 0.6474480151228733,
      "grad_norm": 0.3338054564985125,
      "learning_rate": 9.270330601532855e-05,
      "loss": 0.5647667050361633,
      "step": 685
    },
    {
      "epoch": 0.6483931947069943,
      "grad_norm": 0.31269482191016873,
      "learning_rate": 9.267618259188945e-05,
      "loss": 0.5407567024230957,
      "step": 686
    },
    {
      "epoch": 0.6493383742911153,
      "grad_norm": 0.31865577299938574,
      "learning_rate": 9.264901283329347e-05,
      "loss": 0.5966730117797852,
      "step": 687
    },
    {
      "epoch": 0.6502835538752363,
      "grad_norm": 0.29866720519547724,
      "learning_rate": 9.262179676903986e-05,
      "loss": 0.4618033766746521,
      "step": 688
    },
    {
      "epoch": 0.6512287334593573,
      "grad_norm": 0.3511268156171098,
      "learning_rate": 9.259453442867814e-05,
      "loss": 0.6179370284080505,
      "step": 689
    },
    {
      "epoch": 0.6521739130434783,
      "grad_norm": 0.3604441375722058,
      "learning_rate": 9.256722584180806e-05,
      "loss": 0.5758167505264282,
      "step": 690
    },
    {
      "epoch": 0.6531190926275993,
      "grad_norm": 0.33407582789573675,
      "learning_rate": 9.253987103807958e-05,
      "loss": 0.5970637202262878,
      "step": 691
    },
    {
      "epoch": 0.6540642722117203,
      "grad_norm": 0.4139527526574504,
      "learning_rate": 9.251247004719288e-05,
      "loss": 0.5882635116577148,
      "step": 692
    },
    {
      "epoch": 0.6550094517958412,
      "grad_norm": 0.33460751512074455,
      "learning_rate": 9.248502289889827e-05,
      "loss": 0.6125943660736084,
      "step": 693
    },
    {
      "epoch": 0.6559546313799622,
      "grad_norm": 0.35282775937448424,
      "learning_rate": 9.245752962299612e-05,
      "loss": 0.5914421677589417,
      "step": 694
    },
    {
      "epoch": 0.6568998109640832,
      "grad_norm": 0.3303681611290966,
      "learning_rate": 9.242999024933694e-05,
      "loss": 0.6171237230300903,
      "step": 695
    },
    {
      "epoch": 0.6578449905482041,
      "grad_norm": 0.3270270702430698,
      "learning_rate": 9.24024048078213e-05,
      "loss": 0.5704157948493958,
      "step": 696
    },
    {
      "epoch": 0.6587901701323251,
      "grad_norm": 0.38968197026146667,
      "learning_rate": 9.237477332839975e-05,
      "loss": 0.5840707421302795,
      "step": 697
    },
    {
      "epoch": 0.6597353497164461,
      "grad_norm": 0.3032630351668916,
      "learning_rate": 9.234709584107285e-05,
      "loss": 0.5208728313446045,
      "step": 698
    },
    {
      "epoch": 0.6606805293005671,
      "grad_norm": 0.3632275780978841,
      "learning_rate": 9.231937237589108e-05,
      "loss": 0.660001277923584,
      "step": 699
    },
    {
      "epoch": 0.6616257088846881,
      "grad_norm": 0.33573025972442866,
      "learning_rate": 9.229160296295488e-05,
      "loss": 0.5872658491134644,
      "step": 700
    },
    {
      "epoch": 0.6625708884688091,
      "grad_norm": 0.339524242459544,
      "learning_rate": 9.226378763241454e-05,
      "loss": 0.5647661685943604,
      "step": 701
    },
    {
      "epoch": 0.6635160680529301,
      "grad_norm": 0.33182820777269534,
      "learning_rate": 9.223592641447026e-05,
      "loss": 0.6018141508102417,
      "step": 702
    },
    {
      "epoch": 0.6644612476370511,
      "grad_norm": 0.33836631073667683,
      "learning_rate": 9.220801933937199e-05,
      "loss": 0.5658904910087585,
      "step": 703
    },
    {
      "epoch": 0.665406427221172,
      "grad_norm": 0.3807485723403195,
      "learning_rate": 9.218006643741952e-05,
      "loss": 0.551738977432251,
      "step": 704
    },
    {
      "epoch": 0.666351606805293,
      "grad_norm": 0.3441643094760525,
      "learning_rate": 9.215206773896237e-05,
      "loss": 0.5703752040863037,
      "step": 705
    },
    {
      "epoch": 0.667296786389414,
      "grad_norm": 0.3155968454527295,
      "learning_rate": 9.212402327439982e-05,
      "loss": 0.5828571915626526,
      "step": 706
    },
    {
      "epoch": 0.668241965973535,
      "grad_norm": 0.3161663248809731,
      "learning_rate": 9.209593307418078e-05,
      "loss": 0.5646883845329285,
      "step": 707
    },
    {
      "epoch": 0.6691871455576559,
      "grad_norm": 0.2985242643485724,
      "learning_rate": 9.206779716880387e-05,
      "loss": 0.4955628514289856,
      "step": 708
    },
    {
      "epoch": 0.6701323251417769,
      "grad_norm": 0.3325336960978488,
      "learning_rate": 9.203961558881731e-05,
      "loss": 0.5621753334999084,
      "step": 709
    },
    {
      "epoch": 0.6710775047258979,
      "grad_norm": 0.3722643790685757,
      "learning_rate": 9.201138836481891e-05,
      "loss": 0.6541796922683716,
      "step": 710
    },
    {
      "epoch": 0.6720226843100189,
      "grad_norm": 0.327327775613863,
      "learning_rate": 9.198311552745604e-05,
      "loss": 0.5462191104888916,
      "step": 711
    },
    {
      "epoch": 0.6729678638941399,
      "grad_norm": 0.3351404672135452,
      "learning_rate": 9.19547971074256e-05,
      "loss": 0.6002238392829895,
      "step": 712
    },
    {
      "epoch": 0.6739130434782609,
      "grad_norm": 0.31791295137482817,
      "learning_rate": 9.1926433135474e-05,
      "loss": 0.5498855113983154,
      "step": 713
    },
    {
      "epoch": 0.6748582230623819,
      "grad_norm": 0.3338301194630996,
      "learning_rate": 9.189802364239705e-05,
      "loss": 0.5738036036491394,
      "step": 714
    },
    {
      "epoch": 0.6758034026465028,
      "grad_norm": 0.31209241789921244,
      "learning_rate": 9.186956865904003e-05,
      "loss": 0.6109306216239929,
      "step": 715
    },
    {
      "epoch": 0.6767485822306238,
      "grad_norm": 0.3213073119872088,
      "learning_rate": 9.18410682162976e-05,
      "loss": 0.5274949669837952,
      "step": 716
    },
    {
      "epoch": 0.6776937618147448,
      "grad_norm": 0.4026597482950423,
      "learning_rate": 9.18125223451138e-05,
      "loss": 0.5923879146575928,
      "step": 717
    },
    {
      "epoch": 0.6786389413988658,
      "grad_norm": 0.3304671628090505,
      "learning_rate": 9.178393107648193e-05,
      "loss": 0.5679522752761841,
      "step": 718
    },
    {
      "epoch": 0.6795841209829868,
      "grad_norm": 0.37725285205567016,
      "learning_rate": 9.175529444144463e-05,
      "loss": 0.5787641406059265,
      "step": 719
    },
    {
      "epoch": 0.6805293005671077,
      "grad_norm": 0.36290875810762474,
      "learning_rate": 9.172661247109382e-05,
      "loss": 0.6399126052856445,
      "step": 720
    },
    {
      "epoch": 0.6814744801512287,
      "grad_norm": 0.3411342280528808,
      "learning_rate": 9.169788519657057e-05,
      "loss": 0.593219518661499,
      "step": 721
    },
    {
      "epoch": 0.6824196597353497,
      "grad_norm": 0.32941160924371726,
      "learning_rate": 9.166911264906517e-05,
      "loss": 0.6293588280677795,
      "step": 722
    },
    {
      "epoch": 0.6833648393194707,
      "grad_norm": 0.3259111068988502,
      "learning_rate": 9.164029485981709e-05,
      "loss": 0.5297510623931885,
      "step": 723
    },
    {
      "epoch": 0.6843100189035917,
      "grad_norm": 0.32889351961622004,
      "learning_rate": 9.161143186011492e-05,
      "loss": 0.5781282186508179,
      "step": 724
    },
    {
      "epoch": 0.6852551984877127,
      "grad_norm": 0.3601888388946764,
      "learning_rate": 9.158252368129628e-05,
      "loss": 0.5346832275390625,
      "step": 725
    },
    {
      "epoch": 0.6862003780718336,
      "grad_norm": 0.33608149986786157,
      "learning_rate": 9.15535703547479e-05,
      "loss": 0.5458328723907471,
      "step": 726
    },
    {
      "epoch": 0.6871455576559546,
      "grad_norm": 0.33678722520947857,
      "learning_rate": 9.15245719119055e-05,
      "loss": 0.6070656776428223,
      "step": 727
    },
    {
      "epoch": 0.6880907372400756,
      "grad_norm": 0.33095961951583586,
      "learning_rate": 9.14955283842538e-05,
      "loss": 0.6221878528594971,
      "step": 728
    },
    {
      "epoch": 0.6890359168241966,
      "grad_norm": 0.3761704452709564,
      "learning_rate": 9.146643980332647e-05,
      "loss": 0.6337128281593323,
      "step": 729
    },
    {
      "epoch": 0.6899810964083176,
      "grad_norm": 0.32337784338537895,
      "learning_rate": 9.143730620070608e-05,
      "loss": 0.6010251045227051,
      "step": 730
    },
    {
      "epoch": 0.6909262759924386,
      "grad_norm": 0.34420865692180536,
      "learning_rate": 9.140812760802411e-05,
      "loss": 0.6469215154647827,
      "step": 731
    },
    {
      "epoch": 0.6918714555765595,
      "grad_norm": 0.35632815288141784,
      "learning_rate": 9.137890405696086e-05,
      "loss": 0.6641390323638916,
      "step": 732
    },
    {
      "epoch": 0.6928166351606805,
      "grad_norm": 0.3495992214000321,
      "learning_rate": 9.134963557924543e-05,
      "loss": 0.5361427664756775,
      "step": 733
    },
    {
      "epoch": 0.6937618147448015,
      "grad_norm": 0.334767621372606,
      "learning_rate": 9.132032220665574e-05,
      "loss": 0.5452948212623596,
      "step": 734
    },
    {
      "epoch": 0.6947069943289225,
      "grad_norm": 0.36235713569074884,
      "learning_rate": 9.129096397101843e-05,
      "loss": 0.5450464487075806,
      "step": 735
    },
    {
      "epoch": 0.6956521739130435,
      "grad_norm": 0.29762622472194034,
      "learning_rate": 9.126156090420888e-05,
      "loss": 0.5325775146484375,
      "step": 736
    },
    {
      "epoch": 0.6965973534971645,
      "grad_norm": 0.34804814373372595,
      "learning_rate": 9.123211303815109e-05,
      "loss": 0.7042575478553772,
      "step": 737
    },
    {
      "epoch": 0.6975425330812854,
      "grad_norm": 0.333036377506224,
      "learning_rate": 9.120262040481776e-05,
      "loss": 0.5966389179229736,
      "step": 738
    },
    {
      "epoch": 0.6984877126654064,
      "grad_norm": 0.3428809198065424,
      "learning_rate": 9.117308303623012e-05,
      "loss": 0.6494941711425781,
      "step": 739
    },
    {
      "epoch": 0.6994328922495274,
      "grad_norm": 0.3720397398192089,
      "learning_rate": 9.114350096445803e-05,
      "loss": 0.5872294902801514,
      "step": 740
    },
    {
      "epoch": 0.7003780718336484,
      "grad_norm": 0.3277713647652778,
      "learning_rate": 9.111387422161989e-05,
      "loss": 0.5815536975860596,
      "step": 741
    },
    {
      "epoch": 0.7013232514177694,
      "grad_norm": 0.4139253450363024,
      "learning_rate": 9.10842028398826e-05,
      "loss": 0.5946831703186035,
      "step": 742
    },
    {
      "epoch": 0.7022684310018904,
      "grad_norm": 0.34186327440037345,
      "learning_rate": 9.105448685146145e-05,
      "loss": 0.5891484022140503,
      "step": 743
    },
    {
      "epoch": 0.7032136105860114,
      "grad_norm": 0.34975312279519705,
      "learning_rate": 9.102472628862028e-05,
      "loss": 0.566165566444397,
      "step": 744
    },
    {
      "epoch": 0.7041587901701323,
      "grad_norm": 0.34266899671722645,
      "learning_rate": 9.099492118367123e-05,
      "loss": 0.7015571594238281,
      "step": 745
    },
    {
      "epoch": 0.7051039697542533,
      "grad_norm": 0.34440197088584,
      "learning_rate": 9.096507156897486e-05,
      "loss": 0.6671074032783508,
      "step": 746
    },
    {
      "epoch": 0.7060491493383743,
      "grad_norm": 0.34327890968983815,
      "learning_rate": 9.093517747694001e-05,
      "loss": 0.5432665348052979,
      "step": 747
    },
    {
      "epoch": 0.7069943289224953,
      "grad_norm": 0.383788586360317,
      "learning_rate": 9.090523894002385e-05,
      "loss": 0.49029800295829773,
      "step": 748
    },
    {
      "epoch": 0.7079395085066162,
      "grad_norm": 0.3243739990481308,
      "learning_rate": 9.087525599073179e-05,
      "loss": 0.5867433547973633,
      "step": 749
    },
    {
      "epoch": 0.7088846880907372,
      "grad_norm": 0.3391741415492159,
      "learning_rate": 9.084522866161746e-05,
      "loss": 0.5583163499832153,
      "step": 750
    },
    {
      "epoch": 0.7098298676748582,
      "grad_norm": 0.3202548074361154,
      "learning_rate": 9.081515698528267e-05,
      "loss": 0.5204033851623535,
      "step": 751
    },
    {
      "epoch": 0.7107750472589792,
      "grad_norm": 0.3513887479988686,
      "learning_rate": 9.078504099437739e-05,
      "loss": 0.5990594029426575,
      "step": 752
    },
    {
      "epoch": 0.7117202268431002,
      "grad_norm": 0.3701690224085548,
      "learning_rate": 9.075488072159969e-05,
      "loss": 0.6873788833618164,
      "step": 753
    },
    {
      "epoch": 0.7126654064272212,
      "grad_norm": 0.34409795477528654,
      "learning_rate": 9.072467619969572e-05,
      "loss": 0.5670281648635864,
      "step": 754
    },
    {
      "epoch": 0.7136105860113422,
      "grad_norm": 0.3424192780972742,
      "learning_rate": 9.069442746145971e-05,
      "loss": 0.6809086203575134,
      "step": 755
    },
    {
      "epoch": 0.7145557655954632,
      "grad_norm": 0.38205809253197065,
      "learning_rate": 9.066413453973383e-05,
      "loss": 0.670640230178833,
      "step": 756
    },
    {
      "epoch": 0.7155009451795841,
      "grad_norm": 0.32035799534029546,
      "learning_rate": 9.06337974674083e-05,
      "loss": 0.5516180992126465,
      "step": 757
    },
    {
      "epoch": 0.7164461247637051,
      "grad_norm": 0.3220012307894953,
      "learning_rate": 9.060341627742121e-05,
      "loss": 0.5999091863632202,
      "step": 758
    },
    {
      "epoch": 0.717391304347826,
      "grad_norm": 0.37619498300186494,
      "learning_rate": 9.057299100275856e-05,
      "loss": 0.6089802980422974,
      "step": 759
    },
    {
      "epoch": 0.718336483931947,
      "grad_norm": 0.3206718476573975,
      "learning_rate": 9.054252167645425e-05,
      "loss": 0.4593515992164612,
      "step": 760
    },
    {
      "epoch": 0.719281663516068,
      "grad_norm": 0.33913181563556977,
      "learning_rate": 9.051200833158999e-05,
      "loss": 0.5758387446403503,
      "step": 761
    },
    {
      "epoch": 0.720226843100189,
      "grad_norm": 0.3384130997687583,
      "learning_rate": 9.048145100129529e-05,
      "loss": 0.5774932503700256,
      "step": 762
    },
    {
      "epoch": 0.72117202268431,
      "grad_norm": 0.323796555970056,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.6051476001739502,
      "step": 763
    },
    {
      "epoch": 0.722117202268431,
      "grad_norm": 0.34573214347120657,
      "learning_rate": 9.042020451717124e-05,
      "loss": 0.6143999099731445,
      "step": 764
    },
    {
      "epoch": 0.723062381852552,
      "grad_norm": 0.32225068954767555,
      "learning_rate": 9.038951542983956e-05,
      "loss": 0.5983526706695557,
      "step": 765
    },
    {
      "epoch": 0.724007561436673,
      "grad_norm": 0.353193196559549,
      "learning_rate": 9.035878249007263e-05,
      "loss": 0.678943932056427,
      "step": 766
    },
    {
      "epoch": 0.724952741020794,
      "grad_norm": 0.3601469425309262,
      "learning_rate": 9.032800573123838e-05,
      "loss": 0.6037911176681519,
      "step": 767
    },
    {
      "epoch": 0.725897920604915,
      "grad_norm": 0.4013138119928066,
      "learning_rate": 9.02971851867523e-05,
      "loss": 0.7431052923202515,
      "step": 768
    },
    {
      "epoch": 0.7268431001890359,
      "grad_norm": 0.3828759865377993,
      "learning_rate": 9.026632089007746e-05,
      "loss": 0.6403226852416992,
      "step": 769
    },
    {
      "epoch": 0.7277882797731569,
      "grad_norm": 0.3024665639388889,
      "learning_rate": 9.023541287472435e-05,
      "loss": 0.5426170229911804,
      "step": 770
    },
    {
      "epoch": 0.7287334593572778,
      "grad_norm": 0.3371562274917901,
      "learning_rate": 9.020446117425101e-05,
      "loss": 0.5473175048828125,
      "step": 771
    },
    {
      "epoch": 0.7296786389413988,
      "grad_norm": 0.34238902264225934,
      "learning_rate": 9.017346582226289e-05,
      "loss": 0.620408296585083,
      "step": 772
    },
    {
      "epoch": 0.7306238185255198,
      "grad_norm": 0.28826465714254246,
      "learning_rate": 9.014242685241281e-05,
      "loss": 0.508825421333313,
      "step": 773
    },
    {
      "epoch": 0.7315689981096408,
      "grad_norm": 0.35002047354494537,
      "learning_rate": 9.011134429840098e-05,
      "loss": 0.5483698844909668,
      "step": 774
    },
    {
      "epoch": 0.7325141776937618,
      "grad_norm": 0.3158457765112785,
      "learning_rate": 9.008021819397487e-05,
      "loss": 0.5078508853912354,
      "step": 775
    },
    {
      "epoch": 0.7334593572778828,
      "grad_norm": 0.3837644236444848,
      "learning_rate": 9.004904857292932e-05,
      "loss": 0.6113440990447998,
      "step": 776
    },
    {
      "epoch": 0.7344045368620038,
      "grad_norm": 0.3325056668467959,
      "learning_rate": 9.001783546910636e-05,
      "loss": 0.4990871250629425,
      "step": 777
    },
    {
      "epoch": 0.7353497164461248,
      "grad_norm": 0.3557775545303774,
      "learning_rate": 8.998657891639524e-05,
      "loss": 0.6218876838684082,
      "step": 778
    },
    {
      "epoch": 0.7362948960302458,
      "grad_norm": 0.3422818600117004,
      "learning_rate": 8.995527894873239e-05,
      "loss": 0.6014370918273926,
      "step": 779
    },
    {
      "epoch": 0.7372400756143668,
      "grad_norm": 0.31525269550233004,
      "learning_rate": 8.992393560010137e-05,
      "loss": 0.5002023577690125,
      "step": 780
    },
    {
      "epoch": 0.7381852551984878,
      "grad_norm": 0.3256335252705099,
      "learning_rate": 8.98925489045329e-05,
      "loss": 0.582957923412323,
      "step": 781
    },
    {
      "epoch": 0.7391304347826086,
      "grad_norm": 0.31638153705702066,
      "learning_rate": 8.986111889610462e-05,
      "loss": 0.5396153926849365,
      "step": 782
    },
    {
      "epoch": 0.7400756143667296,
      "grad_norm": 0.3529201011137394,
      "learning_rate": 8.982964560894136e-05,
      "loss": 0.6551094055175781,
      "step": 783
    },
    {
      "epoch": 0.7410207939508506,
      "grad_norm": 0.36508538399076296,
      "learning_rate": 8.979812907721486e-05,
      "loss": 0.6506434082984924,
      "step": 784
    },
    {
      "epoch": 0.7419659735349716,
      "grad_norm": 0.3202486715781944,
      "learning_rate": 8.976656933514378e-05,
      "loss": 0.5493451356887817,
      "step": 785
    },
    {
      "epoch": 0.7429111531190926,
      "grad_norm": 0.31874314392825337,
      "learning_rate": 8.973496641699374e-05,
      "loss": 0.5209483504295349,
      "step": 786
    },
    {
      "epoch": 0.7438563327032136,
      "grad_norm": 0.324856935105145,
      "learning_rate": 8.97033203570773e-05,
      "loss": 0.5625597238540649,
      "step": 787
    },
    {
      "epoch": 0.7448015122873346,
      "grad_norm": 0.3133516437676061,
      "learning_rate": 8.96716311897537e-05,
      "loss": 0.5825165510177612,
      "step": 788
    },
    {
      "epoch": 0.7457466918714556,
      "grad_norm": 0.34445795748324964,
      "learning_rate": 8.963989894942912e-05,
      "loss": 0.54851895570755,
      "step": 789
    },
    {
      "epoch": 0.7466918714555766,
      "grad_norm": 0.3383139493898251,
      "learning_rate": 8.960812367055646e-05,
      "loss": 0.5722180008888245,
      "step": 790
    },
    {
      "epoch": 0.7476370510396976,
      "grad_norm": 0.3395511006452421,
      "learning_rate": 8.957630538763537e-05,
      "loss": 0.6150497198104858,
      "step": 791
    },
    {
      "epoch": 0.7485822306238186,
      "grad_norm": 0.36107745323846147,
      "learning_rate": 8.954444413521214e-05,
      "loss": 0.6510486602783203,
      "step": 792
    },
    {
      "epoch": 0.7495274102079396,
      "grad_norm": 0.36285537424588726,
      "learning_rate": 8.951253994787975e-05,
      "loss": 0.6148955821990967,
      "step": 793
    },
    {
      "epoch": 0.7504725897920604,
      "grad_norm": 0.34147462797603173,
      "learning_rate": 8.948059286027779e-05,
      "loss": 0.5363318920135498,
      "step": 794
    },
    {
      "epoch": 0.7514177693761814,
      "grad_norm": 0.3240697346265293,
      "learning_rate": 8.944860290709244e-05,
      "loss": 0.6190799474716187,
      "step": 795
    },
    {
      "epoch": 0.7523629489603024,
      "grad_norm": 0.31939603686156964,
      "learning_rate": 8.941657012305643e-05,
      "loss": 0.5211721658706665,
      "step": 796
    },
    {
      "epoch": 0.7533081285444234,
      "grad_norm": 0.33892202359333445,
      "learning_rate": 8.938449454294893e-05,
      "loss": 0.5587764978408813,
      "step": 797
    },
    {
      "epoch": 0.7542533081285444,
      "grad_norm": 0.3336156237242459,
      "learning_rate": 8.935237620159565e-05,
      "loss": 0.5625824928283691,
      "step": 798
    },
    {
      "epoch": 0.7551984877126654,
      "grad_norm": 0.35708959720530015,
      "learning_rate": 8.93202151338687e-05,
      "loss": 0.6562711596488953,
      "step": 799
    },
    {
      "epoch": 0.7561436672967864,
      "grad_norm": 0.3001762903717714,
      "learning_rate": 8.928801137468654e-05,
      "loss": 0.4926139712333679,
      "step": 800
    },
    {
      "epoch": 0.7570888468809074,
      "grad_norm": 0.30303667861572775,
      "learning_rate": 8.925576495901405e-05,
      "loss": 0.5090972781181335,
      "step": 801
    },
    {
      "epoch": 0.7580340264650284,
      "grad_norm": 0.3373874647143659,
      "learning_rate": 8.922347592186238e-05,
      "loss": 0.5190160274505615,
      "step": 802
    },
    {
      "epoch": 0.7589792060491494,
      "grad_norm": 0.3475511036585332,
      "learning_rate": 8.919114429828897e-05,
      "loss": 0.588807225227356,
      "step": 803
    },
    {
      "epoch": 0.7599243856332704,
      "grad_norm": 0.35103939422677083,
      "learning_rate": 8.915877012339751e-05,
      "loss": 0.5425076484680176,
      "step": 804
    },
    {
      "epoch": 0.7608695652173914,
      "grad_norm": 0.3650821546567092,
      "learning_rate": 8.912635343233784e-05,
      "loss": 0.6311471462249756,
      "step": 805
    },
    {
      "epoch": 0.7618147448015122,
      "grad_norm": 0.28787085548118374,
      "learning_rate": 8.909389426030603e-05,
      "loss": 0.5453073978424072,
      "step": 806
    },
    {
      "epoch": 0.7627599243856332,
      "grad_norm": 0.3330569234866635,
      "learning_rate": 8.906139264254422e-05,
      "loss": 0.5542840361595154,
      "step": 807
    },
    {
      "epoch": 0.7637051039697542,
      "grad_norm": 0.3292294492752479,
      "learning_rate": 8.902884861434065e-05,
      "loss": 0.6399264335632324,
      "step": 808
    },
    {
      "epoch": 0.7646502835538752,
      "grad_norm": 0.33054166360598963,
      "learning_rate": 8.899626221102964e-05,
      "loss": 0.584404468536377,
      "step": 809
    },
    {
      "epoch": 0.7655954631379962,
      "grad_norm": 0.3260378774264709,
      "learning_rate": 8.896363346799146e-05,
      "loss": 0.5194777846336365,
      "step": 810
    },
    {
      "epoch": 0.7665406427221172,
      "grad_norm": 0.3448674916202805,
      "learning_rate": 8.89309624206524e-05,
      "loss": 0.628821611404419,
      "step": 811
    },
    {
      "epoch": 0.7674858223062382,
      "grad_norm": 0.33239970374474187,
      "learning_rate": 8.889824910448463e-05,
      "loss": 0.6281082630157471,
      "step": 812
    },
    {
      "epoch": 0.7684310018903592,
      "grad_norm": 0.3217448484936856,
      "learning_rate": 8.886549355500628e-05,
      "loss": 0.539427638053894,
      "step": 813
    },
    {
      "epoch": 0.7693761814744802,
      "grad_norm": 0.32372791822574004,
      "learning_rate": 8.883269580778126e-05,
      "loss": 0.565782904624939,
      "step": 814
    },
    {
      "epoch": 0.7703213610586012,
      "grad_norm": 0.357451568244404,
      "learning_rate": 8.879985589841937e-05,
      "loss": 0.5804085731506348,
      "step": 815
    },
    {
      "epoch": 0.7712665406427222,
      "grad_norm": 0.3364468317158487,
      "learning_rate": 8.876697386257614e-05,
      "loss": 0.5743640661239624,
      "step": 816
    },
    {
      "epoch": 0.7722117202268431,
      "grad_norm": 0.3194186805386807,
      "learning_rate": 8.873404973595285e-05,
      "loss": 0.5452738404273987,
      "step": 817
    },
    {
      "epoch": 0.7731568998109641,
      "grad_norm": 0.38454951441409196,
      "learning_rate": 8.870108355429645e-05,
      "loss": 0.7016465663909912,
      "step": 818
    },
    {
      "epoch": 0.774102079395085,
      "grad_norm": 0.3403083004190255,
      "learning_rate": 8.86680753533996e-05,
      "loss": 0.5867838859558105,
      "step": 819
    },
    {
      "epoch": 0.775047258979206,
      "grad_norm": 0.3203249285490356,
      "learning_rate": 8.863502516910058e-05,
      "loss": 0.4775681495666504,
      "step": 820
    },
    {
      "epoch": 0.775992438563327,
      "grad_norm": 0.39879921731278883,
      "learning_rate": 8.860193303728319e-05,
      "loss": 0.5532201528549194,
      "step": 821
    },
    {
      "epoch": 0.776937618147448,
      "grad_norm": 0.3475159378168436,
      "learning_rate": 8.856879899387686e-05,
      "loss": 0.525693416595459,
      "step": 822
    },
    {
      "epoch": 0.777882797731569,
      "grad_norm": 0.2923508590148474,
      "learning_rate": 8.853562307485647e-05,
      "loss": 0.46732297539711,
      "step": 823
    },
    {
      "epoch": 0.77882797731569,
      "grad_norm": 0.3679889952932248,
      "learning_rate": 8.850240531624236e-05,
      "loss": 0.5720497965812683,
      "step": 824
    },
    {
      "epoch": 0.779773156899811,
      "grad_norm": 0.311776258355539,
      "learning_rate": 8.846914575410034e-05,
      "loss": 0.5420946478843689,
      "step": 825
    },
    {
      "epoch": 0.780718336483932,
      "grad_norm": 0.3504529352389468,
      "learning_rate": 8.843584442454158e-05,
      "loss": 0.6371745467185974,
      "step": 826
    },
    {
      "epoch": 0.781663516068053,
      "grad_norm": 0.2904439152171306,
      "learning_rate": 8.840250136372259e-05,
      "loss": 0.4990834593772888,
      "step": 827
    },
    {
      "epoch": 0.782608695652174,
      "grad_norm": 0.3399574573864736,
      "learning_rate": 8.836911660784521e-05,
      "loss": 0.5316848754882812,
      "step": 828
    },
    {
      "epoch": 0.7835538752362949,
      "grad_norm": 0.34222127251120565,
      "learning_rate": 8.833569019315653e-05,
      "loss": 0.5650478601455688,
      "step": 829
    },
    {
      "epoch": 0.7844990548204159,
      "grad_norm": 0.33061666465127526,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.5605599880218506,
      "step": 830
    },
    {
      "epoch": 0.7854442344045368,
      "grad_norm": 0.33261640320316077,
      "learning_rate": 8.826871253255984e-05,
      "loss": 0.6091066598892212,
      "step": 831
    },
    {
      "epoch": 0.7863894139886578,
      "grad_norm": 0.30592171503743243,
      "learning_rate": 8.8235161359372e-05,
      "loss": 0.5505962371826172,
      "step": 832
    },
    {
      "epoch": 0.7873345935727788,
      "grad_norm": 0.3558220704776237,
      "learning_rate": 8.820156867281319e-05,
      "loss": 0.5606878995895386,
      "step": 833
    },
    {
      "epoch": 0.7882797731568998,
      "grad_norm": 0.3047021449631442,
      "learning_rate": 8.816793450935625e-05,
      "loss": 0.547127902507782,
      "step": 834
    },
    {
      "epoch": 0.7892249527410208,
      "grad_norm": 0.35376546067206616,
      "learning_rate": 8.81342589055191e-05,
      "loss": 0.5704728364944458,
      "step": 835
    },
    {
      "epoch": 0.7901701323251418,
      "grad_norm": 0.32589881524391195,
      "learning_rate": 8.810054189786461e-05,
      "loss": 0.5659046173095703,
      "step": 836
    },
    {
      "epoch": 0.7911153119092628,
      "grad_norm": 0.3280858933061696,
      "learning_rate": 8.806678352300062e-05,
      "loss": 0.5819205045700073,
      "step": 837
    },
    {
      "epoch": 0.7920604914933838,
      "grad_norm": 0.3511757225711003,
      "learning_rate": 8.803298381757988e-05,
      "loss": 0.5203192234039307,
      "step": 838
    },
    {
      "epoch": 0.7930056710775047,
      "grad_norm": 0.3419364169631027,
      "learning_rate": 8.799914281830001e-05,
      "loss": 0.6447063684463501,
      "step": 839
    },
    {
      "epoch": 0.7939508506616257,
      "grad_norm": 0.3181488727171322,
      "learning_rate": 8.79652605619035e-05,
      "loss": 0.5635242462158203,
      "step": 840
    },
    {
      "epoch": 0.7948960302457467,
      "grad_norm": 0.3661649181017646,
      "learning_rate": 8.793133708517758e-05,
      "loss": 0.534828782081604,
      "step": 841
    },
    {
      "epoch": 0.7958412098298677,
      "grad_norm": 0.3063433984772507,
      "learning_rate": 8.789737242495428e-05,
      "loss": 0.5489747524261475,
      "step": 842
    },
    {
      "epoch": 0.7967863894139886,
      "grad_norm": 0.3435180685753377,
      "learning_rate": 8.786336661811031e-05,
      "loss": 0.6090033054351807,
      "step": 843
    },
    {
      "epoch": 0.7977315689981096,
      "grad_norm": 0.3365418041979608,
      "learning_rate": 8.782931970156707e-05,
      "loss": 0.6043979525566101,
      "step": 844
    },
    {
      "epoch": 0.7986767485822306,
      "grad_norm": 0.2863380250741046,
      "learning_rate": 8.77952317122906e-05,
      "loss": 0.47669416666030884,
      "step": 845
    },
    {
      "epoch": 0.7996219281663516,
      "grad_norm": 0.3431511519219471,
      "learning_rate": 8.776110268729153e-05,
      "loss": 0.607347846031189,
      "step": 846
    },
    {
      "epoch": 0.8005671077504726,
      "grad_norm": 0.32591092212233097,
      "learning_rate": 8.772693266362503e-05,
      "loss": 0.5159330368041992,
      "step": 847
    },
    {
      "epoch": 0.8015122873345936,
      "grad_norm": 0.30730781960729076,
      "learning_rate": 8.769272167839082e-05,
      "loss": 0.5445238351821899,
      "step": 848
    },
    {
      "epoch": 0.8024574669187146,
      "grad_norm": 0.3631460071092267,
      "learning_rate": 8.765846976873305e-05,
      "loss": 0.6209941506385803,
      "step": 849
    },
    {
      "epoch": 0.8034026465028355,
      "grad_norm": 0.3419294318201916,
      "learning_rate": 8.762417697184033e-05,
      "loss": 0.5780820250511169,
      "step": 850
    },
    {
      "epoch": 0.8043478260869565,
      "grad_norm": 0.3549073791748884,
      "learning_rate": 8.758984332494565e-05,
      "loss": 0.6059643030166626,
      "step": 851
    },
    {
      "epoch": 0.8052930056710775,
      "grad_norm": 0.3581524581532561,
      "learning_rate": 8.755546886532638e-05,
      "loss": 0.629655122756958,
      "step": 852
    },
    {
      "epoch": 0.8062381852551985,
      "grad_norm": 0.34581266696384255,
      "learning_rate": 8.752105363030414e-05,
      "loss": 0.6075328588485718,
      "step": 853
    },
    {
      "epoch": 0.8071833648393195,
      "grad_norm": 0.3696799634410484,
      "learning_rate": 8.748659765724488e-05,
      "loss": 0.6427270174026489,
      "step": 854
    },
    {
      "epoch": 0.8081285444234405,
      "grad_norm": 0.33174356127589877,
      "learning_rate": 8.745210098355878e-05,
      "loss": 0.5859735012054443,
      "step": 855
    },
    {
      "epoch": 0.8090737240075614,
      "grad_norm": 0.331913918804586,
      "learning_rate": 8.741756364670018e-05,
      "loss": 0.6096954941749573,
      "step": 856
    },
    {
      "epoch": 0.8100189035916824,
      "grad_norm": 0.3296633726524113,
      "learning_rate": 8.738298568416756e-05,
      "loss": 0.6119064092636108,
      "step": 857
    },
    {
      "epoch": 0.8109640831758034,
      "grad_norm": 0.34652677651298985,
      "learning_rate": 8.734836713350358e-05,
      "loss": 0.5950466394424438,
      "step": 858
    },
    {
      "epoch": 0.8119092627599244,
      "grad_norm": 0.3227977469868169,
      "learning_rate": 8.731370803229487e-05,
      "loss": 0.5704362392425537,
      "step": 859
    },
    {
      "epoch": 0.8128544423440454,
      "grad_norm": 0.35197728122964594,
      "learning_rate": 8.727900841817215e-05,
      "loss": 0.6311769485473633,
      "step": 860
    },
    {
      "epoch": 0.8137996219281664,
      "grad_norm": 0.32173161586560206,
      "learning_rate": 8.724426832881015e-05,
      "loss": 0.6004565954208374,
      "step": 861
    },
    {
      "epoch": 0.8147448015122873,
      "grad_norm": 0.32606806713962655,
      "learning_rate": 8.720948780192746e-05,
      "loss": 0.5663374066352844,
      "step": 862
    },
    {
      "epoch": 0.8156899810964083,
      "grad_norm": 0.3402970946534143,
      "learning_rate": 8.717466687528664e-05,
      "loss": 0.6377445459365845,
      "step": 863
    },
    {
      "epoch": 0.8166351606805293,
      "grad_norm": 0.32289507749457746,
      "learning_rate": 8.71398055866941e-05,
      "loss": 0.5530605912208557,
      "step": 864
    },
    {
      "epoch": 0.8175803402646503,
      "grad_norm": 0.35265139447631766,
      "learning_rate": 8.710490397400006e-05,
      "loss": 0.6040802001953125,
      "step": 865
    },
    {
      "epoch": 0.8185255198487713,
      "grad_norm": 0.30191477856419063,
      "learning_rate": 8.706996207509853e-05,
      "loss": 0.5256832838058472,
      "step": 866
    },
    {
      "epoch": 0.8194706994328923,
      "grad_norm": 0.3075478077384235,
      "learning_rate": 8.703497992792728e-05,
      "loss": 0.6391090154647827,
      "step": 867
    },
    {
      "epoch": 0.8204158790170132,
      "grad_norm": 0.3020248448550408,
      "learning_rate": 8.699995757046774e-05,
      "loss": 0.5322891473770142,
      "step": 868
    },
    {
      "epoch": 0.8213610586011342,
      "grad_norm": 0.33784410268236476,
      "learning_rate": 8.696489504074501e-05,
      "loss": 0.5735327005386353,
      "step": 869
    },
    {
      "epoch": 0.8223062381852552,
      "grad_norm": 0.3183690250115243,
      "learning_rate": 8.692979237682786e-05,
      "loss": 0.5889484882354736,
      "step": 870
    },
    {
      "epoch": 0.8232514177693762,
      "grad_norm": 0.3267574135652416,
      "learning_rate": 8.689464961682852e-05,
      "loss": 0.6460490226745605,
      "step": 871
    },
    {
      "epoch": 0.8241965973534972,
      "grad_norm": 0.32451747109546975,
      "learning_rate": 8.685946679890288e-05,
      "loss": 0.5998455286026001,
      "step": 872
    },
    {
      "epoch": 0.8251417769376181,
      "grad_norm": 0.3367612219735067,
      "learning_rate": 8.682424396125024e-05,
      "loss": 0.5769582986831665,
      "step": 873
    },
    {
      "epoch": 0.8260869565217391,
      "grad_norm": 0.3366907242218308,
      "learning_rate": 8.678898114211338e-05,
      "loss": 0.619125247001648,
      "step": 874
    },
    {
      "epoch": 0.8270321361058601,
      "grad_norm": 0.33682546546131814,
      "learning_rate": 8.675367837977849e-05,
      "loss": 0.6162038445472717,
      "step": 875
    },
    {
      "epoch": 0.8279773156899811,
      "grad_norm": 0.35576005326323235,
      "learning_rate": 8.671833571257512e-05,
      "loss": 0.6362898945808411,
      "step": 876
    },
    {
      "epoch": 0.8289224952741021,
      "grad_norm": 0.33569741265827663,
      "learning_rate": 8.668295317887614e-05,
      "loss": 0.5754570960998535,
      "step": 877
    },
    {
      "epoch": 0.8298676748582231,
      "grad_norm": 0.3630615192310791,
      "learning_rate": 8.664753081709773e-05,
      "loss": 0.6256622076034546,
      "step": 878
    },
    {
      "epoch": 0.8308128544423441,
      "grad_norm": 0.35452617526942215,
      "learning_rate": 8.661206866569931e-05,
      "loss": 0.6753613948822021,
      "step": 879
    },
    {
      "epoch": 0.831758034026465,
      "grad_norm": 0.37017011697218993,
      "learning_rate": 8.657656676318346e-05,
      "loss": 0.5713620185852051,
      "step": 880
    },
    {
      "epoch": 0.832703213610586,
      "grad_norm": 0.3969664473175766,
      "learning_rate": 8.654102514809595e-05,
      "loss": 0.5944727063179016,
      "step": 881
    },
    {
      "epoch": 0.833648393194707,
      "grad_norm": 0.3314628199498843,
      "learning_rate": 8.650544385902568e-05,
      "loss": 0.5271555185317993,
      "step": 882
    },
    {
      "epoch": 0.834593572778828,
      "grad_norm": 0.3228248104375323,
      "learning_rate": 8.64698229346046e-05,
      "loss": 0.6418903470039368,
      "step": 883
    },
    {
      "epoch": 0.8355387523629489,
      "grad_norm": 0.346219768321841,
      "learning_rate": 8.643416241350771e-05,
      "loss": 0.558115541934967,
      "step": 884
    },
    {
      "epoch": 0.8364839319470699,
      "grad_norm": 0.3467792048525173,
      "learning_rate": 8.639846233445301e-05,
      "loss": 0.5193544626235962,
      "step": 885
    },
    {
      "epoch": 0.8374291115311909,
      "grad_norm": 0.35076936679181225,
      "learning_rate": 8.636272273620141e-05,
      "loss": 0.5656487941741943,
      "step": 886
    },
    {
      "epoch": 0.8383742911153119,
      "grad_norm": 0.41275306475508833,
      "learning_rate": 8.632694365755676e-05,
      "loss": 0.6541579961776733,
      "step": 887
    },
    {
      "epoch": 0.8393194706994329,
      "grad_norm": 0.31626164779055554,
      "learning_rate": 8.629112513736577e-05,
      "loss": 0.5199324488639832,
      "step": 888
    },
    {
      "epoch": 0.8402646502835539,
      "grad_norm": 0.3435087954096943,
      "learning_rate": 8.625526721451798e-05,
      "loss": 0.6210706233978271,
      "step": 889
    },
    {
      "epoch": 0.8412098298676749,
      "grad_norm": 0.3371816282298623,
      "learning_rate": 8.621936992794568e-05,
      "loss": 0.5545634031295776,
      "step": 890
    },
    {
      "epoch": 0.8421550094517959,
      "grad_norm": 0.2754647220261133,
      "learning_rate": 8.618343331662397e-05,
      "loss": 0.5040866136550903,
      "step": 891
    },
    {
      "epoch": 0.8431001890359168,
      "grad_norm": 0.3725594110423086,
      "learning_rate": 8.614745741957054e-05,
      "loss": 0.6336383819580078,
      "step": 892
    },
    {
      "epoch": 0.8440453686200378,
      "grad_norm": 0.3403579492844374,
      "learning_rate": 8.611144227584582e-05,
      "loss": 0.6933382749557495,
      "step": 893
    },
    {
      "epoch": 0.8449905482041588,
      "grad_norm": 0.33273304082546723,
      "learning_rate": 8.607538792455282e-05,
      "loss": 0.5916414856910706,
      "step": 894
    },
    {
      "epoch": 0.8459357277882797,
      "grad_norm": 0.3368252391941159,
      "learning_rate": 8.603929440483713e-05,
      "loss": 0.5381198525428772,
      "step": 895
    },
    {
      "epoch": 0.8468809073724007,
      "grad_norm": 0.34767688750029496,
      "learning_rate": 8.600316175588684e-05,
      "loss": 0.6333449482917786,
      "step": 896
    },
    {
      "epoch": 0.8478260869565217,
      "grad_norm": 0.358852171193461,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.5922762155532837,
      "step": 897
    },
    {
      "epoch": 0.8487712665406427,
      "grad_norm": 0.3228719145622494,
      "learning_rate": 8.593077922724733e-05,
      "loss": 0.6594687104225159,
      "step": 898
    },
    {
      "epoch": 0.8497164461247637,
      "grad_norm": 0.3400112209924279,
      "learning_rate": 8.589452942614655e-05,
      "loss": 0.620603084564209,
      "step": 899
    },
    {
      "epoch": 0.8506616257088847,
      "grad_norm": 0.3787889638936399,
      "learning_rate": 8.585824065298806e-05,
      "loss": 0.6358669996261597,
      "step": 900
    },
    {
      "epoch": 0.8516068052930057,
      "grad_norm": 0.30089457654468704,
      "learning_rate": 8.582191294717191e-05,
      "loss": 0.5445858836174011,
      "step": 901
    },
    {
      "epoch": 0.8525519848771267,
      "grad_norm": 0.360480657192104,
      "learning_rate": 8.578554634814051e-05,
      "loss": 0.5587654709815979,
      "step": 902
    },
    {
      "epoch": 0.8534971644612477,
      "grad_norm": 0.3376854225648232,
      "learning_rate": 8.574914089537845e-05,
      "loss": 0.5384718179702759,
      "step": 903
    },
    {
      "epoch": 0.8544423440453687,
      "grad_norm": 0.3345264942614641,
      "learning_rate": 8.571269662841253e-05,
      "loss": 0.6188182234764099,
      "step": 904
    },
    {
      "epoch": 0.8553875236294896,
      "grad_norm": 0.34603637250403724,
      "learning_rate": 8.567621358681165e-05,
      "loss": 0.6056867837905884,
      "step": 905
    },
    {
      "epoch": 0.8563327032136105,
      "grad_norm": 0.33613879073885977,
      "learning_rate": 8.563969181018687e-05,
      "loss": 0.5191941261291504,
      "step": 906
    },
    {
      "epoch": 0.8572778827977315,
      "grad_norm": 0.31468644192565,
      "learning_rate": 8.560313133819125e-05,
      "loss": 0.5207089185714722,
      "step": 907
    },
    {
      "epoch": 0.8582230623818525,
      "grad_norm": 0.32601544469349075,
      "learning_rate": 8.556653221051989e-05,
      "loss": 0.558337390422821,
      "step": 908
    },
    {
      "epoch": 0.8591682419659735,
      "grad_norm": 0.3276993242378383,
      "learning_rate": 8.552989446690987e-05,
      "loss": 0.6073688864707947,
      "step": 909
    },
    {
      "epoch": 0.8601134215500945,
      "grad_norm": 0.32323515659780516,
      "learning_rate": 8.549321814714018e-05,
      "loss": 0.5968376398086548,
      "step": 910
    },
    {
      "epoch": 0.8610586011342155,
      "grad_norm": 0.3447689381889637,
      "learning_rate": 8.545650329103167e-05,
      "loss": 0.5920617580413818,
      "step": 911
    },
    {
      "epoch": 0.8620037807183365,
      "grad_norm": 0.27100190593608725,
      "learning_rate": 8.54197499384471e-05,
      "loss": 0.45354998111724854,
      "step": 912
    },
    {
      "epoch": 0.8629489603024575,
      "grad_norm": 0.3338378710432014,
      "learning_rate": 8.538295812929096e-05,
      "loss": 0.5669863224029541,
      "step": 913
    },
    {
      "epoch": 0.8638941398865785,
      "grad_norm": 0.3111880738493051,
      "learning_rate": 8.534612790350953e-05,
      "loss": 0.5374206900596619,
      "step": 914
    },
    {
      "epoch": 0.8648393194706995,
      "grad_norm": 0.37218016955050703,
      "learning_rate": 8.530925930109078e-05,
      "loss": 0.5519886016845703,
      "step": 915
    },
    {
      "epoch": 0.8657844990548205,
      "grad_norm": 0.3108074807238979,
      "learning_rate": 8.527235236206436e-05,
      "loss": 0.5425099730491638,
      "step": 916
    },
    {
      "epoch": 0.8667296786389413,
      "grad_norm": 0.34390416718266154,
      "learning_rate": 8.523540712650154e-05,
      "loss": 0.5577598810195923,
      "step": 917
    },
    {
      "epoch": 0.8676748582230623,
      "grad_norm": 0.3743476972260953,
      "learning_rate": 8.519842363451516e-05,
      "loss": 0.5855838060379028,
      "step": 918
    },
    {
      "epoch": 0.8686200378071833,
      "grad_norm": 0.34040187855015014,
      "learning_rate": 8.516140192625962e-05,
      "loss": 0.5935339331626892,
      "step": 919
    },
    {
      "epoch": 0.8695652173913043,
      "grad_norm": 0.35456224577827883,
      "learning_rate": 8.51243420419308e-05,
      "loss": 0.5663225650787354,
      "step": 920
    },
    {
      "epoch": 0.8705103969754253,
      "grad_norm": 0.36082074773000805,
      "learning_rate": 8.508724402176602e-05,
      "loss": 0.5715097188949585,
      "step": 921
    },
    {
      "epoch": 0.8714555765595463,
      "grad_norm": 0.36047183574170794,
      "learning_rate": 8.5050107906044e-05,
      "loss": 0.6112238168716431,
      "step": 922
    },
    {
      "epoch": 0.8724007561436673,
      "grad_norm": 0.3598647067290654,
      "learning_rate": 8.501293373508485e-05,
      "loss": 0.5935978293418884,
      "step": 923
    },
    {
      "epoch": 0.8733459357277883,
      "grad_norm": 0.32604457140635346,
      "learning_rate": 8.497572154924997e-05,
      "loss": 0.5630688071250916,
      "step": 924
    },
    {
      "epoch": 0.8742911153119093,
      "grad_norm": 0.30902805685814433,
      "learning_rate": 8.493847138894209e-05,
      "loss": 0.5242061614990234,
      "step": 925
    },
    {
      "epoch": 0.8752362948960303,
      "grad_norm": 0.3335408720381236,
      "learning_rate": 8.490118329460509e-05,
      "loss": 0.586224377155304,
      "step": 926
    },
    {
      "epoch": 0.8761814744801513,
      "grad_norm": 0.3641696774689055,
      "learning_rate": 8.486385730672407e-05,
      "loss": 0.5616868138313293,
      "step": 927
    },
    {
      "epoch": 0.8771266540642723,
      "grad_norm": 0.34127805455986504,
      "learning_rate": 8.48264934658253e-05,
      "loss": 0.5934858322143555,
      "step": 928
    },
    {
      "epoch": 0.8780718336483931,
      "grad_norm": 0.3362844152765701,
      "learning_rate": 8.47890918124761e-05,
      "loss": 0.6053175926208496,
      "step": 929
    },
    {
      "epoch": 0.8790170132325141,
      "grad_norm": 0.29218442587492904,
      "learning_rate": 8.475165238728489e-05,
      "loss": 0.5871474146842957,
      "step": 930
    },
    {
      "epoch": 0.8799621928166351,
      "grad_norm": 0.361730540432299,
      "learning_rate": 8.471417523090108e-05,
      "loss": 0.5799127817153931,
      "step": 931
    },
    {
      "epoch": 0.8809073724007561,
      "grad_norm": 0.30988338135254784,
      "learning_rate": 8.467666038401502e-05,
      "loss": 0.5906287431716919,
      "step": 932
    },
    {
      "epoch": 0.8818525519848771,
      "grad_norm": 0.30613792201765655,
      "learning_rate": 8.463910788735805e-05,
      "loss": 0.5513879060745239,
      "step": 933
    },
    {
      "epoch": 0.8827977315689981,
      "grad_norm": 0.37561115845292375,
      "learning_rate": 8.46015177817023e-05,
      "loss": 0.6075446605682373,
      "step": 934
    },
    {
      "epoch": 0.8837429111531191,
      "grad_norm": 0.31058548401655517,
      "learning_rate": 8.456389010786083e-05,
      "loss": 0.5224324464797974,
      "step": 935
    },
    {
      "epoch": 0.8846880907372401,
      "grad_norm": 0.327040668396255,
      "learning_rate": 8.45262249066874e-05,
      "loss": 0.5274821519851685,
      "step": 936
    },
    {
      "epoch": 0.8856332703213611,
      "grad_norm": 0.3570218152682555,
      "learning_rate": 8.448852221907656e-05,
      "loss": 0.5499608516693115,
      "step": 937
    },
    {
      "epoch": 0.8865784499054821,
      "grad_norm": 0.3136214547349336,
      "learning_rate": 8.445078208596357e-05,
      "loss": 0.531125545501709,
      "step": 938
    },
    {
      "epoch": 0.8875236294896031,
      "grad_norm": 0.348943226599309,
      "learning_rate": 8.441300454832432e-05,
      "loss": 0.5893869400024414,
      "step": 939
    },
    {
      "epoch": 0.888468809073724,
      "grad_norm": 0.3077480566808413,
      "learning_rate": 8.43751896471753e-05,
      "loss": 0.5132349729537964,
      "step": 940
    },
    {
      "epoch": 0.889413988657845,
      "grad_norm": 0.35241160055693005,
      "learning_rate": 8.433733742357361e-05,
      "loss": 0.6446377038955688,
      "step": 941
    },
    {
      "epoch": 0.8903591682419659,
      "grad_norm": 0.34632026451050607,
      "learning_rate": 8.429944791861687e-05,
      "loss": 0.5648569464683533,
      "step": 942
    },
    {
      "epoch": 0.8913043478260869,
      "grad_norm": 0.3766205931048789,
      "learning_rate": 8.426152117344313e-05,
      "loss": 0.6127818822860718,
      "step": 943
    },
    {
      "epoch": 0.8922495274102079,
      "grad_norm": 0.32491412296986233,
      "learning_rate": 8.42235572292309e-05,
      "loss": 0.6108893156051636,
      "step": 944
    },
    {
      "epoch": 0.8931947069943289,
      "grad_norm": 0.3289981184294011,
      "learning_rate": 8.418555612719911e-05,
      "loss": 0.6494847536087036,
      "step": 945
    },
    {
      "epoch": 0.8941398865784499,
      "grad_norm": 0.3380830080168739,
      "learning_rate": 8.414751790860697e-05,
      "loss": 0.569890022277832,
      "step": 946
    },
    {
      "epoch": 0.8950850661625709,
      "grad_norm": 0.35727809470298005,
      "learning_rate": 8.410944261475404e-05,
      "loss": 0.6030163168907166,
      "step": 947
    },
    {
      "epoch": 0.8960302457466919,
      "grad_norm": 0.31652650514171327,
      "learning_rate": 8.407133028698011e-05,
      "loss": 0.5311115980148315,
      "step": 948
    },
    {
      "epoch": 0.8969754253308129,
      "grad_norm": 0.3313864141846018,
      "learning_rate": 8.40331809666652e-05,
      "loss": 0.5332129001617432,
      "step": 949
    },
    {
      "epoch": 0.8979206049149339,
      "grad_norm": 0.32367646232798464,
      "learning_rate": 8.399499469522947e-05,
      "loss": 0.5928565263748169,
      "step": 950
    },
    {
      "epoch": 0.8988657844990549,
      "grad_norm": 0.360165483202246,
      "learning_rate": 8.395677151413319e-05,
      "loss": 0.6258502006530762,
      "step": 951
    },
    {
      "epoch": 0.8998109640831758,
      "grad_norm": 0.3428149212127628,
      "learning_rate": 8.391851146487675e-05,
      "loss": 0.6106274127960205,
      "step": 952
    },
    {
      "epoch": 0.9007561436672968,
      "grad_norm": 0.3021280682405449,
      "learning_rate": 8.388021458900054e-05,
      "loss": 0.5224393010139465,
      "step": 953
    },
    {
      "epoch": 0.9017013232514177,
      "grad_norm": 0.31377522767100885,
      "learning_rate": 8.384188092808493e-05,
      "loss": 0.56197190284729,
      "step": 954
    },
    {
      "epoch": 0.9026465028355387,
      "grad_norm": 0.3816289687316642,
      "learning_rate": 8.380351052375022e-05,
      "loss": 0.6004618406295776,
      "step": 955
    },
    {
      "epoch": 0.9035916824196597,
      "grad_norm": 0.30687181672377917,
      "learning_rate": 8.376510341765666e-05,
      "loss": 0.5154225826263428,
      "step": 956
    },
    {
      "epoch": 0.9045368620037807,
      "grad_norm": 0.32863242004035115,
      "learning_rate": 8.37266596515043e-05,
      "loss": 0.5520099401473999,
      "step": 957
    },
    {
      "epoch": 0.9054820415879017,
      "grad_norm": 0.3411468198193523,
      "learning_rate": 8.368817926703296e-05,
      "loss": 0.575688898563385,
      "step": 958
    },
    {
      "epoch": 0.9064272211720227,
      "grad_norm": 0.345714337497686,
      "learning_rate": 8.364966230602228e-05,
      "loss": 0.6091079711914062,
      "step": 959
    },
    {
      "epoch": 0.9073724007561437,
      "grad_norm": 0.32166212525481236,
      "learning_rate": 8.361110881029161e-05,
      "loss": 0.5530925989151001,
      "step": 960
    },
    {
      "epoch": 0.9083175803402647,
      "grad_norm": 0.3215238491274843,
      "learning_rate": 8.357251882169994e-05,
      "loss": 0.5914909243583679,
      "step": 961
    },
    {
      "epoch": 0.9092627599243857,
      "grad_norm": 0.3430719584798943,
      "learning_rate": 8.353389238214588e-05,
      "loss": 0.6528674960136414,
      "step": 962
    },
    {
      "epoch": 0.9102079395085066,
      "grad_norm": 0.301841156767362,
      "learning_rate": 8.349522953356763e-05,
      "loss": 0.507624626159668,
      "step": 963
    },
    {
      "epoch": 0.9111531190926276,
      "grad_norm": 0.33382412030541764,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.6037786602973938,
      "step": 964
    },
    {
      "epoch": 0.9120982986767486,
      "grad_norm": 0.3276042602189977,
      "learning_rate": 8.341779477728896e-05,
      "loss": 0.5713250637054443,
      "step": 965
    },
    {
      "epoch": 0.9130434782608695,
      "grad_norm": 0.33555221404863694,
      "learning_rate": 8.33790229536624e-05,
      "loss": 0.6059105396270752,
      "step": 966
    },
    {
      "epoch": 0.9139886578449905,
      "grad_norm": 0.3446736694880658,
      "learning_rate": 8.334021488915931e-05,
      "loss": 0.5730043649673462,
      "step": 967
    },
    {
      "epoch": 0.9149338374291115,
      "grad_norm": 0.35434970277999056,
      "learning_rate": 8.330137062591504e-05,
      "loss": 0.6256241798400879,
      "step": 968
    },
    {
      "epoch": 0.9158790170132325,
      "grad_norm": 0.3598125932506581,
      "learning_rate": 8.326249020610432e-05,
      "loss": 0.6257157325744629,
      "step": 969
    },
    {
      "epoch": 0.9168241965973535,
      "grad_norm": 0.32841781942656856,
      "learning_rate": 8.322357367194109e-05,
      "loss": 0.568682074546814,
      "step": 970
    },
    {
      "epoch": 0.9177693761814745,
      "grad_norm": 0.313676264335114,
      "learning_rate": 8.318462106567852e-05,
      "loss": 0.49590882658958435,
      "step": 971
    },
    {
      "epoch": 0.9187145557655955,
      "grad_norm": 0.3903207617379238,
      "learning_rate": 8.314563242960893e-05,
      "loss": 0.5435889363288879,
      "step": 972
    },
    {
      "epoch": 0.9196597353497165,
      "grad_norm": 0.3570305979385101,
      "learning_rate": 8.310660780606377e-05,
      "loss": 0.6679445505142212,
      "step": 973
    },
    {
      "epoch": 0.9206049149338374,
      "grad_norm": 0.34358170528731796,
      "learning_rate": 8.306754723741357e-05,
      "loss": 0.631051242351532,
      "step": 974
    },
    {
      "epoch": 0.9215500945179584,
      "grad_norm": 0.32820748588238696,
      "learning_rate": 8.302845076606786e-05,
      "loss": 0.6148118376731873,
      "step": 975
    },
    {
      "epoch": 0.9224952741020794,
      "grad_norm": 0.3564425124373959,
      "learning_rate": 8.298931843447517e-05,
      "loss": 0.6256043910980225,
      "step": 976
    },
    {
      "epoch": 0.9234404536862004,
      "grad_norm": 0.36417790490235946,
      "learning_rate": 8.295015028512298e-05,
      "loss": 0.6155723333358765,
      "step": 977
    },
    {
      "epoch": 0.9243856332703214,
      "grad_norm": 0.31357761474651563,
      "learning_rate": 8.291094636053763e-05,
      "loss": 0.4804495573043823,
      "step": 978
    },
    {
      "epoch": 0.9253308128544423,
      "grad_norm": 0.3286382527849331,
      "learning_rate": 8.28717067032843e-05,
      "loss": 0.5794953107833862,
      "step": 979
    },
    {
      "epoch": 0.9262759924385633,
      "grad_norm": 0.34041227733107626,
      "learning_rate": 8.283243135596701e-05,
      "loss": 0.5873276591300964,
      "step": 980
    },
    {
      "epoch": 0.9272211720226843,
      "grad_norm": 0.32895693362147366,
      "learning_rate": 8.279312036122848e-05,
      "loss": 0.5581244230270386,
      "step": 981
    },
    {
      "epoch": 0.9281663516068053,
      "grad_norm": 0.350814906841522,
      "learning_rate": 8.275377376175015e-05,
      "loss": 0.6132566332817078,
      "step": 982
    },
    {
      "epoch": 0.9291115311909263,
      "grad_norm": 0.3930647360107493,
      "learning_rate": 8.271439160025215e-05,
      "loss": 0.602157711982727,
      "step": 983
    },
    {
      "epoch": 0.9300567107750473,
      "grad_norm": 0.32937209093472536,
      "learning_rate": 8.267497391949318e-05,
      "loss": 0.515630841255188,
      "step": 984
    },
    {
      "epoch": 0.9310018903591682,
      "grad_norm": 0.32017420259489915,
      "learning_rate": 8.263552076227048e-05,
      "loss": 0.5074448585510254,
      "step": 985
    },
    {
      "epoch": 0.9319470699432892,
      "grad_norm": 0.3651657528946463,
      "learning_rate": 8.25960321714199e-05,
      "loss": 0.7083311080932617,
      "step": 986
    },
    {
      "epoch": 0.9328922495274102,
      "grad_norm": 0.3246072756304128,
      "learning_rate": 8.255650818981568e-05,
      "loss": 0.6071174144744873,
      "step": 987
    },
    {
      "epoch": 0.9338374291115312,
      "grad_norm": 0.37059795711617005,
      "learning_rate": 8.251694886037052e-05,
      "loss": 0.6656999588012695,
      "step": 988
    },
    {
      "epoch": 0.9347826086956522,
      "grad_norm": 0.34036470776314603,
      "learning_rate": 8.247735422603548e-05,
      "loss": 0.5541384220123291,
      "step": 989
    },
    {
      "epoch": 0.9357277882797732,
      "grad_norm": 0.3642850127682722,
      "learning_rate": 8.243772432979997e-05,
      "loss": 0.5614625811576843,
      "step": 990
    },
    {
      "epoch": 0.9366729678638941,
      "grad_norm": 0.330064105550449,
      "learning_rate": 8.239805921469168e-05,
      "loss": 0.5757848620414734,
      "step": 991
    },
    {
      "epoch": 0.9376181474480151,
      "grad_norm": 0.3261340118943475,
      "learning_rate": 8.235835892377655e-05,
      "loss": 0.5598487854003906,
      "step": 992
    },
    {
      "epoch": 0.9385633270321361,
      "grad_norm": 0.3648125117189749,
      "learning_rate": 8.231862350015866e-05,
      "loss": 0.6362212896347046,
      "step": 993
    },
    {
      "epoch": 0.9395085066162571,
      "grad_norm": 0.3677445435699814,
      "learning_rate": 8.22788529869803e-05,
      "loss": 0.6263105869293213,
      "step": 994
    },
    {
      "epoch": 0.9404536862003781,
      "grad_norm": 0.3307737109273409,
      "learning_rate": 8.223904742742181e-05,
      "loss": 0.6217019557952881,
      "step": 995
    },
    {
      "epoch": 0.941398865784499,
      "grad_norm": 0.29143132562335566,
      "learning_rate": 8.219920686470164e-05,
      "loss": 0.5148254632949829,
      "step": 996
    },
    {
      "epoch": 0.94234404536862,
      "grad_norm": 0.28927084758899196,
      "learning_rate": 8.215933134207618e-05,
      "loss": 0.5599417686462402,
      "step": 997
    },
    {
      "epoch": 0.943289224952741,
      "grad_norm": 0.3422805889592822,
      "learning_rate": 8.21194209028398e-05,
      "loss": 0.5863127708435059,
      "step": 998
    },
    {
      "epoch": 0.944234404536862,
      "grad_norm": 0.32199716726885436,
      "learning_rate": 8.20794755903248e-05,
      "loss": 0.5515451431274414,
      "step": 999
    },
    {
      "epoch": 0.945179584120983,
      "grad_norm": 0.3284213851621592,
      "learning_rate": 8.203949544790131e-05,
      "loss": 0.5047205686569214,
      "step": 1000
    },
    {
      "epoch": 0.946124763705104,
      "grad_norm": 0.3503143744266985,
      "learning_rate": 8.19994805189773e-05,
      "loss": 0.5660998821258545,
      "step": 1001
    },
    {
      "epoch": 0.947069943289225,
      "grad_norm": 0.32304944977069056,
      "learning_rate": 8.19594308469985e-05,
      "loss": 0.5563942193984985,
      "step": 1002
    },
    {
      "epoch": 0.9480151228733459,
      "grad_norm": 0.32640385393649696,
      "learning_rate": 8.19193464754484e-05,
      "loss": 0.5366355776786804,
      "step": 1003
    },
    {
      "epoch": 0.9489603024574669,
      "grad_norm": 0.31957513261638115,
      "learning_rate": 8.187922744784808e-05,
      "loss": 0.5836501121520996,
      "step": 1004
    },
    {
      "epoch": 0.9499054820415879,
      "grad_norm": 0.344343033655704,
      "learning_rate": 8.183907380775631e-05,
      "loss": 0.6069638133049011,
      "step": 1005
    },
    {
      "epoch": 0.9508506616257089,
      "grad_norm": 0.3833098201280509,
      "learning_rate": 8.179888559876943e-05,
      "loss": 0.5965646505355835,
      "step": 1006
    },
    {
      "epoch": 0.9517958412098299,
      "grad_norm": 0.331609181941832,
      "learning_rate": 8.17586628645213e-05,
      "loss": 0.5339845418930054,
      "step": 1007
    },
    {
      "epoch": 0.9527410207939508,
      "grad_norm": 0.30553529565718573,
      "learning_rate": 8.17184056486833e-05,
      "loss": 0.5603302717208862,
      "step": 1008
    },
    {
      "epoch": 0.9536862003780718,
      "grad_norm": 0.34099188974654515,
      "learning_rate": 8.16781139949642e-05,
      "loss": 0.5675671696662903,
      "step": 1009
    },
    {
      "epoch": 0.9546313799621928,
      "grad_norm": 0.30443098141583164,
      "learning_rate": 8.163778794711019e-05,
      "loss": 0.4620763659477234,
      "step": 1010
    },
    {
      "epoch": 0.9555765595463138,
      "grad_norm": 0.3395963353863875,
      "learning_rate": 8.159742754890481e-05,
      "loss": 0.5935757160186768,
      "step": 1011
    },
    {
      "epoch": 0.9565217391304348,
      "grad_norm": 0.3410525448310156,
      "learning_rate": 8.155703284416885e-05,
      "loss": 0.5883910059928894,
      "step": 1012
    },
    {
      "epoch": 0.9574669187145558,
      "grad_norm": 0.3513419301511129,
      "learning_rate": 8.151660387676039e-05,
      "loss": 0.6457619071006775,
      "step": 1013
    },
    {
      "epoch": 0.9584120982986768,
      "grad_norm": 0.3078298560016145,
      "learning_rate": 8.147614069057473e-05,
      "loss": 0.626125693321228,
      "step": 1014
    },
    {
      "epoch": 0.9593572778827977,
      "grad_norm": 0.3154834494226425,
      "learning_rate": 8.143564332954425e-05,
      "loss": 0.49725663661956787,
      "step": 1015
    },
    {
      "epoch": 0.9603024574669187,
      "grad_norm": 0.30686964104338305,
      "learning_rate": 8.13951118376385e-05,
      "loss": 0.5565370321273804,
      "step": 1016
    },
    {
      "epoch": 0.9612476370510397,
      "grad_norm": 0.3475903230698172,
      "learning_rate": 8.135454625886408e-05,
      "loss": 0.5786968469619751,
      "step": 1017
    },
    {
      "epoch": 0.9621928166351607,
      "grad_norm": 0.3545558094445856,
      "learning_rate": 8.131394663726453e-05,
      "loss": 0.5984129905700684,
      "step": 1018
    },
    {
      "epoch": 0.9631379962192816,
      "grad_norm": 0.3441294108877493,
      "learning_rate": 8.127331301692045e-05,
      "loss": 0.5508743524551392,
      "step": 1019
    },
    {
      "epoch": 0.9640831758034026,
      "grad_norm": 0.3358666312102586,
      "learning_rate": 8.123264544194933e-05,
      "loss": 0.5199525952339172,
      "step": 1020
    },
    {
      "epoch": 0.9650283553875236,
      "grad_norm": 0.3129885115893956,
      "learning_rate": 8.119194395650544e-05,
      "loss": 0.5244097709655762,
      "step": 1021
    },
    {
      "epoch": 0.9659735349716446,
      "grad_norm": 0.4104112583282951,
      "learning_rate": 8.115120860477998e-05,
      "loss": 0.6023027896881104,
      "step": 1022
    },
    {
      "epoch": 0.9669187145557656,
      "grad_norm": 0.3310147086610026,
      "learning_rate": 8.111043943100087e-05,
      "loss": 0.6022363901138306,
      "step": 1023
    },
    {
      "epoch": 0.9678638941398866,
      "grad_norm": 0.3075855846865967,
      "learning_rate": 8.106963647943274e-05,
      "loss": 0.5665014386177063,
      "step": 1024
    },
    {
      "epoch": 0.9688090737240076,
      "grad_norm": 0.3285853003039665,
      "learning_rate": 8.10287997943769e-05,
      "loss": 0.5746585130691528,
      "step": 1025
    },
    {
      "epoch": 0.9697542533081286,
      "grad_norm": 0.3120351911563498,
      "learning_rate": 8.098792942017132e-05,
      "loss": 0.5823804140090942,
      "step": 1026
    },
    {
      "epoch": 0.9706994328922496,
      "grad_norm": 0.352394669421066,
      "learning_rate": 8.09470254011905e-05,
      "loss": 0.563854992389679,
      "step": 1027
    },
    {
      "epoch": 0.9716446124763705,
      "grad_norm": 0.3809925133767755,
      "learning_rate": 8.090608778184551e-05,
      "loss": 0.6597002148628235,
      "step": 1028
    },
    {
      "epoch": 0.9725897920604915,
      "grad_norm": 0.3067107858097164,
      "learning_rate": 8.086511660658388e-05,
      "loss": 0.5306185483932495,
      "step": 1029
    },
    {
      "epoch": 0.9735349716446124,
      "grad_norm": 0.3510109591320633,
      "learning_rate": 8.082411191988957e-05,
      "loss": 0.5493956208229065,
      "step": 1030
    },
    {
      "epoch": 0.9744801512287334,
      "grad_norm": 0.36975524463869386,
      "learning_rate": 8.07830737662829e-05,
      "loss": 0.6106983423233032,
      "step": 1031
    },
    {
      "epoch": 0.9754253308128544,
      "grad_norm": 0.35797314984073014,
      "learning_rate": 8.074200219032063e-05,
      "loss": 0.6216443777084351,
      "step": 1032
    },
    {
      "epoch": 0.9763705103969754,
      "grad_norm": 0.3387627462848737,
      "learning_rate": 8.070089723659566e-05,
      "loss": 0.566179633140564,
      "step": 1033
    },
    {
      "epoch": 0.9773156899810964,
      "grad_norm": 0.5851515504724873,
      "learning_rate": 8.065975894973725e-05,
      "loss": 0.567028284072876,
      "step": 1034
    },
    {
      "epoch": 0.9782608695652174,
      "grad_norm": 0.35989697372510915,
      "learning_rate": 8.061858737441078e-05,
      "loss": 0.5599448680877686,
      "step": 1035
    },
    {
      "epoch": 0.9792060491493384,
      "grad_norm": 0.348481815582967,
      "learning_rate": 8.057738255531781e-05,
      "loss": 0.6032788753509521,
      "step": 1036
    },
    {
      "epoch": 0.9801512287334594,
      "grad_norm": 0.33083205740815524,
      "learning_rate": 8.053614453719598e-05,
      "loss": 0.5532835721969604,
      "step": 1037
    },
    {
      "epoch": 0.9810964083175804,
      "grad_norm": 0.3269415112116866,
      "learning_rate": 8.049487336481897e-05,
      "loss": 0.5081336498260498,
      "step": 1038
    },
    {
      "epoch": 0.9820415879017014,
      "grad_norm": 0.3862768249194153,
      "learning_rate": 8.045356908299648e-05,
      "loss": 0.5399956703186035,
      "step": 1039
    },
    {
      "epoch": 0.9829867674858223,
      "grad_norm": 0.33446421725708003,
      "learning_rate": 8.04122317365741e-05,
      "loss": 0.5454647541046143,
      "step": 1040
    },
    {
      "epoch": 0.9839319470699432,
      "grad_norm": 0.3352683450726139,
      "learning_rate": 8.037086137043343e-05,
      "loss": 0.6063953638076782,
      "step": 1041
    },
    {
      "epoch": 0.9848771266540642,
      "grad_norm": 0.3358874626397624,
      "learning_rate": 8.032945802949179e-05,
      "loss": 0.6309429407119751,
      "step": 1042
    },
    {
      "epoch": 0.9858223062381852,
      "grad_norm": 0.36417418238896887,
      "learning_rate": 8.02880217587024e-05,
      "loss": 0.562352180480957,
      "step": 1043
    },
    {
      "epoch": 0.9867674858223062,
      "grad_norm": 0.32396827519862165,
      "learning_rate": 8.024655260305418e-05,
      "loss": 0.5922744274139404,
      "step": 1044
    },
    {
      "epoch": 0.9877126654064272,
      "grad_norm": 0.3609547709811774,
      "learning_rate": 8.020505060757179e-05,
      "loss": 0.5951582193374634,
      "step": 1045
    },
    {
      "epoch": 0.9886578449905482,
      "grad_norm": 0.29779850340405734,
      "learning_rate": 8.01635158173155e-05,
      "loss": 0.46443837881088257,
      "step": 1046
    },
    {
      "epoch": 0.9896030245746692,
      "grad_norm": 0.3567744083303144,
      "learning_rate": 8.012194827738123e-05,
      "loss": 0.5857369899749756,
      "step": 1047
    },
    {
      "epoch": 0.9905482041587902,
      "grad_norm": 0.31646206333328875,
      "learning_rate": 8.008034803290044e-05,
      "loss": 0.5488166213035583,
      "step": 1048
    },
    {
      "epoch": 0.9914933837429112,
      "grad_norm": 0.3216978316624403,
      "learning_rate": 8.003871512904011e-05,
      "loss": 0.5582402944564819,
      "step": 1049
    },
    {
      "epoch": 0.9924385633270322,
      "grad_norm": 0.3691551786778532,
      "learning_rate": 7.999704961100266e-05,
      "loss": 0.6012238264083862,
      "step": 1050
    },
    {
      "epoch": 0.9933837429111532,
      "grad_norm": 0.3419363047799858,
      "learning_rate": 7.995535152402591e-05,
      "loss": 0.5658949613571167,
      "step": 1051
    },
    {
      "epoch": 0.994328922495274,
      "grad_norm": 0.3185468201810075,
      "learning_rate": 7.99136209133831e-05,
      "loss": 0.5790666937828064,
      "step": 1052
    },
    {
      "epoch": 0.995274102079395,
      "grad_norm": 0.30895018158314874,
      "learning_rate": 7.987185782438272e-05,
      "loss": 0.5372473001480103,
      "step": 1053
    },
    {
      "epoch": 0.996219281663516,
      "grad_norm": 0.3420451064779633,
      "learning_rate": 7.983006230236853e-05,
      "loss": 0.5918558835983276,
      "step": 1054
    },
    {
      "epoch": 0.997164461247637,
      "grad_norm": 0.3175420493508935,
      "learning_rate": 7.978823439271958e-05,
      "loss": 0.5741146802902222,
      "step": 1055
    },
    {
      "epoch": 0.998109640831758,
      "grad_norm": 0.3243643349601419,
      "learning_rate": 7.974637414084995e-05,
      "loss": 0.5844846963882446,
      "step": 1056
    },
    {
      "epoch": 0.999054820415879,
      "grad_norm": 0.3440882336159603,
      "learning_rate": 7.970448159220896e-05,
      "loss": 0.5554633140563965,
      "step": 1057
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3381998306420692,
      "learning_rate": 7.966255679228093e-05,
      "loss": 0.5672557353973389,
      "step": 1058
    },
    {
      "epoch": 1.0009451795841209,
      "grad_norm": 0.3006054472529061,
      "learning_rate": 7.96205997865852e-05,
      "loss": 0.5346417427062988,
      "step": 1059
    },
    {
      "epoch": 1.001890359168242,
      "grad_norm": 0.30424558842156435,
      "learning_rate": 7.957861062067614e-05,
      "loss": 0.5003029108047485,
      "step": 1060
    },
    {
      "epoch": 1.0028355387523629,
      "grad_norm": 0.31105299451654794,
      "learning_rate": 7.953658934014291e-05,
      "loss": 0.5437217950820923,
      "step": 1061
    },
    {
      "epoch": 1.003780718336484,
      "grad_norm": 0.32724742610158236,
      "learning_rate": 7.949453599060965e-05,
      "loss": 0.4736807346343994,
      "step": 1062
    },
    {
      "epoch": 1.0047258979206048,
      "grad_norm": 0.35380160637682545,
      "learning_rate": 7.94524506177353e-05,
      "loss": 0.46323662996292114,
      "step": 1063
    },
    {
      "epoch": 1.005671077504726,
      "grad_norm": 0.3307054223271345,
      "learning_rate": 7.941033326721355e-05,
      "loss": 0.5694730281829834,
      "step": 1064
    },
    {
      "epoch": 1.0066162570888468,
      "grad_norm": 0.38634501987912007,
      "learning_rate": 7.936818398477279e-05,
      "loss": 0.5222971439361572,
      "step": 1065
    },
    {
      "epoch": 1.007561436672968,
      "grad_norm": 0.34542597794463636,
      "learning_rate": 7.93260028161761e-05,
      "loss": 0.5696039199829102,
      "step": 1066
    },
    {
      "epoch": 1.0085066162570888,
      "grad_norm": 0.3579584082900781,
      "learning_rate": 7.92837898072212e-05,
      "loss": 0.5431487560272217,
      "step": 1067
    },
    {
      "epoch": 1.00945179584121,
      "grad_norm": 0.34657984282048937,
      "learning_rate": 7.924154500374038e-05,
      "loss": 0.5724254846572876,
      "step": 1068
    },
    {
      "epoch": 1.0103969754253308,
      "grad_norm": 0.3291139220006367,
      "learning_rate": 7.919926845160037e-05,
      "loss": 0.4852656126022339,
      "step": 1069
    },
    {
      "epoch": 1.011342155009452,
      "grad_norm": 0.3355828425334763,
      "learning_rate": 7.915696019670249e-05,
      "loss": 0.5050979852676392,
      "step": 1070
    },
    {
      "epoch": 1.0122873345935728,
      "grad_norm": 0.3391219088342965,
      "learning_rate": 7.91146202849824e-05,
      "loss": 0.5407587289810181,
      "step": 1071
    },
    {
      "epoch": 1.0132325141776937,
      "grad_norm": 0.3685116890552503,
      "learning_rate": 7.907224876241014e-05,
      "loss": 0.49539831280708313,
      "step": 1072
    },
    {
      "epoch": 1.0141776937618148,
      "grad_norm": 0.3509020462567203,
      "learning_rate": 7.902984567499015e-05,
      "loss": 0.5415762662887573,
      "step": 1073
    },
    {
      "epoch": 1.0151228733459357,
      "grad_norm": 0.3368755648242806,
      "learning_rate": 7.8987411068761e-05,
      "loss": 0.5116532444953918,
      "step": 1074
    },
    {
      "epoch": 1.0160680529300568,
      "grad_norm": 0.35863332020696453,
      "learning_rate": 7.894494498979557e-05,
      "loss": 0.5285764932632446,
      "step": 1075
    },
    {
      "epoch": 1.0170132325141776,
      "grad_norm": 0.347682499371311,
      "learning_rate": 7.890244748420095e-05,
      "loss": 0.5746331810951233,
      "step": 1076
    },
    {
      "epoch": 1.0179584120982987,
      "grad_norm": 0.3639549294723415,
      "learning_rate": 7.885991859811825e-05,
      "loss": 0.5355871915817261,
      "step": 1077
    },
    {
      "epoch": 1.0189035916824196,
      "grad_norm": 0.36033061575312403,
      "learning_rate": 7.881735837772274e-05,
      "loss": 0.5308500528335571,
      "step": 1078
    },
    {
      "epoch": 1.0198487712665407,
      "grad_norm": 0.3646905868981388,
      "learning_rate": 7.877476686922364e-05,
      "loss": 0.5174593329429626,
      "step": 1079
    },
    {
      "epoch": 1.0207939508506616,
      "grad_norm": 0.364935809143971,
      "learning_rate": 7.873214411886419e-05,
      "loss": 0.5646743774414062,
      "step": 1080
    },
    {
      "epoch": 1.0217391304347827,
      "grad_norm": 0.3231907128339707,
      "learning_rate": 7.868949017292153e-05,
      "loss": 0.5031501054763794,
      "step": 1081
    },
    {
      "epoch": 1.0226843100189036,
      "grad_norm": 0.3318367469713632,
      "learning_rate": 7.864680507770667e-05,
      "loss": 0.4745025634765625,
      "step": 1082
    },
    {
      "epoch": 1.0236294896030245,
      "grad_norm": 0.3485130722900509,
      "learning_rate": 7.860408887956443e-05,
      "loss": 0.5057541728019714,
      "step": 1083
    },
    {
      "epoch": 1.0245746691871456,
      "grad_norm": 0.3459065262837187,
      "learning_rate": 7.856134162487345e-05,
      "loss": 0.5050945281982422,
      "step": 1084
    },
    {
      "epoch": 1.0255198487712665,
      "grad_norm": 0.31917881200950676,
      "learning_rate": 7.851856336004604e-05,
      "loss": 0.5122698545455933,
      "step": 1085
    },
    {
      "epoch": 1.0264650283553876,
      "grad_norm": 0.38296980656824,
      "learning_rate": 7.847575413152816e-05,
      "loss": 0.5004159212112427,
      "step": 1086
    },
    {
      "epoch": 1.0274102079395084,
      "grad_norm": 0.34890795050921697,
      "learning_rate": 7.843291398579946e-05,
      "loss": 0.5094386339187622,
      "step": 1087
    },
    {
      "epoch": 1.0283553875236295,
      "grad_norm": 0.35572124560406254,
      "learning_rate": 7.839004296937311e-05,
      "loss": 0.5903792381286621,
      "step": 1088
    },
    {
      "epoch": 1.0293005671077504,
      "grad_norm": 0.34090523101603487,
      "learning_rate": 7.83471411287958e-05,
      "loss": 0.4651232659816742,
      "step": 1089
    },
    {
      "epoch": 1.0302457466918715,
      "grad_norm": 0.3573594693438729,
      "learning_rate": 7.830420851064766e-05,
      "loss": 0.5233197808265686,
      "step": 1090
    },
    {
      "epoch": 1.0311909262759924,
      "grad_norm": 0.39141600780958213,
      "learning_rate": 7.826124516154231e-05,
      "loss": 0.5438803434371948,
      "step": 1091
    },
    {
      "epoch": 1.0321361058601135,
      "grad_norm": 0.3377991290260998,
      "learning_rate": 7.821825112812669e-05,
      "loss": 0.5199178457260132,
      "step": 1092
    },
    {
      "epoch": 1.0330812854442344,
      "grad_norm": 0.35510627387367455,
      "learning_rate": 7.817522645708105e-05,
      "loss": 0.5120576620101929,
      "step": 1093
    },
    {
      "epoch": 1.0340264650283555,
      "grad_norm": 0.3259105052559221,
      "learning_rate": 7.813217119511892e-05,
      "loss": 0.4927425682544708,
      "step": 1094
    },
    {
      "epoch": 1.0349716446124764,
      "grad_norm": 0.3451050664721306,
      "learning_rate": 7.808908538898702e-05,
      "loss": 0.49643760919570923,
      "step": 1095
    },
    {
      "epoch": 1.0359168241965973,
      "grad_norm": 0.37627496576313124,
      "learning_rate": 7.804596908546529e-05,
      "loss": 0.5733413696289062,
      "step": 1096
    },
    {
      "epoch": 1.0368620037807184,
      "grad_norm": 0.30964563600116085,
      "learning_rate": 7.800282233136671e-05,
      "loss": 0.4565189480781555,
      "step": 1097
    },
    {
      "epoch": 1.0378071833648392,
      "grad_norm": 0.37817389874497026,
      "learning_rate": 7.795964517353735e-05,
      "loss": 0.5787177085876465,
      "step": 1098
    },
    {
      "epoch": 1.0387523629489603,
      "grad_norm": 0.3550613863436304,
      "learning_rate": 7.791643765885632e-05,
      "loss": 0.4861932396888733,
      "step": 1099
    },
    {
      "epoch": 1.0396975425330812,
      "grad_norm": 0.36424466550496865,
      "learning_rate": 7.787319983423563e-05,
      "loss": 0.49278080463409424,
      "step": 1100
    },
    {
      "epoch": 1.0406427221172023,
      "grad_norm": 0.3600111820162566,
      "learning_rate": 7.782993174662028e-05,
      "loss": 0.5108716487884521,
      "step": 1101
    },
    {
      "epoch": 1.0415879017013232,
      "grad_norm": 0.341102066600554,
      "learning_rate": 7.778663344298805e-05,
      "loss": 0.4439137578010559,
      "step": 1102
    },
    {
      "epoch": 1.0425330812854443,
      "grad_norm": 0.3999255194549586,
      "learning_rate": 7.774330497034954e-05,
      "loss": 0.4878424406051636,
      "step": 1103
    },
    {
      "epoch": 1.0434782608695652,
      "grad_norm": 0.34206875519869795,
      "learning_rate": 7.769994637574816e-05,
      "loss": 0.49135303497314453,
      "step": 1104
    },
    {
      "epoch": 1.0444234404536863,
      "grad_norm": 0.396270732830356,
      "learning_rate": 7.765655770625997e-05,
      "loss": 0.5880895853042603,
      "step": 1105
    },
    {
      "epoch": 1.0453686200378072,
      "grad_norm": 0.37445857481163236,
      "learning_rate": 7.761313900899372e-05,
      "loss": 0.5919753313064575,
      "step": 1106
    },
    {
      "epoch": 1.0463137996219283,
      "grad_norm": 0.33361721810976497,
      "learning_rate": 7.756969033109071e-05,
      "loss": 0.4871576726436615,
      "step": 1107
    },
    {
      "epoch": 1.0472589792060492,
      "grad_norm": 0.36212839843117284,
      "learning_rate": 7.752621171972484e-05,
      "loss": 0.44937264919281006,
      "step": 1108
    },
    {
      "epoch": 1.04820415879017,
      "grad_norm": 0.31770196938499207,
      "learning_rate": 7.748270322210251e-05,
      "loss": 0.43894317746162415,
      "step": 1109
    },
    {
      "epoch": 1.0491493383742911,
      "grad_norm": 0.34019012915529667,
      "learning_rate": 7.743916488546254e-05,
      "loss": 0.49072593450546265,
      "step": 1110
    },
    {
      "epoch": 1.050094517958412,
      "grad_norm": 0.39702425572288286,
      "learning_rate": 7.739559675707615e-05,
      "loss": 0.6074655055999756,
      "step": 1111
    },
    {
      "epoch": 1.0510396975425331,
      "grad_norm": 0.3287320590874013,
      "learning_rate": 7.735199888424694e-05,
      "loss": 0.5266655683517456,
      "step": 1112
    },
    {
      "epoch": 1.051984877126654,
      "grad_norm": 0.35734068642357897,
      "learning_rate": 7.730837131431075e-05,
      "loss": 0.5470563769340515,
      "step": 1113
    },
    {
      "epoch": 1.052930056710775,
      "grad_norm": 0.3478189849289446,
      "learning_rate": 7.726471409463572e-05,
      "loss": 0.5421780347824097,
      "step": 1114
    },
    {
      "epoch": 1.053875236294896,
      "grad_norm": 0.3237135296310244,
      "learning_rate": 7.722102727262215e-05,
      "loss": 0.48919299244880676,
      "step": 1115
    },
    {
      "epoch": 1.054820415879017,
      "grad_norm": 0.3545036997020006,
      "learning_rate": 7.717731089570246e-05,
      "loss": 0.48566973209381104,
      "step": 1116
    },
    {
      "epoch": 1.055765595463138,
      "grad_norm": 0.3733407862119998,
      "learning_rate": 7.713356501134123e-05,
      "loss": 0.5404857397079468,
      "step": 1117
    },
    {
      "epoch": 1.056710775047259,
      "grad_norm": 0.3705370638596789,
      "learning_rate": 7.708978966703503e-05,
      "loss": 0.5791646242141724,
      "step": 1118
    },
    {
      "epoch": 1.05765595463138,
      "grad_norm": 0.3465612646001813,
      "learning_rate": 7.704598491031238e-05,
      "loss": 0.5418069958686829,
      "step": 1119
    },
    {
      "epoch": 1.0586011342155008,
      "grad_norm": 0.33264921133479924,
      "learning_rate": 7.700215078873379e-05,
      "loss": 0.5093839764595032,
      "step": 1120
    },
    {
      "epoch": 1.059546313799622,
      "grad_norm": 0.3857196214698816,
      "learning_rate": 7.695828734989166e-05,
      "loss": 0.5572558641433716,
      "step": 1121
    },
    {
      "epoch": 1.0604914933837428,
      "grad_norm": 0.38477266594171566,
      "learning_rate": 7.691439464141014e-05,
      "loss": 0.6182390451431274,
      "step": 1122
    },
    {
      "epoch": 1.061436672967864,
      "grad_norm": 0.37822781766719804,
      "learning_rate": 7.687047271094527e-05,
      "loss": 0.5502234101295471,
      "step": 1123
    },
    {
      "epoch": 1.0623818525519848,
      "grad_norm": 0.3224181251870461,
      "learning_rate": 7.682652160618477e-05,
      "loss": 0.5187748074531555,
      "step": 1124
    },
    {
      "epoch": 1.063327032136106,
      "grad_norm": 0.3769676221355212,
      "learning_rate": 7.678254137484797e-05,
      "loss": 0.4995424747467041,
      "step": 1125
    },
    {
      "epoch": 1.0642722117202268,
      "grad_norm": 0.34223199699462975,
      "learning_rate": 7.673853206468593e-05,
      "loss": 0.49892014265060425,
      "step": 1126
    },
    {
      "epoch": 1.065217391304348,
      "grad_norm": 0.3745702567424658,
      "learning_rate": 7.669449372348121e-05,
      "loss": 0.5614603161811829,
      "step": 1127
    },
    {
      "epoch": 1.0661625708884688,
      "grad_norm": 0.37910867346422883,
      "learning_rate": 7.665042639904791e-05,
      "loss": 0.5236904621124268,
      "step": 1128
    },
    {
      "epoch": 1.0671077504725899,
      "grad_norm": 0.3526549787752036,
      "learning_rate": 7.660633013923164e-05,
      "loss": 0.517768144607544,
      "step": 1129
    },
    {
      "epoch": 1.0680529300567108,
      "grad_norm": 0.3620564876589945,
      "learning_rate": 7.656220499190936e-05,
      "loss": 0.5644733905792236,
      "step": 1130
    },
    {
      "epoch": 1.0689981096408319,
      "grad_norm": 0.34721758507082334,
      "learning_rate": 7.65180510049894e-05,
      "loss": 0.51356041431427,
      "step": 1131
    },
    {
      "epoch": 1.0699432892249527,
      "grad_norm": 0.3391884728135079,
      "learning_rate": 7.64738682264115e-05,
      "loss": 0.5052676200866699,
      "step": 1132
    },
    {
      "epoch": 1.0708884688090736,
      "grad_norm": 0.3668428423343394,
      "learning_rate": 7.64296567041465e-05,
      "loss": 0.5176701545715332,
      "step": 1133
    },
    {
      "epoch": 1.0718336483931947,
      "grad_norm": 0.36845622750757573,
      "learning_rate": 7.63854164861966e-05,
      "loss": 0.5350835919380188,
      "step": 1134
    },
    {
      "epoch": 1.0727788279773156,
      "grad_norm": 0.3675099157037802,
      "learning_rate": 7.634114762059504e-05,
      "loss": 0.5398349165916443,
      "step": 1135
    },
    {
      "epoch": 1.0737240075614367,
      "grad_norm": 0.4121960170242065,
      "learning_rate": 7.629685015540624e-05,
      "loss": 0.5729745626449585,
      "step": 1136
    },
    {
      "epoch": 1.0746691871455576,
      "grad_norm": 0.34685970316596226,
      "learning_rate": 7.625252413872564e-05,
      "loss": 0.5097899436950684,
      "step": 1137
    },
    {
      "epoch": 1.0756143667296787,
      "grad_norm": 0.372328947216553,
      "learning_rate": 7.62081696186797e-05,
      "loss": 0.5738967657089233,
      "step": 1138
    },
    {
      "epoch": 1.0765595463137996,
      "grad_norm": 0.3211871394964878,
      "learning_rate": 7.616378664342576e-05,
      "loss": 0.4503703713417053,
      "step": 1139
    },
    {
      "epoch": 1.0775047258979207,
      "grad_norm": 0.3825434905069455,
      "learning_rate": 7.611937526115218e-05,
      "loss": 0.6126779317855835,
      "step": 1140
    },
    {
      "epoch": 1.0784499054820416,
      "grad_norm": 0.330740695756803,
      "learning_rate": 7.607493552007805e-05,
      "loss": 0.4941519796848297,
      "step": 1141
    },
    {
      "epoch": 1.0793950850661627,
      "grad_norm": 0.3100647013340297,
      "learning_rate": 7.603046746845327e-05,
      "loss": 0.5109092593193054,
      "step": 1142
    },
    {
      "epoch": 1.0803402646502835,
      "grad_norm": 0.3447384419152866,
      "learning_rate": 7.598597115455853e-05,
      "loss": 0.5275675654411316,
      "step": 1143
    },
    {
      "epoch": 1.0812854442344046,
      "grad_norm": 0.3239035109902177,
      "learning_rate": 7.594144662670517e-05,
      "loss": 0.465610146522522,
      "step": 1144
    },
    {
      "epoch": 1.0822306238185255,
      "grad_norm": 0.3544250122793593,
      "learning_rate": 7.589689393323514e-05,
      "loss": 0.5390779972076416,
      "step": 1145
    },
    {
      "epoch": 1.0831758034026464,
      "grad_norm": 0.3472576904258959,
      "learning_rate": 7.585231312252103e-05,
      "loss": 0.5006203055381775,
      "step": 1146
    },
    {
      "epoch": 1.0841209829867675,
      "grad_norm": 0.32742045248928625,
      "learning_rate": 7.58077042429659e-05,
      "loss": 0.5011980533599854,
      "step": 1147
    },
    {
      "epoch": 1.0850661625708884,
      "grad_norm": 0.34570186811509307,
      "learning_rate": 7.576306734300331e-05,
      "loss": 0.5281603932380676,
      "step": 1148
    },
    {
      "epoch": 1.0860113421550095,
      "grad_norm": 0.35360557174664703,
      "learning_rate": 7.571840247109727e-05,
      "loss": 0.5011874437332153,
      "step": 1149
    },
    {
      "epoch": 1.0869565217391304,
      "grad_norm": 0.3763028616683949,
      "learning_rate": 7.56737096757421e-05,
      "loss": 0.5844608545303345,
      "step": 1150
    },
    {
      "epoch": 1.0879017013232515,
      "grad_norm": 0.34329249059330375,
      "learning_rate": 7.562898900546249e-05,
      "loss": 0.5164328217506409,
      "step": 1151
    },
    {
      "epoch": 1.0888468809073724,
      "grad_norm": 0.34923108279350734,
      "learning_rate": 7.558424050881336e-05,
      "loss": 0.4902428388595581,
      "step": 1152
    },
    {
      "epoch": 1.0897920604914935,
      "grad_norm": 0.32781131283398607,
      "learning_rate": 7.553946423437987e-05,
      "loss": 0.48593682050704956,
      "step": 1153
    },
    {
      "epoch": 1.0907372400756143,
      "grad_norm": 0.3813245752177321,
      "learning_rate": 7.549466023077732e-05,
      "loss": 0.5180710554122925,
      "step": 1154
    },
    {
      "epoch": 1.0916824196597354,
      "grad_norm": 0.3842389846293309,
      "learning_rate": 7.544982854665113e-05,
      "loss": 0.5598089694976807,
      "step": 1155
    },
    {
      "epoch": 1.0926275992438563,
      "grad_norm": 0.3990732765317015,
      "learning_rate": 7.540496923067675e-05,
      "loss": 0.514994740486145,
      "step": 1156
    },
    {
      "epoch": 1.0935727788279772,
      "grad_norm": 0.39332242339206613,
      "learning_rate": 7.536008233155967e-05,
      "loss": 0.5316749811172485,
      "step": 1157
    },
    {
      "epoch": 1.0945179584120983,
      "grad_norm": 0.4014257802494534,
      "learning_rate": 7.53151678980353e-05,
      "loss": 0.5306562185287476,
      "step": 1158
    },
    {
      "epoch": 1.0954631379962192,
      "grad_norm": 0.40683438706511843,
      "learning_rate": 7.527022597886895e-05,
      "loss": 0.5280157923698425,
      "step": 1159
    },
    {
      "epoch": 1.0964083175803403,
      "grad_norm": 0.35093009096756766,
      "learning_rate": 7.522525662285575e-05,
      "loss": 0.471016526222229,
      "step": 1160
    },
    {
      "epoch": 1.0973534971644612,
      "grad_norm": 0.3492795279703145,
      "learning_rate": 7.518025987882067e-05,
      "loss": 0.49131473898887634,
      "step": 1161
    },
    {
      "epoch": 1.0982986767485823,
      "grad_norm": 0.3700365512654156,
      "learning_rate": 7.513523579561839e-05,
      "loss": 0.5022586584091187,
      "step": 1162
    },
    {
      "epoch": 1.0992438563327032,
      "grad_norm": 0.3664141178889443,
      "learning_rate": 7.509018442213326e-05,
      "loss": 0.5361422300338745,
      "step": 1163
    },
    {
      "epoch": 1.1001890359168243,
      "grad_norm": 0.37375062682771937,
      "learning_rate": 7.504510580727926e-05,
      "loss": 0.4838642477989197,
      "step": 1164
    },
    {
      "epoch": 1.1011342155009451,
      "grad_norm": 0.34430513916870054,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.4914611876010895,
      "step": 1165
    },
    {
      "epoch": 1.1020793950850662,
      "grad_norm": 0.3454919369087684,
      "learning_rate": 7.495486704926855e-05,
      "loss": 0.5167262554168701,
      "step": 1166
    },
    {
      "epoch": 1.1030245746691871,
      "grad_norm": 0.3576388983732975,
      "learning_rate": 7.490970700408745e-05,
      "loss": 0.5493753552436829,
      "step": 1167
    },
    {
      "epoch": 1.103969754253308,
      "grad_norm": 0.35935529279830153,
      "learning_rate": 7.486451991348872e-05,
      "loss": 0.4854198694229126,
      "step": 1168
    },
    {
      "epoch": 1.1049149338374291,
      "grad_norm": 0.36258162310086284,
      "learning_rate": 7.481930582653368e-05,
      "loss": 0.534584641456604,
      "step": 1169
    },
    {
      "epoch": 1.10586011342155,
      "grad_norm": 0.40318375940631485,
      "learning_rate": 7.4774064792313e-05,
      "loss": 0.5997729301452637,
      "step": 1170
    },
    {
      "epoch": 1.106805293005671,
      "grad_norm": 0.37672199330449385,
      "learning_rate": 7.472879685994658e-05,
      "loss": 0.5467474460601807,
      "step": 1171
    },
    {
      "epoch": 1.107750472589792,
      "grad_norm": 0.3418465330088291,
      "learning_rate": 7.468350207858355e-05,
      "loss": 0.5197471380233765,
      "step": 1172
    },
    {
      "epoch": 1.108695652173913,
      "grad_norm": 0.3349900385215164,
      "learning_rate": 7.463818049740217e-05,
      "loss": 0.49897703528404236,
      "step": 1173
    },
    {
      "epoch": 1.109640831758034,
      "grad_norm": 0.33468059192631994,
      "learning_rate": 7.459283216560981e-05,
      "loss": 0.4752257168292999,
      "step": 1174
    },
    {
      "epoch": 1.110586011342155,
      "grad_norm": 0.37668593900780745,
      "learning_rate": 7.454745713244289e-05,
      "loss": 0.6101627945899963,
      "step": 1175
    },
    {
      "epoch": 1.111531190926276,
      "grad_norm": 0.3700638959184255,
      "learning_rate": 7.450205544716682e-05,
      "loss": 0.5882896184921265,
      "step": 1176
    },
    {
      "epoch": 1.112476370510397,
      "grad_norm": 0.3610223681322468,
      "learning_rate": 7.445662715907591e-05,
      "loss": 0.5122951865196228,
      "step": 1177
    },
    {
      "epoch": 1.113421550094518,
      "grad_norm": 0.4097372566646611,
      "learning_rate": 7.44111723174934e-05,
      "loss": 0.5727459788322449,
      "step": 1178
    },
    {
      "epoch": 1.114366729678639,
      "grad_norm": 0.34636616171395296,
      "learning_rate": 7.436569097177135e-05,
      "loss": 0.5205846428871155,
      "step": 1179
    },
    {
      "epoch": 1.11531190926276,
      "grad_norm": 0.365389539912882,
      "learning_rate": 7.432018317129056e-05,
      "loss": 0.5097216963768005,
      "step": 1180
    },
    {
      "epoch": 1.116257088846881,
      "grad_norm": 0.3497823602185728,
      "learning_rate": 7.427464896546061e-05,
      "loss": 0.49792999029159546,
      "step": 1181
    },
    {
      "epoch": 1.117202268431002,
      "grad_norm": 0.3657780259767824,
      "learning_rate": 7.422908840371971e-05,
      "loss": 0.46849507093429565,
      "step": 1182
    },
    {
      "epoch": 1.1181474480151228,
      "grad_norm": 0.3778320189962276,
      "learning_rate": 7.41835015355347e-05,
      "loss": 0.564968466758728,
      "step": 1183
    },
    {
      "epoch": 1.1190926275992439,
      "grad_norm": 0.39466565603873227,
      "learning_rate": 7.413788841040099e-05,
      "loss": 0.5672471523284912,
      "step": 1184
    },
    {
      "epoch": 1.1200378071833648,
      "grad_norm": 0.3270032664846419,
      "learning_rate": 7.409224907784247e-05,
      "loss": 0.48062390089035034,
      "step": 1185
    },
    {
      "epoch": 1.1209829867674859,
      "grad_norm": 0.39727291794839165,
      "learning_rate": 7.40465835874115e-05,
      "loss": 0.5376319885253906,
      "step": 1186
    },
    {
      "epoch": 1.1219281663516067,
      "grad_norm": 0.3967154845320211,
      "learning_rate": 7.400089198868885e-05,
      "loss": 0.5801239609718323,
      "step": 1187
    },
    {
      "epoch": 1.1228733459357279,
      "grad_norm": 0.3551827103415185,
      "learning_rate": 7.395517433128365e-05,
      "loss": 0.4701177477836609,
      "step": 1188
    },
    {
      "epoch": 1.1238185255198487,
      "grad_norm": 0.3845704913837229,
      "learning_rate": 7.390943066483326e-05,
      "loss": 0.5192392468452454,
      "step": 1189
    },
    {
      "epoch": 1.1247637051039698,
      "grad_norm": 0.3577816262792218,
      "learning_rate": 7.386366103900337e-05,
      "loss": 0.4965108633041382,
      "step": 1190
    },
    {
      "epoch": 1.1257088846880907,
      "grad_norm": 0.3549074535398388,
      "learning_rate": 7.381786550348774e-05,
      "loss": 0.4944078028202057,
      "step": 1191
    },
    {
      "epoch": 1.1266540642722118,
      "grad_norm": 0.3415475962658178,
      "learning_rate": 7.377204410800838e-05,
      "loss": 0.45291605591773987,
      "step": 1192
    },
    {
      "epoch": 1.1275992438563327,
      "grad_norm": 0.41300724617139584,
      "learning_rate": 7.372619690231534e-05,
      "loss": 0.5654923915863037,
      "step": 1193
    },
    {
      "epoch": 1.1285444234404536,
      "grad_norm": 0.3786401205306675,
      "learning_rate": 7.368032393618662e-05,
      "loss": 0.5425905585289001,
      "step": 1194
    },
    {
      "epoch": 1.1294896030245747,
      "grad_norm": 0.4307953595281302,
      "learning_rate": 7.363442525942826e-05,
      "loss": 0.5892547965049744,
      "step": 1195
    },
    {
      "epoch": 1.1304347826086956,
      "grad_norm": 0.3793555143598211,
      "learning_rate": 7.358850092187424e-05,
      "loss": 0.5525598526000977,
      "step": 1196
    },
    {
      "epoch": 1.1313799621928167,
      "grad_norm": 0.3846240957576434,
      "learning_rate": 7.354255097338634e-05,
      "loss": 0.5356988310813904,
      "step": 1197
    },
    {
      "epoch": 1.1323251417769375,
      "grad_norm": 0.3668195186117827,
      "learning_rate": 7.349657546385413e-05,
      "loss": 0.5182729363441467,
      "step": 1198
    },
    {
      "epoch": 1.1332703213610587,
      "grad_norm": 0.43717137667868444,
      "learning_rate": 7.345057444319505e-05,
      "loss": 0.6351199150085449,
      "step": 1199
    },
    {
      "epoch": 1.1342155009451795,
      "grad_norm": 0.35079552598055086,
      "learning_rate": 7.34045479613541e-05,
      "loss": 0.5227787494659424,
      "step": 1200
    },
    {
      "epoch": 1.1351606805293006,
      "grad_norm": 0.36295296081976824,
      "learning_rate": 7.335849606830401e-05,
      "loss": 0.5436674356460571,
      "step": 1201
    },
    {
      "epoch": 1.1361058601134215,
      "grad_norm": 0.2989614287576837,
      "learning_rate": 7.331241881404507e-05,
      "loss": 0.4210548996925354,
      "step": 1202
    },
    {
      "epoch": 1.1370510396975426,
      "grad_norm": 0.3201887814463189,
      "learning_rate": 7.326631624860509e-05,
      "loss": 0.5177375078201294,
      "step": 1203
    },
    {
      "epoch": 1.1379962192816635,
      "grad_norm": 0.33579330364358506,
      "learning_rate": 7.322018842203941e-05,
      "loss": 0.46858710050582886,
      "step": 1204
    },
    {
      "epoch": 1.1389413988657844,
      "grad_norm": 0.3493382565684786,
      "learning_rate": 7.317403538443078e-05,
      "loss": 0.5482692718505859,
      "step": 1205
    },
    {
      "epoch": 1.1398865784499055,
      "grad_norm": 0.36883899945468446,
      "learning_rate": 7.312785718588927e-05,
      "loss": 0.521681547164917,
      "step": 1206
    },
    {
      "epoch": 1.1408317580340266,
      "grad_norm": 0.36983155669263273,
      "learning_rate": 7.308165387655231e-05,
      "loss": 0.5452646017074585,
      "step": 1207
    },
    {
      "epoch": 1.1417769376181475,
      "grad_norm": 0.40073535488129297,
      "learning_rate": 7.303542550658462e-05,
      "loss": 0.6034570932388306,
      "step": 1208
    },
    {
      "epoch": 1.1427221172022684,
      "grad_norm": 0.4061330548795422,
      "learning_rate": 7.29891721261781e-05,
      "loss": 0.6063923239707947,
      "step": 1209
    },
    {
      "epoch": 1.1436672967863895,
      "grad_norm": 0.35609218091544,
      "learning_rate": 7.294289378555179e-05,
      "loss": 0.546364426612854,
      "step": 1210
    },
    {
      "epoch": 1.1446124763705103,
      "grad_norm": 0.3474593743450535,
      "learning_rate": 7.289659053495186e-05,
      "loss": 0.49704423546791077,
      "step": 1211
    },
    {
      "epoch": 1.1455576559546314,
      "grad_norm": 0.3439488566111658,
      "learning_rate": 7.28502624246515e-05,
      "loss": 0.47920137643814087,
      "step": 1212
    },
    {
      "epoch": 1.1465028355387523,
      "grad_norm": 0.33149375764713795,
      "learning_rate": 7.280390950495093e-05,
      "loss": 0.48399949073791504,
      "step": 1213
    },
    {
      "epoch": 1.1474480151228734,
      "grad_norm": 0.3498981293324763,
      "learning_rate": 7.275753182617723e-05,
      "loss": 0.5116795897483826,
      "step": 1214
    },
    {
      "epoch": 1.1483931947069943,
      "grad_norm": 0.3776389412239106,
      "learning_rate": 7.271112943868448e-05,
      "loss": 0.5556368827819824,
      "step": 1215
    },
    {
      "epoch": 1.1493383742911154,
      "grad_norm": 0.31542741466020513,
      "learning_rate": 7.266470239285347e-05,
      "loss": 0.44255930185317993,
      "step": 1216
    },
    {
      "epoch": 1.1502835538752363,
      "grad_norm": 0.35428942059960766,
      "learning_rate": 7.261825073909182e-05,
      "loss": 0.528002917766571,
      "step": 1217
    },
    {
      "epoch": 1.1512287334593574,
      "grad_norm": 0.3495697053153477,
      "learning_rate": 7.257177452783388e-05,
      "loss": 0.5553412437438965,
      "step": 1218
    },
    {
      "epoch": 1.1521739130434783,
      "grad_norm": 0.3805192004395827,
      "learning_rate": 7.252527380954062e-05,
      "loss": 0.5770350694656372,
      "step": 1219
    },
    {
      "epoch": 1.1531190926275992,
      "grad_norm": 0.3686904347203821,
      "learning_rate": 7.247874863469964e-05,
      "loss": 0.5202410817146301,
      "step": 1220
    },
    {
      "epoch": 1.1540642722117203,
      "grad_norm": 0.4264447961543811,
      "learning_rate": 7.243219905382511e-05,
      "loss": 0.6264290809631348,
      "step": 1221
    },
    {
      "epoch": 1.1550094517958411,
      "grad_norm": 0.38918036656243893,
      "learning_rate": 7.238562511745768e-05,
      "loss": 0.4828234910964966,
      "step": 1222
    },
    {
      "epoch": 1.1559546313799622,
      "grad_norm": 0.33645940200500324,
      "learning_rate": 7.233902687616445e-05,
      "loss": 0.4688301682472229,
      "step": 1223
    },
    {
      "epoch": 1.1568998109640831,
      "grad_norm": 0.3688665320898013,
      "learning_rate": 7.229240438053889e-05,
      "loss": 0.548723578453064,
      "step": 1224
    },
    {
      "epoch": 1.1578449905482042,
      "grad_norm": 0.3847082170664231,
      "learning_rate": 7.224575768120083e-05,
      "loss": 0.5758494734764099,
      "step": 1225
    },
    {
      "epoch": 1.158790170132325,
      "grad_norm": 0.37192928431138145,
      "learning_rate": 7.219908682879637e-05,
      "loss": 0.5438143610954285,
      "step": 1226
    },
    {
      "epoch": 1.1597353497164462,
      "grad_norm": 0.3538460851027304,
      "learning_rate": 7.215239187399782e-05,
      "loss": 0.5124550461769104,
      "step": 1227
    },
    {
      "epoch": 1.160680529300567,
      "grad_norm": 0.36316658978286553,
      "learning_rate": 7.210567286750369e-05,
      "loss": 0.5085729360580444,
      "step": 1228
    },
    {
      "epoch": 1.1616257088846882,
      "grad_norm": 0.33268531763435744,
      "learning_rate": 7.205892986003857e-05,
      "loss": 0.5345269441604614,
      "step": 1229
    },
    {
      "epoch": 1.162570888468809,
      "grad_norm": 0.37064989739992016,
      "learning_rate": 7.201216290235312e-05,
      "loss": 0.5229178667068481,
      "step": 1230
    },
    {
      "epoch": 1.16351606805293,
      "grad_norm": 0.34469346476640794,
      "learning_rate": 7.196537204522401e-05,
      "loss": 0.470956414937973,
      "step": 1231
    },
    {
      "epoch": 1.164461247637051,
      "grad_norm": 0.38302123184190856,
      "learning_rate": 7.191855733945387e-05,
      "loss": 0.5397948622703552,
      "step": 1232
    },
    {
      "epoch": 1.165406427221172,
      "grad_norm": 0.3570313021805296,
      "learning_rate": 7.18717188358712e-05,
      "loss": 0.5354431867599487,
      "step": 1233
    },
    {
      "epoch": 1.166351606805293,
      "grad_norm": 0.41090418879305285,
      "learning_rate": 7.182485658533036e-05,
      "loss": 0.5589978694915771,
      "step": 1234
    },
    {
      "epoch": 1.167296786389414,
      "grad_norm": 0.29953352396131494,
      "learning_rate": 7.177797063871146e-05,
      "loss": 0.3858043849468231,
      "step": 1235
    },
    {
      "epoch": 1.168241965973535,
      "grad_norm": 0.3661554103510045,
      "learning_rate": 7.173106104692033e-05,
      "loss": 0.5312099456787109,
      "step": 1236
    },
    {
      "epoch": 1.169187145557656,
      "grad_norm": 0.3728942706224277,
      "learning_rate": 7.168412786088857e-05,
      "loss": 0.5064578056335449,
      "step": 1237
    },
    {
      "epoch": 1.170132325141777,
      "grad_norm": 0.3586159540659491,
      "learning_rate": 7.163717113157329e-05,
      "loss": 0.534149169921875,
      "step": 1238
    },
    {
      "epoch": 1.1710775047258979,
      "grad_norm": 0.38417957199043984,
      "learning_rate": 7.15901909099572e-05,
      "loss": 0.523589015007019,
      "step": 1239
    },
    {
      "epoch": 1.172022684310019,
      "grad_norm": 0.4054850243462678,
      "learning_rate": 7.154318724704853e-05,
      "loss": 0.6000632643699646,
      "step": 1240
    },
    {
      "epoch": 1.1729678638941399,
      "grad_norm": 0.3695012215996357,
      "learning_rate": 7.149616019388093e-05,
      "loss": 0.5746753215789795,
      "step": 1241
    },
    {
      "epoch": 1.1739130434782608,
      "grad_norm": 0.33635529670437925,
      "learning_rate": 7.144910980151345e-05,
      "loss": 0.4966840445995331,
      "step": 1242
    },
    {
      "epoch": 1.1748582230623819,
      "grad_norm": 0.36327541557645165,
      "learning_rate": 7.140203612103053e-05,
      "loss": 0.5231446027755737,
      "step": 1243
    },
    {
      "epoch": 1.175803402646503,
      "grad_norm": 0.3276134143556871,
      "learning_rate": 7.135493920354184e-05,
      "loss": 0.39921241998672485,
      "step": 1244
    },
    {
      "epoch": 1.1767485822306238,
      "grad_norm": 0.3825946733719089,
      "learning_rate": 7.130781910018227e-05,
      "loss": 0.5199316740036011,
      "step": 1245
    },
    {
      "epoch": 1.1776937618147447,
      "grad_norm": 0.3908617876156442,
      "learning_rate": 7.126067586211198e-05,
      "loss": 0.477749764919281,
      "step": 1246
    },
    {
      "epoch": 1.1786389413988658,
      "grad_norm": 0.35994442034268714,
      "learning_rate": 7.12135095405161e-05,
      "loss": 0.5132631659507751,
      "step": 1247
    },
    {
      "epoch": 1.1795841209829867,
      "grad_norm": 0.3713089948137971,
      "learning_rate": 7.116632018660496e-05,
      "loss": 0.5144317150115967,
      "step": 1248
    },
    {
      "epoch": 1.1805293005671078,
      "grad_norm": 0.3201534803937335,
      "learning_rate": 7.111910785161381e-05,
      "loss": 0.4505634903907776,
      "step": 1249
    },
    {
      "epoch": 1.1814744801512287,
      "grad_norm": 0.36732225047730876,
      "learning_rate": 7.107187258680287e-05,
      "loss": 0.5085432529449463,
      "step": 1250
    },
    {
      "epoch": 1.1824196597353498,
      "grad_norm": 0.39733149567185266,
      "learning_rate": 7.102461444345731e-05,
      "loss": 0.5730095505714417,
      "step": 1251
    },
    {
      "epoch": 1.1833648393194707,
      "grad_norm": 0.3595813407493992,
      "learning_rate": 7.097733347288708e-05,
      "loss": 0.526576042175293,
      "step": 1252
    },
    {
      "epoch": 1.1843100189035918,
      "grad_norm": 0.34673395127841206,
      "learning_rate": 7.093002972642689e-05,
      "loss": 0.41745734214782715,
      "step": 1253
    },
    {
      "epoch": 1.1852551984877127,
      "grad_norm": 0.3437212502833106,
      "learning_rate": 7.088270325543627e-05,
      "loss": 0.4641576409339905,
      "step": 1254
    },
    {
      "epoch": 1.1862003780718338,
      "grad_norm": 0.38091663122735214,
      "learning_rate": 7.083535411129933e-05,
      "loss": 0.5604367852210999,
      "step": 1255
    },
    {
      "epoch": 1.1871455576559546,
      "grad_norm": 0.33960820900742744,
      "learning_rate": 7.078798234542488e-05,
      "loss": 0.46768924593925476,
      "step": 1256
    },
    {
      "epoch": 1.1880907372400755,
      "grad_norm": 0.3727543453703319,
      "learning_rate": 7.074058800924624e-05,
      "loss": 0.5147897005081177,
      "step": 1257
    },
    {
      "epoch": 1.1890359168241966,
      "grad_norm": 0.3968190056412378,
      "learning_rate": 7.06931711542212e-05,
      "loss": 0.5977680683135986,
      "step": 1258
    },
    {
      "epoch": 1.1899810964083175,
      "grad_norm": 0.39897736088139607,
      "learning_rate": 7.064573183183211e-05,
      "loss": 0.6144324541091919,
      "step": 1259
    },
    {
      "epoch": 1.1909262759924386,
      "grad_norm": 0.33459343053194923,
      "learning_rate": 7.059827009358563e-05,
      "loss": 0.49748900532722473,
      "step": 1260
    },
    {
      "epoch": 1.1918714555765595,
      "grad_norm": 0.36501695437942866,
      "learning_rate": 7.055078599101275e-05,
      "loss": 0.5454192161560059,
      "step": 1261
    },
    {
      "epoch": 1.1928166351606806,
      "grad_norm": 0.35127577247546055,
      "learning_rate": 7.050327957566881e-05,
      "loss": 0.4637009799480438,
      "step": 1262
    },
    {
      "epoch": 1.1937618147448015,
      "grad_norm": 0.34831040185762846,
      "learning_rate": 7.04557508991333e-05,
      "loss": 0.491697758436203,
      "step": 1263
    },
    {
      "epoch": 1.1947069943289226,
      "grad_norm": 0.3245220273432368,
      "learning_rate": 7.040820001300992e-05,
      "loss": 0.4935090243816376,
      "step": 1264
    },
    {
      "epoch": 1.1956521739130435,
      "grad_norm": 0.38143244975895885,
      "learning_rate": 7.036062696892648e-05,
      "loss": 0.5432577729225159,
      "step": 1265
    },
    {
      "epoch": 1.1965973534971646,
      "grad_norm": 0.34684580341614557,
      "learning_rate": 7.031303181853486e-05,
      "loss": 0.4969443082809448,
      "step": 1266
    },
    {
      "epoch": 1.1975425330812854,
      "grad_norm": 0.36458083039104133,
      "learning_rate": 7.026541461351092e-05,
      "loss": 0.5072838068008423,
      "step": 1267
    },
    {
      "epoch": 1.1984877126654063,
      "grad_norm": 0.3709333555824172,
      "learning_rate": 7.021777540555447e-05,
      "loss": 0.5708043575286865,
      "step": 1268
    },
    {
      "epoch": 1.1994328922495274,
      "grad_norm": 0.39100917299688587,
      "learning_rate": 7.017011424638922e-05,
      "loss": 0.6133058071136475,
      "step": 1269
    },
    {
      "epoch": 1.2003780718336483,
      "grad_norm": 0.3712747704666411,
      "learning_rate": 7.012243118776269e-05,
      "loss": 0.5000534653663635,
      "step": 1270
    },
    {
      "epoch": 1.2013232514177694,
      "grad_norm": 0.3715126742214629,
      "learning_rate": 7.00747262814462e-05,
      "loss": 0.4528438150882721,
      "step": 1271
    },
    {
      "epoch": 1.2022684310018903,
      "grad_norm": 0.39743884679953645,
      "learning_rate": 7.002699957923478e-05,
      "loss": 0.5365614295005798,
      "step": 1272
    },
    {
      "epoch": 1.2032136105860114,
      "grad_norm": 0.37348583350950193,
      "learning_rate": 6.997925113294716e-05,
      "loss": 0.5510760545730591,
      "step": 1273
    },
    {
      "epoch": 1.2041587901701323,
      "grad_norm": 0.36426163457408917,
      "learning_rate": 6.993148099442561e-05,
      "loss": 0.5367915630340576,
      "step": 1274
    },
    {
      "epoch": 1.2051039697542534,
      "grad_norm": 0.39779050019392587,
      "learning_rate": 6.988368921553601e-05,
      "loss": 0.532168984413147,
      "step": 1275
    },
    {
      "epoch": 1.2060491493383743,
      "grad_norm": 0.367318017287535,
      "learning_rate": 6.983587584816769e-05,
      "loss": 0.5133687257766724,
      "step": 1276
    },
    {
      "epoch": 1.2069943289224954,
      "grad_norm": 0.36670745029110713,
      "learning_rate": 6.978804094423345e-05,
      "loss": 0.5269822478294373,
      "step": 1277
    },
    {
      "epoch": 1.2079395085066162,
      "grad_norm": 0.34396016503642235,
      "learning_rate": 6.97401845556695e-05,
      "loss": 0.5127684473991394,
      "step": 1278
    },
    {
      "epoch": 1.2088846880907371,
      "grad_norm": 0.4212930571488827,
      "learning_rate": 6.96923067344353e-05,
      "loss": 0.5505731701850891,
      "step": 1279
    },
    {
      "epoch": 1.2098298676748582,
      "grad_norm": 0.32247185163532277,
      "learning_rate": 6.964440753251366e-05,
      "loss": 0.42618414759635925,
      "step": 1280
    },
    {
      "epoch": 1.210775047258979,
      "grad_norm": 0.3478157214622607,
      "learning_rate": 6.959648700191055e-05,
      "loss": 0.462557852268219,
      "step": 1281
    },
    {
      "epoch": 1.2117202268431002,
      "grad_norm": 0.36877008918609366,
      "learning_rate": 6.95485451946551e-05,
      "loss": 0.5122299194335938,
      "step": 1282
    },
    {
      "epoch": 1.212665406427221,
      "grad_norm": 0.4079707472114189,
      "learning_rate": 6.95005821627996e-05,
      "loss": 0.6498516201972961,
      "step": 1283
    },
    {
      "epoch": 1.2136105860113422,
      "grad_norm": 0.31712628847768054,
      "learning_rate": 6.94525979584193e-05,
      "loss": 0.43410825729370117,
      "step": 1284
    },
    {
      "epoch": 1.214555765595463,
      "grad_norm": 0.348397837700932,
      "learning_rate": 6.940459263361249e-05,
      "loss": 0.4416144788265228,
      "step": 1285
    },
    {
      "epoch": 1.2155009451795842,
      "grad_norm": 0.4025771257113921,
      "learning_rate": 6.935656624050039e-05,
      "loss": 0.541399359703064,
      "step": 1286
    },
    {
      "epoch": 1.216446124763705,
      "grad_norm": 0.36408066924083465,
      "learning_rate": 6.930851883122707e-05,
      "loss": 0.48609209060668945,
      "step": 1287
    },
    {
      "epoch": 1.2173913043478262,
      "grad_norm": 0.369369819518516,
      "learning_rate": 6.926045045795944e-05,
      "loss": 0.5751364231109619,
      "step": 1288
    },
    {
      "epoch": 1.218336483931947,
      "grad_norm": 0.36065692683844525,
      "learning_rate": 6.921236117288716e-05,
      "loss": 0.5009491443634033,
      "step": 1289
    },
    {
      "epoch": 1.2192816635160681,
      "grad_norm": 0.38891389403973714,
      "learning_rate": 6.91642510282226e-05,
      "loss": 0.5577160120010376,
      "step": 1290
    },
    {
      "epoch": 1.220226843100189,
      "grad_norm": 0.3575885412193187,
      "learning_rate": 6.911612007620078e-05,
      "loss": 0.48055657744407654,
      "step": 1291
    },
    {
      "epoch": 1.2211720226843101,
      "grad_norm": 0.34799202181541805,
      "learning_rate": 6.90679683690793e-05,
      "loss": 0.47260233759880066,
      "step": 1292
    },
    {
      "epoch": 1.222117202268431,
      "grad_norm": 0.41252695876449225,
      "learning_rate": 6.90197959591383e-05,
      "loss": 0.5673738718032837,
      "step": 1293
    },
    {
      "epoch": 1.223062381852552,
      "grad_norm": 0.38098379476265265,
      "learning_rate": 6.897160289868042e-05,
      "loss": 0.4979380667209625,
      "step": 1294
    },
    {
      "epoch": 1.224007561436673,
      "grad_norm": 0.3368543101302703,
      "learning_rate": 6.892338924003067e-05,
      "loss": 0.5273658037185669,
      "step": 1295
    },
    {
      "epoch": 1.2249527410207939,
      "grad_norm": 0.35144349823812326,
      "learning_rate": 6.88751550355365e-05,
      "loss": 0.5396562218666077,
      "step": 1296
    },
    {
      "epoch": 1.225897920604915,
      "grad_norm": 0.35880828965017836,
      "learning_rate": 6.88269003375676e-05,
      "loss": 0.5075440406799316,
      "step": 1297
    },
    {
      "epoch": 1.2268431001890359,
      "grad_norm": 0.3387594402689807,
      "learning_rate": 6.87786251985159e-05,
      "loss": 0.5191459655761719,
      "step": 1298
    },
    {
      "epoch": 1.227788279773157,
      "grad_norm": 0.3577305303212637,
      "learning_rate": 6.873032967079561e-05,
      "loss": 0.5200031399726868,
      "step": 1299
    },
    {
      "epoch": 1.2287334593572778,
      "grad_norm": 0.33480277998700736,
      "learning_rate": 6.868201380684299e-05,
      "loss": 0.521730363368988,
      "step": 1300
    },
    {
      "epoch": 1.229678638941399,
      "grad_norm": 0.3262776946873231,
      "learning_rate": 6.863367765911644e-05,
      "loss": 0.4312354326248169,
      "step": 1301
    },
    {
      "epoch": 1.2306238185255198,
      "grad_norm": 0.34299839268914883,
      "learning_rate": 6.858532128009633e-05,
      "loss": 0.5296234488487244,
      "step": 1302
    },
    {
      "epoch": 1.231568998109641,
      "grad_norm": 0.3376673054828742,
      "learning_rate": 6.853694472228503e-05,
      "loss": 0.515418529510498,
      "step": 1303
    },
    {
      "epoch": 1.2325141776937618,
      "grad_norm": 0.33451021983562523,
      "learning_rate": 6.848854803820683e-05,
      "loss": 0.5076371431350708,
      "step": 1304
    },
    {
      "epoch": 1.2334593572778827,
      "grad_norm": 0.36848153123838695,
      "learning_rate": 6.844013128040782e-05,
      "loss": 0.5906152129173279,
      "step": 1305
    },
    {
      "epoch": 1.2344045368620038,
      "grad_norm": 0.36734486279346873,
      "learning_rate": 6.839169450145595e-05,
      "loss": 0.4506158232688904,
      "step": 1306
    },
    {
      "epoch": 1.2353497164461247,
      "grad_norm": 0.3634538124273742,
      "learning_rate": 6.834323775394088e-05,
      "loss": 0.5308840274810791,
      "step": 1307
    },
    {
      "epoch": 1.2362948960302458,
      "grad_norm": 0.348433177463924,
      "learning_rate": 6.829476109047391e-05,
      "loss": 0.5066437125205994,
      "step": 1308
    },
    {
      "epoch": 1.2372400756143667,
      "grad_norm": 0.3326072024716305,
      "learning_rate": 6.824626456368801e-05,
      "loss": 0.4518665671348572,
      "step": 1309
    },
    {
      "epoch": 1.2381852551984878,
      "grad_norm": 0.3310585383704418,
      "learning_rate": 6.819774822623772e-05,
      "loss": 0.46915119886398315,
      "step": 1310
    },
    {
      "epoch": 1.2391304347826086,
      "grad_norm": 0.3895795783098224,
      "learning_rate": 6.814921213079908e-05,
      "loss": 0.5377604961395264,
      "step": 1311
    },
    {
      "epoch": 1.2400756143667298,
      "grad_norm": 0.36612252695740827,
      "learning_rate": 6.810065633006956e-05,
      "loss": 0.5349949598312378,
      "step": 1312
    },
    {
      "epoch": 1.2410207939508506,
      "grad_norm": 0.34818754774548766,
      "learning_rate": 6.805208087676806e-05,
      "loss": 0.5118746757507324,
      "step": 1313
    },
    {
      "epoch": 1.2419659735349717,
      "grad_norm": 0.3789940490397744,
      "learning_rate": 6.800348582363477e-05,
      "loss": 0.5397946238517761,
      "step": 1314
    },
    {
      "epoch": 1.2429111531190926,
      "grad_norm": 0.3776802325210702,
      "learning_rate": 6.795487122343124e-05,
      "loss": 0.5094753503799438,
      "step": 1315
    },
    {
      "epoch": 1.2438563327032135,
      "grad_norm": 0.42825750897595954,
      "learning_rate": 6.790623712894015e-05,
      "loss": 0.5050976872444153,
      "step": 1316
    },
    {
      "epoch": 1.2448015122873346,
      "grad_norm": 0.3472169432752961,
      "learning_rate": 6.785758359296539e-05,
      "loss": 0.478814035654068,
      "step": 1317
    },
    {
      "epoch": 1.2457466918714555,
      "grad_norm": 0.350481055314269,
      "learning_rate": 6.780891066833199e-05,
      "loss": 0.5066851377487183,
      "step": 1318
    },
    {
      "epoch": 1.2466918714555766,
      "grad_norm": 0.3819256172960927,
      "learning_rate": 6.776021840788596e-05,
      "loss": 0.5227097868919373,
      "step": 1319
    },
    {
      "epoch": 1.2476370510396975,
      "grad_norm": 0.3848995370719036,
      "learning_rate": 6.771150686449436e-05,
      "loss": 0.5394572615623474,
      "step": 1320
    },
    {
      "epoch": 1.2485822306238186,
      "grad_norm": 0.35207750463820514,
      "learning_rate": 6.766277609104518e-05,
      "loss": 0.5152580142021179,
      "step": 1321
    },
    {
      "epoch": 1.2495274102079394,
      "grad_norm": 0.4248474628664681,
      "learning_rate": 6.761402614044725e-05,
      "loss": 0.5678708553314209,
      "step": 1322
    },
    {
      "epoch": 1.2504725897920606,
      "grad_norm": 0.4292892971208418,
      "learning_rate": 6.756525706563026e-05,
      "loss": 0.5143340826034546,
      "step": 1323
    },
    {
      "epoch": 1.2514177693761814,
      "grad_norm": 0.37642382974920946,
      "learning_rate": 6.751646891954467e-05,
      "loss": 0.4825954735279083,
      "step": 1324
    },
    {
      "epoch": 1.2523629489603025,
      "grad_norm": 0.36199649372038495,
      "learning_rate": 6.74676617551616e-05,
      "loss": 0.5108698606491089,
      "step": 1325
    },
    {
      "epoch": 1.2533081285444234,
      "grad_norm": 0.36932576029154973,
      "learning_rate": 6.741883562547287e-05,
      "loss": 0.584618091583252,
      "step": 1326
    },
    {
      "epoch": 1.2542533081285443,
      "grad_norm": 0.3657024306642165,
      "learning_rate": 6.736999058349086e-05,
      "loss": 0.514549732208252,
      "step": 1327
    },
    {
      "epoch": 1.2551984877126654,
      "grad_norm": 0.3690028843510558,
      "learning_rate": 6.732112668224847e-05,
      "loss": 0.5615096688270569,
      "step": 1328
    },
    {
      "epoch": 1.2561436672967865,
      "grad_norm": 0.35549757804584087,
      "learning_rate": 6.727224397479913e-05,
      "loss": 0.538159966468811,
      "step": 1329
    },
    {
      "epoch": 1.2570888468809074,
      "grad_norm": 0.36326631212016436,
      "learning_rate": 6.722334251421665e-05,
      "loss": 0.5376110076904297,
      "step": 1330
    },
    {
      "epoch": 1.2580340264650283,
      "grad_norm": 0.3795655013638314,
      "learning_rate": 6.71744223535952e-05,
      "loss": 0.5330298542976379,
      "step": 1331
    },
    {
      "epoch": 1.2589792060491494,
      "grad_norm": 0.36294654752345074,
      "learning_rate": 6.712548354604929e-05,
      "loss": 0.507486879825592,
      "step": 1332
    },
    {
      "epoch": 1.2599243856332702,
      "grad_norm": 0.3378907870960561,
      "learning_rate": 6.707652614471359e-05,
      "loss": 0.4896286725997925,
      "step": 1333
    },
    {
      "epoch": 1.2608695652173914,
      "grad_norm": 0.3953734488355936,
      "learning_rate": 6.702755020274306e-05,
      "loss": 0.531862735748291,
      "step": 1334
    },
    {
      "epoch": 1.2618147448015122,
      "grad_norm": 0.34409545819398907,
      "learning_rate": 6.697855577331274e-05,
      "loss": 0.4746660888195038,
      "step": 1335
    },
    {
      "epoch": 1.2627599243856333,
      "grad_norm": 0.3433096401896328,
      "learning_rate": 6.692954290961774e-05,
      "loss": 0.5090203285217285,
      "step": 1336
    },
    {
      "epoch": 1.2637051039697542,
      "grad_norm": 0.3509249134347413,
      "learning_rate": 6.68805116648732e-05,
      "loss": 0.5314555168151855,
      "step": 1337
    },
    {
      "epoch": 1.264650283553875,
      "grad_norm": 0.33321972752204965,
      "learning_rate": 6.683146209231421e-05,
      "loss": 0.4640154540538788,
      "step": 1338
    },
    {
      "epoch": 1.2655954631379962,
      "grad_norm": 0.37157623657139927,
      "learning_rate": 6.678239424519575e-05,
      "loss": 0.5795703530311584,
      "step": 1339
    },
    {
      "epoch": 1.2665406427221173,
      "grad_norm": 0.380199869156107,
      "learning_rate": 6.673330817679265e-05,
      "loss": 0.5228967666625977,
      "step": 1340
    },
    {
      "epoch": 1.2674858223062382,
      "grad_norm": 0.33766166627451444,
      "learning_rate": 6.668420394039953e-05,
      "loss": 0.4996538758277893,
      "step": 1341
    },
    {
      "epoch": 1.268431001890359,
      "grad_norm": 0.40738284994242574,
      "learning_rate": 6.663508158933074e-05,
      "loss": 0.5891445875167847,
      "step": 1342
    },
    {
      "epoch": 1.2693761814744802,
      "grad_norm": 0.37522668550292665,
      "learning_rate": 6.658594117692027e-05,
      "loss": 0.5397313237190247,
      "step": 1343
    },
    {
      "epoch": 1.2703213610586013,
      "grad_norm": 0.35629632333077926,
      "learning_rate": 6.653678275652174e-05,
      "loss": 0.5193596482276917,
      "step": 1344
    },
    {
      "epoch": 1.2712665406427222,
      "grad_norm": 0.37421185256846023,
      "learning_rate": 6.648760638150832e-05,
      "loss": 0.5331536531448364,
      "step": 1345
    },
    {
      "epoch": 1.272211720226843,
      "grad_norm": 0.3884124987880379,
      "learning_rate": 6.64384121052727e-05,
      "loss": 0.5021524429321289,
      "step": 1346
    },
    {
      "epoch": 1.2731568998109641,
      "grad_norm": 0.36129664040632287,
      "learning_rate": 6.638919998122694e-05,
      "loss": 0.5348781943321228,
      "step": 1347
    },
    {
      "epoch": 1.274102079395085,
      "grad_norm": 0.3191856671614576,
      "learning_rate": 6.633997006280252e-05,
      "loss": 0.44733619689941406,
      "step": 1348
    },
    {
      "epoch": 1.2750472589792061,
      "grad_norm": 0.3546660322854816,
      "learning_rate": 6.629072240345028e-05,
      "loss": 0.5254180431365967,
      "step": 1349
    },
    {
      "epoch": 1.275992438563327,
      "grad_norm": 0.3633564966495802,
      "learning_rate": 6.624145705664023e-05,
      "loss": 0.49134132266044617,
      "step": 1350
    },
    {
      "epoch": 1.276937618147448,
      "grad_norm": 0.37325096215587644,
      "learning_rate": 6.619217407586167e-05,
      "loss": 0.5368967056274414,
      "step": 1351
    },
    {
      "epoch": 1.277882797731569,
      "grad_norm": 0.3850142129268895,
      "learning_rate": 6.614287351462298e-05,
      "loss": 0.5465425848960876,
      "step": 1352
    },
    {
      "epoch": 1.2788279773156899,
      "grad_norm": 0.4224998840661803,
      "learning_rate": 6.609355542645166e-05,
      "loss": 0.5885909795761108,
      "step": 1353
    },
    {
      "epoch": 1.279773156899811,
      "grad_norm": 0.3887823304649113,
      "learning_rate": 6.604421986489427e-05,
      "loss": 0.5911107063293457,
      "step": 1354
    },
    {
      "epoch": 1.280718336483932,
      "grad_norm": 0.37051713247421236,
      "learning_rate": 6.599486688351628e-05,
      "loss": 0.5506333112716675,
      "step": 1355
    },
    {
      "epoch": 1.281663516068053,
      "grad_norm": 0.37767868456678133,
      "learning_rate": 6.594549653590211e-05,
      "loss": 0.5080118179321289,
      "step": 1356
    },
    {
      "epoch": 1.2826086956521738,
      "grad_norm": 0.39541737176992914,
      "learning_rate": 6.589610887565503e-05,
      "loss": 0.5952708721160889,
      "step": 1357
    },
    {
      "epoch": 1.283553875236295,
      "grad_norm": 0.3278697048293944,
      "learning_rate": 6.584670395639707e-05,
      "loss": 0.4589739441871643,
      "step": 1358
    },
    {
      "epoch": 1.2844990548204158,
      "grad_norm": 0.3570201506858837,
      "learning_rate": 6.579728183176907e-05,
      "loss": 0.489423930644989,
      "step": 1359
    },
    {
      "epoch": 1.285444234404537,
      "grad_norm": 0.38474042104482076,
      "learning_rate": 6.574784255543051e-05,
      "loss": 0.5101546049118042,
      "step": 1360
    },
    {
      "epoch": 1.2863894139886578,
      "grad_norm": 0.37656548583645344,
      "learning_rate": 6.56983861810595e-05,
      "loss": 0.5406765937805176,
      "step": 1361
    },
    {
      "epoch": 1.287334593572779,
      "grad_norm": 0.37328016340874315,
      "learning_rate": 6.564891276235269e-05,
      "loss": 0.5355812311172485,
      "step": 1362
    },
    {
      "epoch": 1.2882797731568998,
      "grad_norm": 0.3616487187301431,
      "learning_rate": 6.559942235302527e-05,
      "loss": 0.5210821628570557,
      "step": 1363
    },
    {
      "epoch": 1.2892249527410207,
      "grad_norm": 0.3904151851127228,
      "learning_rate": 6.554991500681083e-05,
      "loss": 0.49844199419021606,
      "step": 1364
    },
    {
      "epoch": 1.2901701323251418,
      "grad_norm": 0.3347303889028326,
      "learning_rate": 6.550039077746142e-05,
      "loss": 0.5153278112411499,
      "step": 1365
    },
    {
      "epoch": 1.2911153119092629,
      "grad_norm": 0.35035782580639263,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.5226248502731323,
      "step": 1366
    },
    {
      "epoch": 1.2920604914933838,
      "grad_norm": 0.4019771608404968,
      "learning_rate": 6.540129188445731e-05,
      "loss": 0.6150473356246948,
      "step": 1367
    },
    {
      "epoch": 1.2930056710775046,
      "grad_norm": 0.3795007885688876,
      "learning_rate": 6.535171732839803e-05,
      "loss": 0.5072449445724487,
      "step": 1368
    },
    {
      "epoch": 1.2939508506616257,
      "grad_norm": 0.3574247962861576,
      "learning_rate": 6.530212610439455e-05,
      "loss": 0.4794924259185791,
      "step": 1369
    },
    {
      "epoch": 1.2948960302457466,
      "grad_norm": 0.3263450307967503,
      "learning_rate": 6.525251826628991e-05,
      "loss": 0.4345635771751404,
      "step": 1370
    },
    {
      "epoch": 1.2958412098298677,
      "grad_norm": 0.41402234521219217,
      "learning_rate": 6.520289386794528e-05,
      "loss": 0.4994678199291229,
      "step": 1371
    },
    {
      "epoch": 1.2967863894139886,
      "grad_norm": 0.38494336979634874,
      "learning_rate": 6.515325296323973e-05,
      "loss": 0.5633077025413513,
      "step": 1372
    },
    {
      "epoch": 1.2977315689981097,
      "grad_norm": 0.36999208017261054,
      "learning_rate": 6.510359560607027e-05,
      "loss": 0.522591233253479,
      "step": 1373
    },
    {
      "epoch": 1.2986767485822306,
      "grad_norm": 0.37603837117019584,
      "learning_rate": 6.505392185035183e-05,
      "loss": 0.5477960109710693,
      "step": 1374
    },
    {
      "epoch": 1.2996219281663515,
      "grad_norm": 0.3538758606754946,
      "learning_rate": 6.500423175001705e-05,
      "loss": 0.5385228395462036,
      "step": 1375
    },
    {
      "epoch": 1.3005671077504726,
      "grad_norm": 0.37743166355060515,
      "learning_rate": 6.495452535901637e-05,
      "loss": 0.5859216451644897,
      "step": 1376
    },
    {
      "epoch": 1.3015122873345937,
      "grad_norm": 0.41030505641462517,
      "learning_rate": 6.490480273131794e-05,
      "loss": 0.529114305973053,
      "step": 1377
    },
    {
      "epoch": 1.3024574669187146,
      "grad_norm": 0.3507775994567242,
      "learning_rate": 6.485506392090748e-05,
      "loss": 0.4469347596168518,
      "step": 1378
    },
    {
      "epoch": 1.3034026465028354,
      "grad_norm": 0.4039078113409359,
      "learning_rate": 6.480530898178836e-05,
      "loss": 0.5543379783630371,
      "step": 1379
    },
    {
      "epoch": 1.3043478260869565,
      "grad_norm": 0.36361728948682437,
      "learning_rate": 6.475553796798135e-05,
      "loss": 0.45133811235427856,
      "step": 1380
    },
    {
      "epoch": 1.3052930056710774,
      "grad_norm": 0.38847362082900083,
      "learning_rate": 6.470575093352477e-05,
      "loss": 0.5170056819915771,
      "step": 1381
    },
    {
      "epoch": 1.3062381852551985,
      "grad_norm": 0.38152698485829906,
      "learning_rate": 6.465594793247429e-05,
      "loss": 0.5809912085533142,
      "step": 1382
    },
    {
      "epoch": 1.3071833648393194,
      "grad_norm": 0.36421665289818117,
      "learning_rate": 6.460612901890294e-05,
      "loss": 0.5019212961196899,
      "step": 1383
    },
    {
      "epoch": 1.3081285444234405,
      "grad_norm": 0.3594739475840755,
      "learning_rate": 6.4556294246901e-05,
      "loss": 0.45289549231529236,
      "step": 1384
    },
    {
      "epoch": 1.3090737240075614,
      "grad_norm": 0.34415087135754857,
      "learning_rate": 6.450644367057597e-05,
      "loss": 0.49870091676712036,
      "step": 1385
    },
    {
      "epoch": 1.3100189035916825,
      "grad_norm": 0.3914478178848421,
      "learning_rate": 6.445657734405252e-05,
      "loss": 0.5358381271362305,
      "step": 1386
    },
    {
      "epoch": 1.3109640831758034,
      "grad_norm": 0.36796792856658933,
      "learning_rate": 6.440669532147246e-05,
      "loss": 0.5357275009155273,
      "step": 1387
    },
    {
      "epoch": 1.3119092627599245,
      "grad_norm": 0.3945499855072847,
      "learning_rate": 6.435679765699455e-05,
      "loss": 0.5448852181434631,
      "step": 1388
    },
    {
      "epoch": 1.3128544423440454,
      "grad_norm": 0.38641997137054923,
      "learning_rate": 6.430688440479461e-05,
      "loss": 0.49601098895072937,
      "step": 1389
    },
    {
      "epoch": 1.3137996219281662,
      "grad_norm": 0.3952162674207108,
      "learning_rate": 6.425695561906537e-05,
      "loss": 0.5643264055252075,
      "step": 1390
    },
    {
      "epoch": 1.3147448015122873,
      "grad_norm": 0.38001771860300576,
      "learning_rate": 6.420701135401639e-05,
      "loss": 0.49873852729797363,
      "step": 1391
    },
    {
      "epoch": 1.3156899810964084,
      "grad_norm": 0.43318694431783655,
      "learning_rate": 6.415705166387405e-05,
      "loss": 0.6077578067779541,
      "step": 1392
    },
    {
      "epoch": 1.3166351606805293,
      "grad_norm": 0.3745451235376059,
      "learning_rate": 6.410707660288155e-05,
      "loss": 0.6014819145202637,
      "step": 1393
    },
    {
      "epoch": 1.3175803402646502,
      "grad_norm": 0.38152607632699503,
      "learning_rate": 6.405708622529864e-05,
      "loss": 0.5580853223800659,
      "step": 1394
    },
    {
      "epoch": 1.3185255198487713,
      "grad_norm": 0.3542518980968763,
      "learning_rate": 6.400708058540182e-05,
      "loss": 0.5102869868278503,
      "step": 1395
    },
    {
      "epoch": 1.3194706994328922,
      "grad_norm": 0.3548676741094233,
      "learning_rate": 6.395705973748414e-05,
      "loss": 0.5329160690307617,
      "step": 1396
    },
    {
      "epoch": 1.3204158790170133,
      "grad_norm": 0.36521766146199314,
      "learning_rate": 6.390702373585509e-05,
      "loss": 0.5053244829177856,
      "step": 1397
    },
    {
      "epoch": 1.3213610586011342,
      "grad_norm": 0.36223412812531847,
      "learning_rate": 6.385697263484068e-05,
      "loss": 0.5038160681724548,
      "step": 1398
    },
    {
      "epoch": 1.3223062381852553,
      "grad_norm": 0.38039168714639815,
      "learning_rate": 6.380690648878331e-05,
      "loss": 0.5455756187438965,
      "step": 1399
    },
    {
      "epoch": 1.3232514177693762,
      "grad_norm": 0.3481165297027521,
      "learning_rate": 6.375682535204167e-05,
      "loss": 0.5413278341293335,
      "step": 1400
    },
    {
      "epoch": 1.324196597353497,
      "grad_norm": 0.38743350873080407,
      "learning_rate": 6.37067292789908e-05,
      "loss": 0.5606579184532166,
      "step": 1401
    },
    {
      "epoch": 1.3251417769376181,
      "grad_norm": 0.3517478765537083,
      "learning_rate": 6.36566183240219e-05,
      "loss": 0.5440981984138489,
      "step": 1402
    },
    {
      "epoch": 1.3260869565217392,
      "grad_norm": 0.33053996505523836,
      "learning_rate": 6.360649254154232e-05,
      "loss": 0.4782882332801819,
      "step": 1403
    },
    {
      "epoch": 1.3270321361058601,
      "grad_norm": 0.37303959505754725,
      "learning_rate": 6.355635198597556e-05,
      "loss": 0.5899040699005127,
      "step": 1404
    },
    {
      "epoch": 1.327977315689981,
      "grad_norm": 0.3566527456597587,
      "learning_rate": 6.350619671176111e-05,
      "loss": 0.49342551827430725,
      "step": 1405
    },
    {
      "epoch": 1.3289224952741021,
      "grad_norm": 0.33728777001433624,
      "learning_rate": 6.345602677335447e-05,
      "loss": 0.49958088994026184,
      "step": 1406
    },
    {
      "epoch": 1.329867674858223,
      "grad_norm": 0.3604274006513613,
      "learning_rate": 6.340584222522707e-05,
      "loss": 0.49888572096824646,
      "step": 1407
    },
    {
      "epoch": 1.330812854442344,
      "grad_norm": 0.36101448501633693,
      "learning_rate": 6.335564312186615e-05,
      "loss": 0.49555736780166626,
      "step": 1408
    },
    {
      "epoch": 1.331758034026465,
      "grad_norm": 0.34560690804513616,
      "learning_rate": 6.330542951777481e-05,
      "loss": 0.39755070209503174,
      "step": 1409
    },
    {
      "epoch": 1.332703213610586,
      "grad_norm": 0.3669859188584307,
      "learning_rate": 6.325520146747189e-05,
      "loss": 0.4894663989543915,
      "step": 1410
    },
    {
      "epoch": 1.333648393194707,
      "grad_norm": 0.3500124771421591,
      "learning_rate": 6.320495902549183e-05,
      "loss": 0.5033903121948242,
      "step": 1411
    },
    {
      "epoch": 1.3345935727788278,
      "grad_norm": 0.384230993467121,
      "learning_rate": 6.315470224638485e-05,
      "loss": 0.513495683670044,
      "step": 1412
    },
    {
      "epoch": 1.335538752362949,
      "grad_norm": 0.3706104968266183,
      "learning_rate": 6.31044311847166e-05,
      "loss": 0.4810448884963989,
      "step": 1413
    },
    {
      "epoch": 1.33648393194707,
      "grad_norm": 0.3892749005323439,
      "learning_rate": 6.30541458950683e-05,
      "loss": 0.5165586471557617,
      "step": 1414
    },
    {
      "epoch": 1.337429111531191,
      "grad_norm": 0.343793985439352,
      "learning_rate": 6.30038464320366e-05,
      "loss": 0.4765288233757019,
      "step": 1415
    },
    {
      "epoch": 1.3383742911153118,
      "grad_norm": 0.4007255990250447,
      "learning_rate": 6.295353285023353e-05,
      "loss": 0.5688680410385132,
      "step": 1416
    },
    {
      "epoch": 1.339319470699433,
      "grad_norm": 0.38124179363211286,
      "learning_rate": 6.290320520428647e-05,
      "loss": 0.5915979146957397,
      "step": 1417
    },
    {
      "epoch": 1.3402646502835538,
      "grad_norm": 0.3646323498943536,
      "learning_rate": 6.285286354883808e-05,
      "loss": 0.5281754732131958,
      "step": 1418
    },
    {
      "epoch": 1.341209829867675,
      "grad_norm": 0.3368499651586634,
      "learning_rate": 6.280250793854619e-05,
      "loss": 0.5128151178359985,
      "step": 1419
    },
    {
      "epoch": 1.3421550094517958,
      "grad_norm": 0.36018951511532094,
      "learning_rate": 6.275213842808383e-05,
      "loss": 0.6080739498138428,
      "step": 1420
    },
    {
      "epoch": 1.3431001890359169,
      "grad_norm": 0.355066885323579,
      "learning_rate": 6.270175507213907e-05,
      "loss": 0.5116724967956543,
      "step": 1421
    },
    {
      "epoch": 1.3440453686200378,
      "grad_norm": 0.3503468916769584,
      "learning_rate": 6.265135792541504e-05,
      "loss": 0.5142219066619873,
      "step": 1422
    },
    {
      "epoch": 1.3449905482041589,
      "grad_norm": 0.3770532263057024,
      "learning_rate": 6.260094704262985e-05,
      "loss": 0.603722333908081,
      "step": 1423
    },
    {
      "epoch": 1.3459357277882797,
      "grad_norm": 0.3878106322320532,
      "learning_rate": 6.255052247851652e-05,
      "loss": 0.5292474031448364,
      "step": 1424
    },
    {
      "epoch": 1.3468809073724008,
      "grad_norm": 0.3825988152879323,
      "learning_rate": 6.250008428782292e-05,
      "loss": 0.5764881372451782,
      "step": 1425
    },
    {
      "epoch": 1.3478260869565217,
      "grad_norm": 0.3640234716586332,
      "learning_rate": 6.244963252531171e-05,
      "loss": 0.5000973343849182,
      "step": 1426
    },
    {
      "epoch": 1.3487712665406426,
      "grad_norm": 0.3821270030828565,
      "learning_rate": 6.239916724576029e-05,
      "loss": 0.5560471415519714,
      "step": 1427
    },
    {
      "epoch": 1.3497164461247637,
      "grad_norm": 0.378330494725968,
      "learning_rate": 6.234868850396072e-05,
      "loss": 0.5698621273040771,
      "step": 1428
    },
    {
      "epoch": 1.3506616257088848,
      "grad_norm": 0.377808306427473,
      "learning_rate": 6.229819635471972e-05,
      "loss": 0.534898042678833,
      "step": 1429
    },
    {
      "epoch": 1.3516068052930057,
      "grad_norm": 0.3610642873731797,
      "learning_rate": 6.224769085285854e-05,
      "loss": 0.5291445255279541,
      "step": 1430
    },
    {
      "epoch": 1.3525519848771266,
      "grad_norm": 0.3501638884448915,
      "learning_rate": 6.219717205321289e-05,
      "loss": 0.4664095640182495,
      "step": 1431
    },
    {
      "epoch": 1.3534971644612477,
      "grad_norm": 0.36598493954226874,
      "learning_rate": 6.214664001063299e-05,
      "loss": 0.515373945236206,
      "step": 1432
    },
    {
      "epoch": 1.3544423440453686,
      "grad_norm": 0.4103691521935304,
      "learning_rate": 6.209609477998338e-05,
      "loss": 0.5261394381523132,
      "step": 1433
    },
    {
      "epoch": 1.3553875236294897,
      "grad_norm": 0.3925827671315023,
      "learning_rate": 6.204553641614298e-05,
      "loss": 0.5285322666168213,
      "step": 1434
    },
    {
      "epoch": 1.3563327032136105,
      "grad_norm": 0.45114324990889704,
      "learning_rate": 6.19949649740049e-05,
      "loss": 0.5447168946266174,
      "step": 1435
    },
    {
      "epoch": 1.3572778827977316,
      "grad_norm": 0.3877082000046806,
      "learning_rate": 6.194438050847647e-05,
      "loss": 0.5598663687705994,
      "step": 1436
    },
    {
      "epoch": 1.3582230623818525,
      "grad_norm": 0.37284918974617653,
      "learning_rate": 6.189378307447921e-05,
      "loss": 0.5179684162139893,
      "step": 1437
    },
    {
      "epoch": 1.3591682419659734,
      "grad_norm": 0.42060233798075075,
      "learning_rate": 6.184317272694867e-05,
      "loss": 0.5087293386459351,
      "step": 1438
    },
    {
      "epoch": 1.3601134215500945,
      "grad_norm": 0.33929823759137206,
      "learning_rate": 6.179254952083437e-05,
      "loss": 0.4907689690589905,
      "step": 1439
    },
    {
      "epoch": 1.3610586011342156,
      "grad_norm": 0.37980781840869304,
      "learning_rate": 6.174191351109995e-05,
      "loss": 0.573150634765625,
      "step": 1440
    },
    {
      "epoch": 1.3620037807183365,
      "grad_norm": 0.36583604634443867,
      "learning_rate": 6.16912647527228e-05,
      "loss": 0.5390379428863525,
      "step": 1441
    },
    {
      "epoch": 1.3629489603024574,
      "grad_norm": 0.4015459129420046,
      "learning_rate": 6.16406033006942e-05,
      "loss": 0.5571323037147522,
      "step": 1442
    },
    {
      "epoch": 1.3638941398865785,
      "grad_norm": 0.3370700941542938,
      "learning_rate": 6.158992921001924e-05,
      "loss": 0.4930906891822815,
      "step": 1443
    },
    {
      "epoch": 1.3648393194706994,
      "grad_norm": 0.4082324653254832,
      "learning_rate": 6.15392425357167e-05,
      "loss": 0.5564974546432495,
      "step": 1444
    },
    {
      "epoch": 1.3657844990548205,
      "grad_norm": 0.3417416347892518,
      "learning_rate": 6.148854333281905e-05,
      "loss": 0.4766826331615448,
      "step": 1445
    },
    {
      "epoch": 1.3667296786389413,
      "grad_norm": 0.3344018350716998,
      "learning_rate": 6.143783165637232e-05,
      "loss": 0.49491623044013977,
      "step": 1446
    },
    {
      "epoch": 1.3676748582230625,
      "grad_norm": 0.41377498799343687,
      "learning_rate": 6.138710756143613e-05,
      "loss": 0.5627220869064331,
      "step": 1447
    },
    {
      "epoch": 1.3686200378071833,
      "grad_norm": 0.3764756048015119,
      "learning_rate": 6.133637110308355e-05,
      "loss": 0.5323675274848938,
      "step": 1448
    },
    {
      "epoch": 1.3695652173913042,
      "grad_norm": 0.3669107892193085,
      "learning_rate": 6.128562233640109e-05,
      "loss": 0.5792803764343262,
      "step": 1449
    },
    {
      "epoch": 1.3705103969754253,
      "grad_norm": 0.3882835286904432,
      "learning_rate": 6.12348613164886e-05,
      "loss": 0.5156582593917847,
      "step": 1450
    },
    {
      "epoch": 1.3714555765595464,
      "grad_norm": 0.3562400868999806,
      "learning_rate": 6.118408809845928e-05,
      "loss": 0.4862077236175537,
      "step": 1451
    },
    {
      "epoch": 1.3724007561436673,
      "grad_norm": 0.3412167481270621,
      "learning_rate": 6.113330273743953e-05,
      "loss": 0.5006967186927795,
      "step": 1452
    },
    {
      "epoch": 1.3733459357277882,
      "grad_norm": 0.40018722861040984,
      "learning_rate": 6.108250528856894e-05,
      "loss": 0.5665547847747803,
      "step": 1453
    },
    {
      "epoch": 1.3742911153119093,
      "grad_norm": 0.34043981774433646,
      "learning_rate": 6.1031695807000286e-05,
      "loss": 0.5277857780456543,
      "step": 1454
    },
    {
      "epoch": 1.3752362948960302,
      "grad_norm": 0.380719528429036,
      "learning_rate": 6.098087434789931e-05,
      "loss": 0.5163791179656982,
      "step": 1455
    },
    {
      "epoch": 1.3761814744801513,
      "grad_norm": 0.3720539070896573,
      "learning_rate": 6.093004096644481e-05,
      "loss": 0.5508084297180176,
      "step": 1456
    },
    {
      "epoch": 1.3771266540642721,
      "grad_norm": 0.3759527008361178,
      "learning_rate": 6.087919571782854e-05,
      "loss": 0.5226770043373108,
      "step": 1457
    },
    {
      "epoch": 1.3780718336483933,
      "grad_norm": 0.3622953388661221,
      "learning_rate": 6.0828338657255126e-05,
      "loss": 0.548148512840271,
      "step": 1458
    },
    {
      "epoch": 1.3790170132325141,
      "grad_norm": 0.3642087135683609,
      "learning_rate": 6.077746983994198e-05,
      "loss": 0.5013360381126404,
      "step": 1459
    },
    {
      "epoch": 1.3799621928166352,
      "grad_norm": 0.390225601657353,
      "learning_rate": 6.0726589321119364e-05,
      "loss": 0.587293267250061,
      "step": 1460
    },
    {
      "epoch": 1.3809073724007561,
      "grad_norm": 0.4270557557453895,
      "learning_rate": 6.0675697156030155e-05,
      "loss": 0.5266384482383728,
      "step": 1461
    },
    {
      "epoch": 1.3818525519848772,
      "grad_norm": 0.3539521161273747,
      "learning_rate": 6.062479339992994e-05,
      "loss": 0.4564582109451294,
      "step": 1462
    },
    {
      "epoch": 1.382797731568998,
      "grad_norm": 0.39004998135180663,
      "learning_rate": 6.057387810808686e-05,
      "loss": 0.5292969942092896,
      "step": 1463
    },
    {
      "epoch": 1.383742911153119,
      "grad_norm": 0.3609272560722653,
      "learning_rate": 6.052295133578156e-05,
      "loss": 0.5628476142883301,
      "step": 1464
    },
    {
      "epoch": 1.38468809073724,
      "grad_norm": 0.38929698717213956,
      "learning_rate": 6.0472013138307235e-05,
      "loss": 0.5717183351516724,
      "step": 1465
    },
    {
      "epoch": 1.3856332703213612,
      "grad_norm": 0.39851122442446857,
      "learning_rate": 6.0421063570969374e-05,
      "loss": 0.630679726600647,
      "step": 1466
    },
    {
      "epoch": 1.386578449905482,
      "grad_norm": 0.3604147400750646,
      "learning_rate": 6.037010268908587e-05,
      "loss": 0.5361441373825073,
      "step": 1467
    },
    {
      "epoch": 1.387523629489603,
      "grad_norm": 0.4361882923396139,
      "learning_rate": 6.0319130547986913e-05,
      "loss": 0.5684278011322021,
      "step": 1468
    },
    {
      "epoch": 1.388468809073724,
      "grad_norm": 0.43096467268312477,
      "learning_rate": 6.0268147203014893e-05,
      "loss": 0.5437805652618408,
      "step": 1469
    },
    {
      "epoch": 1.389413988657845,
      "grad_norm": 0.3975960708837217,
      "learning_rate": 6.021715270952435e-05,
      "loss": 0.5170692801475525,
      "step": 1470
    },
    {
      "epoch": 1.390359168241966,
      "grad_norm": 0.3252667078647034,
      "learning_rate": 6.016614712288198e-05,
      "loss": 0.44107335805892944,
      "step": 1471
    },
    {
      "epoch": 1.391304347826087,
      "grad_norm": 0.3665883394993792,
      "learning_rate": 6.011513049846647e-05,
      "loss": 0.48255544900894165,
      "step": 1472
    },
    {
      "epoch": 1.392249527410208,
      "grad_norm": 0.3672283438134946,
      "learning_rate": 6.0064102891668506e-05,
      "loss": 0.5680907964706421,
      "step": 1473
    },
    {
      "epoch": 1.393194706994329,
      "grad_norm": 0.3579769450228483,
      "learning_rate": 6.001306435789072e-05,
      "loss": 0.5496354103088379,
      "step": 1474
    },
    {
      "epoch": 1.3941398865784498,
      "grad_norm": 0.34921638617412176,
      "learning_rate": 5.9962014952547575e-05,
      "loss": 0.48663556575775146,
      "step": 1475
    },
    {
      "epoch": 1.3950850661625709,
      "grad_norm": 0.37896685056673113,
      "learning_rate": 5.9910954731065386e-05,
      "loss": 0.5633080005645752,
      "step": 1476
    },
    {
      "epoch": 1.396030245746692,
      "grad_norm": 0.36108727106782823,
      "learning_rate": 5.9859883748882164e-05,
      "loss": 0.45966970920562744,
      "step": 1477
    },
    {
      "epoch": 1.3969754253308129,
      "grad_norm": 0.40288869102642444,
      "learning_rate": 5.9808802061447586e-05,
      "loss": 0.490816593170166,
      "step": 1478
    },
    {
      "epoch": 1.3979206049149338,
      "grad_norm": 0.3837303364739356,
      "learning_rate": 5.975770972422303e-05,
      "loss": 0.5582623481750488,
      "step": 1479
    },
    {
      "epoch": 1.3988657844990549,
      "grad_norm": 0.39878721576473036,
      "learning_rate": 5.970660679268138e-05,
      "loss": 0.5131020545959473,
      "step": 1480
    },
    {
      "epoch": 1.3998109640831757,
      "grad_norm": 0.3799299054966021,
      "learning_rate": 5.9655493322307033e-05,
      "loss": 0.5510779619216919,
      "step": 1481
    },
    {
      "epoch": 1.4007561436672968,
      "grad_norm": 0.38708531906837895,
      "learning_rate": 5.9604369368595836e-05,
      "loss": 0.5040088891983032,
      "step": 1482
    },
    {
      "epoch": 1.4017013232514177,
      "grad_norm": 0.3624625800245018,
      "learning_rate": 5.9553234987055006e-05,
      "loss": 0.5210117101669312,
      "step": 1483
    },
    {
      "epoch": 1.4026465028355388,
      "grad_norm": 0.4169178304256326,
      "learning_rate": 5.950209023320308e-05,
      "loss": 0.5446616411209106,
      "step": 1484
    },
    {
      "epoch": 1.4035916824196597,
      "grad_norm": 0.3775682373785461,
      "learning_rate": 5.945093516256989e-05,
      "loss": 0.47594398260116577,
      "step": 1485
    },
    {
      "epoch": 1.4045368620037806,
      "grad_norm": 0.358432832256488,
      "learning_rate": 5.939976983069641e-05,
      "loss": 0.48224392533302307,
      "step": 1486
    },
    {
      "epoch": 1.4054820415879017,
      "grad_norm": 0.38257917526570745,
      "learning_rate": 5.934859429313482e-05,
      "loss": 0.519190788269043,
      "step": 1487
    },
    {
      "epoch": 1.4064272211720228,
      "grad_norm": 0.49421745804648193,
      "learning_rate": 5.9297408605448344e-05,
      "loss": 0.6848883628845215,
      "step": 1488
    },
    {
      "epoch": 1.4073724007561437,
      "grad_norm": 0.3673742163995593,
      "learning_rate": 5.9246212823211225e-05,
      "loss": 0.5137803554534912,
      "step": 1489
    },
    {
      "epoch": 1.4083175803402646,
      "grad_norm": 0.3280066009394793,
      "learning_rate": 5.9195007002008685e-05,
      "loss": 0.43832486867904663,
      "step": 1490
    },
    {
      "epoch": 1.4092627599243857,
      "grad_norm": 0.33266172320648296,
      "learning_rate": 5.914379119743682e-05,
      "loss": 0.4606127142906189,
      "step": 1491
    },
    {
      "epoch": 1.4102079395085065,
      "grad_norm": 0.3926673289538989,
      "learning_rate": 5.909256546510257e-05,
      "loss": 0.5612396001815796,
      "step": 1492
    },
    {
      "epoch": 1.4111531190926276,
      "grad_norm": 0.34563100455772416,
      "learning_rate": 5.90413298606237e-05,
      "loss": 0.49975210428237915,
      "step": 1493
    },
    {
      "epoch": 1.4120982986767485,
      "grad_norm": 0.3738127760821296,
      "learning_rate": 5.899008443962863e-05,
      "loss": 0.502989649772644,
      "step": 1494
    },
    {
      "epoch": 1.4130434782608696,
      "grad_norm": 0.40030752258084235,
      "learning_rate": 5.893882925775648e-05,
      "loss": 0.5804603099822998,
      "step": 1495
    },
    {
      "epoch": 1.4139886578449905,
      "grad_norm": 0.3640311890760167,
      "learning_rate": 5.888756437065693e-05,
      "loss": 0.49941739439964294,
      "step": 1496
    },
    {
      "epoch": 1.4149338374291116,
      "grad_norm": 0.35837304846204326,
      "learning_rate": 5.883628983399025e-05,
      "loss": 0.5472809076309204,
      "step": 1497
    },
    {
      "epoch": 1.4158790170132325,
      "grad_norm": 0.3389977764833276,
      "learning_rate": 5.878500570342713e-05,
      "loss": 0.5484374761581421,
      "step": 1498
    },
    {
      "epoch": 1.4168241965973536,
      "grad_norm": 0.3964398725336757,
      "learning_rate": 5.8733712034648726e-05,
      "loss": 0.5686905384063721,
      "step": 1499
    },
    {
      "epoch": 1.4177693761814745,
      "grad_norm": 0.36459873289052813,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.5438361167907715,
      "step": 1500
    },
    {
      "epoch": 1.4187145557655954,
      "grad_norm": 0.369261499702369,
      "learning_rate": 5.86310963052223e-05,
      "loss": 0.5158875584602356,
      "step": 1501
    },
    {
      "epoch": 1.4196597353497165,
      "grad_norm": 0.37105676306942237,
      "learning_rate": 5.857977435598808e-05,
      "loss": 0.5340439677238464,
      "step": 1502
    },
    {
      "epoch": 1.4206049149338376,
      "grad_norm": 0.3514299090284784,
      "learning_rate": 5.852844309136605e-05,
      "loss": 0.5100448131561279,
      "step": 1503
    },
    {
      "epoch": 1.4215500945179584,
      "grad_norm": 0.37964527698400136,
      "learning_rate": 5.847710256708854e-05,
      "loss": 0.4616706967353821,
      "step": 1504
    },
    {
      "epoch": 1.4224952741020793,
      "grad_norm": 0.3723587220681798,
      "learning_rate": 5.842575283889789e-05,
      "loss": 0.4944632053375244,
      "step": 1505
    },
    {
      "epoch": 1.4234404536862004,
      "grad_norm": 0.33905898512327165,
      "learning_rate": 5.837439396254647e-05,
      "loss": 0.4463622570037842,
      "step": 1506
    },
    {
      "epoch": 1.4243856332703213,
      "grad_norm": 0.34106762637894594,
      "learning_rate": 5.832302599379656e-05,
      "loss": 0.47003352642059326,
      "step": 1507
    },
    {
      "epoch": 1.4253308128544424,
      "grad_norm": 0.3615248436862058,
      "learning_rate": 5.827164898842032e-05,
      "loss": 0.5119285583496094,
      "step": 1508
    },
    {
      "epoch": 1.4262759924385633,
      "grad_norm": 0.4151833783380097,
      "learning_rate": 5.822026300219974e-05,
      "loss": 0.5722465515136719,
      "step": 1509
    },
    {
      "epoch": 1.4272211720226844,
      "grad_norm": 0.3858253804382938,
      "learning_rate": 5.816886809092651e-05,
      "loss": 0.5191408395767212,
      "step": 1510
    },
    {
      "epoch": 1.4281663516068053,
      "grad_norm": 0.3996551649627464,
      "learning_rate": 5.8117464310402056e-05,
      "loss": 0.5113227963447571,
      "step": 1511
    },
    {
      "epoch": 1.4291115311909262,
      "grad_norm": 0.3400391992132418,
      "learning_rate": 5.8066051716437464e-05,
      "loss": 0.43543583154678345,
      "step": 1512
    },
    {
      "epoch": 1.4300567107750473,
      "grad_norm": 0.36877386988742444,
      "learning_rate": 5.8014630364853276e-05,
      "loss": 0.4979449212551117,
      "step": 1513
    },
    {
      "epoch": 1.4310018903591684,
      "grad_norm": 0.37701602200579176,
      "learning_rate": 5.796320031147964e-05,
      "loss": 0.531485915184021,
      "step": 1514
    },
    {
      "epoch": 1.4319470699432892,
      "grad_norm": 0.3528240983525812,
      "learning_rate": 5.7911761612156135e-05,
      "loss": 0.5036394596099854,
      "step": 1515
    },
    {
      "epoch": 1.4328922495274101,
      "grad_norm": 0.3606947721014903,
      "learning_rate": 5.78603143227317e-05,
      "loss": 0.4864853620529175,
      "step": 1516
    },
    {
      "epoch": 1.4338374291115312,
      "grad_norm": 0.3825259655218611,
      "learning_rate": 5.780885849906462e-05,
      "loss": 0.5651298761367798,
      "step": 1517
    },
    {
      "epoch": 1.434782608695652,
      "grad_norm": 0.3804474400665256,
      "learning_rate": 5.775739419702247e-05,
      "loss": 0.5312708616256714,
      "step": 1518
    },
    {
      "epoch": 1.4357277882797732,
      "grad_norm": 0.33381634383358116,
      "learning_rate": 5.770592147248196e-05,
      "loss": 0.5041257739067078,
      "step": 1519
    },
    {
      "epoch": 1.436672967863894,
      "grad_norm": 0.38197567057002824,
      "learning_rate": 5.765444038132901e-05,
      "loss": 0.4971402883529663,
      "step": 1520
    },
    {
      "epoch": 1.4376181474480152,
      "grad_norm": 0.38063226075385825,
      "learning_rate": 5.76029509794586e-05,
      "loss": 0.4947785437107086,
      "step": 1521
    },
    {
      "epoch": 1.438563327032136,
      "grad_norm": 0.323978976285035,
      "learning_rate": 5.7551453322774716e-05,
      "loss": 0.4617655873298645,
      "step": 1522
    },
    {
      "epoch": 1.439508506616257,
      "grad_norm": 0.38574416904883896,
      "learning_rate": 5.749994746719038e-05,
      "loss": 0.6056530475616455,
      "step": 1523
    },
    {
      "epoch": 1.440453686200378,
      "grad_norm": 0.3388420211944713,
      "learning_rate": 5.7448433468627396e-05,
      "loss": 0.49883174896240234,
      "step": 1524
    },
    {
      "epoch": 1.4413988657844992,
      "grad_norm": 0.3148760564172148,
      "learning_rate": 5.73969113830165e-05,
      "loss": 0.408201664686203,
      "step": 1525
    },
    {
      "epoch": 1.44234404536862,
      "grad_norm": 0.3781406172362981,
      "learning_rate": 5.7345381266297194e-05,
      "loss": 0.5076204538345337,
      "step": 1526
    },
    {
      "epoch": 1.443289224952741,
      "grad_norm": 0.34222059416461215,
      "learning_rate": 5.729384317441767e-05,
      "loss": 0.49621760845184326,
      "step": 1527
    },
    {
      "epoch": 1.444234404536862,
      "grad_norm": 0.35179229565127557,
      "learning_rate": 5.7242297163334804e-05,
      "loss": 0.5322737693786621,
      "step": 1528
    },
    {
      "epoch": 1.445179584120983,
      "grad_norm": 0.4222749236847371,
      "learning_rate": 5.7190743289014085e-05,
      "loss": 0.5726200938224792,
      "step": 1529
    },
    {
      "epoch": 1.446124763705104,
      "grad_norm": 0.3568969449987007,
      "learning_rate": 5.713918160742948e-05,
      "loss": 0.4785623252391815,
      "step": 1530
    },
    {
      "epoch": 1.447069943289225,
      "grad_norm": 0.3467120666857666,
      "learning_rate": 5.70876121745635e-05,
      "loss": 0.4292585849761963,
      "step": 1531
    },
    {
      "epoch": 1.448015122873346,
      "grad_norm": 0.3468834861582227,
      "learning_rate": 5.703603504640702e-05,
      "loss": 0.5353392362594604,
      "step": 1532
    },
    {
      "epoch": 1.4489603024574669,
      "grad_norm": 0.37248302233147723,
      "learning_rate": 5.698445027895931e-05,
      "loss": 0.5676983594894409,
      "step": 1533
    },
    {
      "epoch": 1.449905482041588,
      "grad_norm": 0.38931796388318357,
      "learning_rate": 5.69328579282279e-05,
      "loss": 0.5629222393035889,
      "step": 1534
    },
    {
      "epoch": 1.4508506616257089,
      "grad_norm": 0.32419395231076137,
      "learning_rate": 5.688125805022861e-05,
      "loss": 0.45296430587768555,
      "step": 1535
    },
    {
      "epoch": 1.45179584120983,
      "grad_norm": 0.38829350315390443,
      "learning_rate": 5.682965070098534e-05,
      "loss": 0.5911639332771301,
      "step": 1536
    },
    {
      "epoch": 1.4527410207939508,
      "grad_norm": 0.33254667772848295,
      "learning_rate": 5.6778035936530184e-05,
      "loss": 0.47687023878097534,
      "step": 1537
    },
    {
      "epoch": 1.4536862003780717,
      "grad_norm": 0.36223768719889493,
      "learning_rate": 5.672641381290325e-05,
      "loss": 0.5825412273406982,
      "step": 1538
    },
    {
      "epoch": 1.4546313799621928,
      "grad_norm": 0.3533467263452521,
      "learning_rate": 5.6674784386152635e-05,
      "loss": 0.5243592262268066,
      "step": 1539
    },
    {
      "epoch": 1.455576559546314,
      "grad_norm": 0.35843199554660005,
      "learning_rate": 5.66231477123344e-05,
      "loss": 0.5246967077255249,
      "step": 1540
    },
    {
      "epoch": 1.4565217391304348,
      "grad_norm": 0.3486828500759107,
      "learning_rate": 5.657150384751243e-05,
      "loss": 0.5150754451751709,
      "step": 1541
    },
    {
      "epoch": 1.4574669187145557,
      "grad_norm": 0.40383229412540844,
      "learning_rate": 5.651985284775845e-05,
      "loss": 0.4790208637714386,
      "step": 1542
    },
    {
      "epoch": 1.4584120982986768,
      "grad_norm": 0.37534099799734233,
      "learning_rate": 5.646819476915189e-05,
      "loss": 0.5270127654075623,
      "step": 1543
    },
    {
      "epoch": 1.4593572778827977,
      "grad_norm": 0.33753265427955853,
      "learning_rate": 5.641652966777993e-05,
      "loss": 0.4645795226097107,
      "step": 1544
    },
    {
      "epoch": 1.4603024574669188,
      "grad_norm": 0.3726130853310144,
      "learning_rate": 5.636485759973729e-05,
      "loss": 0.49515873193740845,
      "step": 1545
    },
    {
      "epoch": 1.4612476370510397,
      "grad_norm": 0.3759634471704034,
      "learning_rate": 5.631317862112636e-05,
      "loss": 0.5576908588409424,
      "step": 1546
    },
    {
      "epoch": 1.4621928166351608,
      "grad_norm": 0.4098967452237737,
      "learning_rate": 5.6261492788056924e-05,
      "loss": 0.5619882941246033,
      "step": 1547
    },
    {
      "epoch": 1.4631379962192816,
      "grad_norm": 0.39741994512582846,
      "learning_rate": 5.6209800156646294e-05,
      "loss": 0.5621123313903809,
      "step": 1548
    },
    {
      "epoch": 1.4640831758034025,
      "grad_norm": 0.4476571454294351,
      "learning_rate": 5.615810078301912e-05,
      "loss": 0.5707625150680542,
      "step": 1549
    },
    {
      "epoch": 1.4650283553875236,
      "grad_norm": 0.35811337274900434,
      "learning_rate": 5.6106394723307365e-05,
      "loss": 0.45583105087280273,
      "step": 1550
    },
    {
      "epoch": 1.4659735349716447,
      "grad_norm": 0.32228962791744725,
      "learning_rate": 5.6054682033650296e-05,
      "loss": 0.47290241718292236,
      "step": 1551
    },
    {
      "epoch": 1.4669187145557656,
      "grad_norm": 0.3806389646359353,
      "learning_rate": 5.6002962770194335e-05,
      "loss": 0.5461976528167725,
      "step": 1552
    },
    {
      "epoch": 1.4678638941398865,
      "grad_norm": 0.3931533598284275,
      "learning_rate": 5.5951236989093036e-05,
      "loss": 0.5833940505981445,
      "step": 1553
    },
    {
      "epoch": 1.4688090737240076,
      "grad_norm": 0.3639817873452855,
      "learning_rate": 5.589950474650709e-05,
      "loss": 0.5705401301383972,
      "step": 1554
    },
    {
      "epoch": 1.4697542533081285,
      "grad_norm": 0.3627185436078416,
      "learning_rate": 5.584776609860414e-05,
      "loss": 0.5335103273391724,
      "step": 1555
    },
    {
      "epoch": 1.4706994328922496,
      "grad_norm": 0.3638592399147534,
      "learning_rate": 5.579602110155878e-05,
      "loss": 0.5328971147537231,
      "step": 1556
    },
    {
      "epoch": 1.4716446124763705,
      "grad_norm": 0.3861658222047642,
      "learning_rate": 5.574426981155258e-05,
      "loss": 0.5493745803833008,
      "step": 1557
    },
    {
      "epoch": 1.4725897920604916,
      "grad_norm": 0.4027433549886297,
      "learning_rate": 5.5692512284773854e-05,
      "loss": 0.5264089703559875,
      "step": 1558
    },
    {
      "epoch": 1.4735349716446124,
      "grad_norm": 0.3585080237066396,
      "learning_rate": 5.5640748577417734e-05,
      "loss": 0.5163941383361816,
      "step": 1559
    },
    {
      "epoch": 1.4744801512287333,
      "grad_norm": 0.34564404479236555,
      "learning_rate": 5.558897874568604e-05,
      "loss": 0.48280447721481323,
      "step": 1560
    },
    {
      "epoch": 1.4754253308128544,
      "grad_norm": 0.3714259897850016,
      "learning_rate": 5.5537202845787226e-05,
      "loss": 0.4946216344833374,
      "step": 1561
    },
    {
      "epoch": 1.4763705103969755,
      "grad_norm": 0.3704468376007152,
      "learning_rate": 5.5485420933936404e-05,
      "loss": 0.4823738932609558,
      "step": 1562
    },
    {
      "epoch": 1.4773156899810964,
      "grad_norm": 0.34689685288285416,
      "learning_rate": 5.543363306635515e-05,
      "loss": 0.5337408781051636,
      "step": 1563
    },
    {
      "epoch": 1.4782608695652173,
      "grad_norm": 0.38806216840529084,
      "learning_rate": 5.538183929927152e-05,
      "loss": 0.5665512084960938,
      "step": 1564
    },
    {
      "epoch": 1.4792060491493384,
      "grad_norm": 0.3795235280486561,
      "learning_rate": 5.533003968891998e-05,
      "loss": 0.5691730976104736,
      "step": 1565
    },
    {
      "epoch": 1.4801512287334593,
      "grad_norm": 0.4018295114299985,
      "learning_rate": 5.527823429154133e-05,
      "loss": 0.5360079407691956,
      "step": 1566
    },
    {
      "epoch": 1.4810964083175804,
      "grad_norm": 0.3570008144473228,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.49208560585975647,
      "step": 1567
    },
    {
      "epoch": 1.4820415879017013,
      "grad_norm": 0.39283243049041655,
      "learning_rate": 5.5174606360697334e-05,
      "loss": 0.5494649410247803,
      "step": 1568
    },
    {
      "epoch": 1.4829867674858224,
      "grad_norm": 0.42632910142107683,
      "learning_rate": 5.512278393974478e-05,
      "loss": 0.5900541543960571,
      "step": 1569
    },
    {
      "epoch": 1.4839319470699432,
      "grad_norm": 0.4220531966290287,
      "learning_rate": 5.5070955956790594e-05,
      "loss": 0.5219538807868958,
      "step": 1570
    },
    {
      "epoch": 1.4848771266540643,
      "grad_norm": 0.3491950055443269,
      "learning_rate": 5.501912246810638e-05,
      "loss": 0.5022003054618835,
      "step": 1571
    },
    {
      "epoch": 1.4858223062381852,
      "grad_norm": 0.37760001537316856,
      "learning_rate": 5.496728352996973e-05,
      "loss": 0.5482698678970337,
      "step": 1572
    },
    {
      "epoch": 1.4867674858223063,
      "grad_norm": 0.3725682276629769,
      "learning_rate": 5.4915439198664164e-05,
      "loss": 0.518518328666687,
      "step": 1573
    },
    {
      "epoch": 1.4877126654064272,
      "grad_norm": 0.40878591752452925,
      "learning_rate": 5.486358953047904e-05,
      "loss": 0.607046365737915,
      "step": 1574
    },
    {
      "epoch": 1.488657844990548,
      "grad_norm": 0.4443257829756715,
      "learning_rate": 5.4811734581709514e-05,
      "loss": 0.521558403968811,
      "step": 1575
    },
    {
      "epoch": 1.4896030245746692,
      "grad_norm": 0.369074734179134,
      "learning_rate": 5.47598744086565e-05,
      "loss": 0.5074565410614014,
      "step": 1576
    },
    {
      "epoch": 1.4905482041587903,
      "grad_norm": 0.37641390807748343,
      "learning_rate": 5.4708009067626545e-05,
      "loss": 0.483066588640213,
      "step": 1577
    },
    {
      "epoch": 1.4914933837429112,
      "grad_norm": 0.4218397064226925,
      "learning_rate": 5.4656138614931816e-05,
      "loss": 0.6046510338783264,
      "step": 1578
    },
    {
      "epoch": 1.492438563327032,
      "grad_norm": 0.4100127715839473,
      "learning_rate": 5.460426310689005e-05,
      "loss": 0.5343729853630066,
      "step": 1579
    },
    {
      "epoch": 1.4933837429111532,
      "grad_norm": 0.3852974834578726,
      "learning_rate": 5.455238259982448e-05,
      "loss": 0.49311500787734985,
      "step": 1580
    },
    {
      "epoch": 1.494328922495274,
      "grad_norm": 0.4168803314422741,
      "learning_rate": 5.450049715006372e-05,
      "loss": 0.5651973485946655,
      "step": 1581
    },
    {
      "epoch": 1.4952741020793952,
      "grad_norm": 0.4204545596356735,
      "learning_rate": 5.4448606813941805e-05,
      "loss": 0.48860424757003784,
      "step": 1582
    },
    {
      "epoch": 1.496219281663516,
      "grad_norm": 0.38190862524366653,
      "learning_rate": 5.4396711647798015e-05,
      "loss": 0.5801682472229004,
      "step": 1583
    },
    {
      "epoch": 1.4971644612476371,
      "grad_norm": 0.31135775226366735,
      "learning_rate": 5.4344811707976936e-05,
      "loss": 0.44278979301452637,
      "step": 1584
    },
    {
      "epoch": 1.498109640831758,
      "grad_norm": 0.37350158388714716,
      "learning_rate": 5.42929070508283e-05,
      "loss": 0.5242373943328857,
      "step": 1585
    },
    {
      "epoch": 1.499054820415879,
      "grad_norm": 0.37397283166862355,
      "learning_rate": 5.424099773270697e-05,
      "loss": 0.5177803039550781,
      "step": 1586
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.38425025584282396,
      "learning_rate": 5.4189083809972884e-05,
      "loss": 0.5180373191833496,
      "step": 1587
    },
    {
      "epoch": 1.500945179584121,
      "grad_norm": 0.3619735808350762,
      "learning_rate": 5.413716533899096e-05,
      "loss": 0.5393504500389099,
      "step": 1588
    },
    {
      "epoch": 1.501890359168242,
      "grad_norm": 0.3316684455364464,
      "learning_rate": 5.4085242376131015e-05,
      "loss": 0.47502562403678894,
      "step": 1589
    },
    {
      "epoch": 1.5028355387523629,
      "grad_norm": 0.40380507809919974,
      "learning_rate": 5.4033314977767856e-05,
      "loss": 0.5638004541397095,
      "step": 1590
    },
    {
      "epoch": 1.503780718336484,
      "grad_norm": 0.3428366430664611,
      "learning_rate": 5.3981383200281e-05,
      "loss": 0.4959918260574341,
      "step": 1591
    },
    {
      "epoch": 1.504725897920605,
      "grad_norm": 0.4255922148672931,
      "learning_rate": 5.392944710005475e-05,
      "loss": 0.589640736579895,
      "step": 1592
    },
    {
      "epoch": 1.505671077504726,
      "grad_norm": 0.35744267277177244,
      "learning_rate": 5.387750673347814e-05,
      "loss": 0.5434280633926392,
      "step": 1593
    },
    {
      "epoch": 1.5066162570888468,
      "grad_norm": 0.3502142756923529,
      "learning_rate": 5.3825562156944774e-05,
      "loss": 0.4691290855407715,
      "step": 1594
    },
    {
      "epoch": 1.507561436672968,
      "grad_norm": 0.3563661182444055,
      "learning_rate": 5.377361342685287e-05,
      "loss": 0.4908035397529602,
      "step": 1595
    },
    {
      "epoch": 1.5085066162570888,
      "grad_norm": 0.39054297964882767,
      "learning_rate": 5.372166059960514e-05,
      "loss": 0.5866037607192993,
      "step": 1596
    },
    {
      "epoch": 1.5094517958412097,
      "grad_norm": 0.3842954186200336,
      "learning_rate": 5.366970373160874e-05,
      "loss": 0.4649214744567871,
      "step": 1597
    },
    {
      "epoch": 1.5103969754253308,
      "grad_norm": 0.36597340243418736,
      "learning_rate": 5.3617742879275224e-05,
      "loss": 0.4774046540260315,
      "step": 1598
    },
    {
      "epoch": 1.511342155009452,
      "grad_norm": 0.4525299165190927,
      "learning_rate": 5.3565778099020504e-05,
      "loss": 0.5637791156768799,
      "step": 1599
    },
    {
      "epoch": 1.5122873345935728,
      "grad_norm": 0.3124285996767469,
      "learning_rate": 5.351380944726465e-05,
      "loss": 0.434393048286438,
      "step": 1600
    },
    {
      "epoch": 1.5132325141776937,
      "grad_norm": 0.3686905544501689,
      "learning_rate": 5.346183698043207e-05,
      "loss": 0.5675680041313171,
      "step": 1601
    },
    {
      "epoch": 1.5141776937618148,
      "grad_norm": 0.3969393437164199,
      "learning_rate": 5.3409860754951225e-05,
      "loss": 0.5228263139724731,
      "step": 1602
    },
    {
      "epoch": 1.5151228733459359,
      "grad_norm": 0.33437965338234127,
      "learning_rate": 5.335788082725467e-05,
      "loss": 0.4593347907066345,
      "step": 1603
    },
    {
      "epoch": 1.5160680529300568,
      "grad_norm": 0.3752390920495842,
      "learning_rate": 5.3305897253779025e-05,
      "loss": 0.49378642439842224,
      "step": 1604
    },
    {
      "epoch": 1.5170132325141776,
      "grad_norm": 0.37816648598068847,
      "learning_rate": 5.325391009096481e-05,
      "loss": 0.5369211435317993,
      "step": 1605
    },
    {
      "epoch": 1.5179584120982987,
      "grad_norm": 0.37706459299436085,
      "learning_rate": 5.3201919395256475e-05,
      "loss": 0.5428913831710815,
      "step": 1606
    },
    {
      "epoch": 1.5189035916824196,
      "grad_norm": 0.37992227921286437,
      "learning_rate": 5.314992522310229e-05,
      "loss": 0.49954989552497864,
      "step": 1607
    },
    {
      "epoch": 1.5198487712665405,
      "grad_norm": 0.3997985956352674,
      "learning_rate": 5.3097927630954327e-05,
      "loss": 0.5659242272377014,
      "step": 1608
    },
    {
      "epoch": 1.5207939508506616,
      "grad_norm": 0.3923465198910532,
      "learning_rate": 5.3045926675268344e-05,
      "loss": 0.5104411244392395,
      "step": 1609
    },
    {
      "epoch": 1.5217391304347827,
      "grad_norm": 0.36901157747375796,
      "learning_rate": 5.299392241250376e-05,
      "loss": 0.5272958874702454,
      "step": 1610
    },
    {
      "epoch": 1.5226843100189036,
      "grad_norm": 0.38240601572431315,
      "learning_rate": 5.294191489912357e-05,
      "loss": 0.5616653561592102,
      "step": 1611
    },
    {
      "epoch": 1.5236294896030245,
      "grad_norm": 0.3725423778605971,
      "learning_rate": 5.2889904191594334e-05,
      "loss": 0.4891291856765747,
      "step": 1612
    },
    {
      "epoch": 1.5245746691871456,
      "grad_norm": 0.35662709722827907,
      "learning_rate": 5.2837890346386043e-05,
      "loss": 0.4797162711620331,
      "step": 1613
    },
    {
      "epoch": 1.5255198487712667,
      "grad_norm": 0.37049039755672625,
      "learning_rate": 5.278587341997211e-05,
      "loss": 0.5293033719062805,
      "step": 1614
    },
    {
      "epoch": 1.5264650283553876,
      "grad_norm": 0.3387274710850052,
      "learning_rate": 5.27338534688293e-05,
      "loss": 0.46112239360809326,
      "step": 1615
    },
    {
      "epoch": 1.5274102079395084,
      "grad_norm": 0.38706565776302826,
      "learning_rate": 5.268183054943765e-05,
      "loss": 0.5276835560798645,
      "step": 1616
    },
    {
      "epoch": 1.5283553875236295,
      "grad_norm": 0.38907268150854596,
      "learning_rate": 5.262980471828044e-05,
      "loss": 0.5603146553039551,
      "step": 1617
    },
    {
      "epoch": 1.5293005671077504,
      "grad_norm": 0.37185584751465045,
      "learning_rate": 5.257777603184407e-05,
      "loss": 0.489784300327301,
      "step": 1618
    },
    {
      "epoch": 1.5302457466918713,
      "grad_norm": 0.39832777966814525,
      "learning_rate": 5.252574454661808e-05,
      "loss": 0.5741744041442871,
      "step": 1619
    },
    {
      "epoch": 1.5311909262759924,
      "grad_norm": 0.38196034715106475,
      "learning_rate": 5.247371031909505e-05,
      "loss": 0.5808093547821045,
      "step": 1620
    },
    {
      "epoch": 1.5321361058601135,
      "grad_norm": 0.3768831741311102,
      "learning_rate": 5.24216734057705e-05,
      "loss": 0.5865797996520996,
      "step": 1621
    },
    {
      "epoch": 1.5330812854442344,
      "grad_norm": 0.3919303242408739,
      "learning_rate": 5.236963386314291e-05,
      "loss": 0.5346591472625732,
      "step": 1622
    },
    {
      "epoch": 1.5340264650283553,
      "grad_norm": 0.37412005254568853,
      "learning_rate": 5.231759174771357e-05,
      "loss": 0.5131747126579285,
      "step": 1623
    },
    {
      "epoch": 1.5349716446124764,
      "grad_norm": 0.36150330538142245,
      "learning_rate": 5.226554711598659e-05,
      "loss": 0.5446631908416748,
      "step": 1624
    },
    {
      "epoch": 1.5359168241965975,
      "grad_norm": 0.37015393640502414,
      "learning_rate": 5.221350002446882e-05,
      "loss": 0.5645014047622681,
      "step": 1625
    },
    {
      "epoch": 1.5368620037807184,
      "grad_norm": 0.3913891114108142,
      "learning_rate": 5.2161450529669775e-05,
      "loss": 0.4549786448478699,
      "step": 1626
    },
    {
      "epoch": 1.5378071833648392,
      "grad_norm": 0.36959771153958626,
      "learning_rate": 5.210939868810156e-05,
      "loss": 0.4756149649620056,
      "step": 1627
    },
    {
      "epoch": 1.5387523629489603,
      "grad_norm": 0.3380649220133284,
      "learning_rate": 5.205734455627885e-05,
      "loss": 0.4607565999031067,
      "step": 1628
    },
    {
      "epoch": 1.5396975425330814,
      "grad_norm": 0.3710704099809346,
      "learning_rate": 5.2005288190718794e-05,
      "loss": 0.5390820503234863,
      "step": 1629
    },
    {
      "epoch": 1.5406427221172023,
      "grad_norm": 0.39840211378508733,
      "learning_rate": 5.195322964794098e-05,
      "loss": 0.5438518524169922,
      "step": 1630
    },
    {
      "epoch": 1.5415879017013232,
      "grad_norm": 0.44991316378889695,
      "learning_rate": 5.190116898446732e-05,
      "loss": 0.5808957815170288,
      "step": 1631
    },
    {
      "epoch": 1.5425330812854443,
      "grad_norm": 0.423929644665659,
      "learning_rate": 5.18491062568221e-05,
      "loss": 0.5544024109840393,
      "step": 1632
    },
    {
      "epoch": 1.5434782608695652,
      "grad_norm": 0.3700275830441986,
      "learning_rate": 5.17970415215318e-05,
      "loss": 0.5348548889160156,
      "step": 1633
    },
    {
      "epoch": 1.544423440453686,
      "grad_norm": 0.4524031314838702,
      "learning_rate": 5.174497483512506e-05,
      "loss": 0.6148048639297485,
      "step": 1634
    },
    {
      "epoch": 1.5453686200378072,
      "grad_norm": 0.31767931391683385,
      "learning_rate": 5.169290625413268e-05,
      "loss": 0.4230349063873291,
      "step": 1635
    },
    {
      "epoch": 1.5463137996219283,
      "grad_norm": 0.3709069533081141,
      "learning_rate": 5.16408358350875e-05,
      "loss": 0.4601902365684509,
      "step": 1636
    },
    {
      "epoch": 1.5472589792060492,
      "grad_norm": 0.36947518403743,
      "learning_rate": 5.158876363452435e-05,
      "loss": 0.533743143081665,
      "step": 1637
    },
    {
      "epoch": 1.54820415879017,
      "grad_norm": 0.38268353532584026,
      "learning_rate": 5.153668970898002e-05,
      "loss": 0.5375320315361023,
      "step": 1638
    },
    {
      "epoch": 1.5491493383742911,
      "grad_norm": 0.36431356254461483,
      "learning_rate": 5.1484614114993156e-05,
      "loss": 0.5582842826843262,
      "step": 1639
    },
    {
      "epoch": 1.5500945179584122,
      "grad_norm": 0.38061305870376555,
      "learning_rate": 5.143253690910419e-05,
      "loss": 0.5344516038894653,
      "step": 1640
    },
    {
      "epoch": 1.5510396975425331,
      "grad_norm": 0.35739674944883215,
      "learning_rate": 5.138045814785534e-05,
      "loss": 0.5288535356521606,
      "step": 1641
    },
    {
      "epoch": 1.551984877126654,
      "grad_norm": 0.37038514193647193,
      "learning_rate": 5.132837788779049e-05,
      "loss": 0.49398982524871826,
      "step": 1642
    },
    {
      "epoch": 1.552930056710775,
      "grad_norm": 0.35410033809060504,
      "learning_rate": 5.1276296185455185e-05,
      "loss": 0.5047576427459717,
      "step": 1643
    },
    {
      "epoch": 1.553875236294896,
      "grad_norm": 0.3945887750916327,
      "learning_rate": 5.12242130973965e-05,
      "loss": 0.4922971725463867,
      "step": 1644
    },
    {
      "epoch": 1.5548204158790169,
      "grad_norm": 0.4024373849416959,
      "learning_rate": 5.117212868016303e-05,
      "loss": 0.5435211658477783,
      "step": 1645
    },
    {
      "epoch": 1.555765595463138,
      "grad_norm": 0.43136444617127895,
      "learning_rate": 5.112004299030481e-05,
      "loss": 0.5091035962104797,
      "step": 1646
    },
    {
      "epoch": 1.556710775047259,
      "grad_norm": 0.39970195818203524,
      "learning_rate": 5.1067956084373236e-05,
      "loss": 0.501255214214325,
      "step": 1647
    },
    {
      "epoch": 1.55765595463138,
      "grad_norm": 0.3667967189342809,
      "learning_rate": 5.101586801892109e-05,
      "loss": 0.5106174945831299,
      "step": 1648
    },
    {
      "epoch": 1.5586011342155008,
      "grad_norm": 0.37130102542774585,
      "learning_rate": 5.096377885050233e-05,
      "loss": 0.432289719581604,
      "step": 1649
    },
    {
      "epoch": 1.559546313799622,
      "grad_norm": 0.3899991702408732,
      "learning_rate": 5.091168863567215e-05,
      "loss": 0.5257974863052368,
      "step": 1650
    },
    {
      "epoch": 1.560491493383743,
      "grad_norm": 0.3696497981466453,
      "learning_rate": 5.085959743098693e-05,
      "loss": 0.5493028163909912,
      "step": 1651
    },
    {
      "epoch": 1.561436672967864,
      "grad_norm": 0.37066450734132983,
      "learning_rate": 5.080750529300402e-05,
      "loss": 0.5047093629837036,
      "step": 1652
    },
    {
      "epoch": 1.5623818525519848,
      "grad_norm": 0.3927045422997042,
      "learning_rate": 5.0755412278281844e-05,
      "loss": 0.549629271030426,
      "step": 1653
    },
    {
      "epoch": 1.563327032136106,
      "grad_norm": 0.3817615177720619,
      "learning_rate": 5.07033184433798e-05,
      "loss": 0.5310356020927429,
      "step": 1654
    },
    {
      "epoch": 1.5642722117202268,
      "grad_norm": 0.353448991345226,
      "learning_rate": 5.065122384485814e-05,
      "loss": 0.457400381565094,
      "step": 1655
    },
    {
      "epoch": 1.5652173913043477,
      "grad_norm": 0.3862915808772388,
      "learning_rate": 5.059912853927794e-05,
      "loss": 0.5378060340881348,
      "step": 1656
    },
    {
      "epoch": 1.5661625708884688,
      "grad_norm": 0.37419440749769006,
      "learning_rate": 5.054703258320107e-05,
      "loss": 0.48655015230178833,
      "step": 1657
    },
    {
      "epoch": 1.5671077504725899,
      "grad_norm": 0.40569470509510464,
      "learning_rate": 5.049493603319008e-05,
      "loss": 0.565788984298706,
      "step": 1658
    },
    {
      "epoch": 1.5680529300567108,
      "grad_norm": 0.37039335985113303,
      "learning_rate": 5.044283894580819e-05,
      "loss": 0.5302537679672241,
      "step": 1659
    },
    {
      "epoch": 1.5689981096408316,
      "grad_norm": 0.38549089274724563,
      "learning_rate": 5.03907413776192e-05,
      "loss": 0.5615366101264954,
      "step": 1660
    },
    {
      "epoch": 1.5699432892249527,
      "grad_norm": 0.39816538288118525,
      "learning_rate": 5.03386433851874e-05,
      "loss": 0.5447930693626404,
      "step": 1661
    },
    {
      "epoch": 1.5708884688090738,
      "grad_norm": 0.37447626823564356,
      "learning_rate": 5.028654502507759e-05,
      "loss": 0.5775421857833862,
      "step": 1662
    },
    {
      "epoch": 1.5718336483931947,
      "grad_norm": 0.39587470232102756,
      "learning_rate": 5.023444635385493e-05,
      "loss": 0.5989867448806763,
      "step": 1663
    },
    {
      "epoch": 1.5727788279773156,
      "grad_norm": 0.4296018653632832,
      "learning_rate": 5.018234742808493e-05,
      "loss": 0.548826277256012,
      "step": 1664
    },
    {
      "epoch": 1.5737240075614367,
      "grad_norm": 0.3716329668502776,
      "learning_rate": 5.013024830433338e-05,
      "loss": 0.5328786373138428,
      "step": 1665
    },
    {
      "epoch": 1.5746691871455578,
      "grad_norm": 0.3803211881854655,
      "learning_rate": 5.00781490391663e-05,
      "loss": 0.5192501544952393,
      "step": 1666
    },
    {
      "epoch": 1.5756143667296785,
      "grad_norm": 0.3543191049876515,
      "learning_rate": 5.002604968914981e-05,
      "loss": 0.45691555738449097,
      "step": 1667
    },
    {
      "epoch": 1.5765595463137996,
      "grad_norm": 0.3748117527656865,
      "learning_rate": 4.997395031085019e-05,
      "loss": 0.5505545735359192,
      "step": 1668
    },
    {
      "epoch": 1.5775047258979207,
      "grad_norm": 0.37199609342326806,
      "learning_rate": 4.992185096083372e-05,
      "loss": 0.5083546042442322,
      "step": 1669
    },
    {
      "epoch": 1.5784499054820416,
      "grad_norm": 0.3849354283836987,
      "learning_rate": 4.9869751695666615e-05,
      "loss": 0.5764696002006531,
      "step": 1670
    },
    {
      "epoch": 1.5793950850661624,
      "grad_norm": 0.3527039849678605,
      "learning_rate": 4.9817652571915085e-05,
      "loss": 0.46788260340690613,
      "step": 1671
    },
    {
      "epoch": 1.5803402646502835,
      "grad_norm": 0.3648749358076265,
      "learning_rate": 4.9765553646145086e-05,
      "loss": 0.5629303455352783,
      "step": 1672
    },
    {
      "epoch": 1.5812854442344046,
      "grad_norm": 0.42194584669309426,
      "learning_rate": 4.971345497492242e-05,
      "loss": 0.5183531045913696,
      "step": 1673
    },
    {
      "epoch": 1.5822306238185255,
      "grad_norm": 0.3642110404925853,
      "learning_rate": 4.9661356614812615e-05,
      "loss": 0.49611952900886536,
      "step": 1674
    },
    {
      "epoch": 1.5831758034026464,
      "grad_norm": 0.3674945377562542,
      "learning_rate": 4.96092586223808e-05,
      "loss": 0.5112574100494385,
      "step": 1675
    },
    {
      "epoch": 1.5841209829867675,
      "grad_norm": 0.3945055630959171,
      "learning_rate": 4.955716105419182e-05,
      "loss": 0.5291755795478821,
      "step": 1676
    },
    {
      "epoch": 1.5850661625708886,
      "grad_norm": 0.3228113092725764,
      "learning_rate": 4.950506396680993e-05,
      "loss": 0.46474188566207886,
      "step": 1677
    },
    {
      "epoch": 1.5860113421550095,
      "grad_norm": 0.4316949911715895,
      "learning_rate": 4.945296741679894e-05,
      "loss": 0.5299582481384277,
      "step": 1678
    },
    {
      "epoch": 1.5869565217391304,
      "grad_norm": 0.3819032105770187,
      "learning_rate": 4.940087146072208e-05,
      "loss": 0.5099689364433289,
      "step": 1679
    },
    {
      "epoch": 1.5879017013232515,
      "grad_norm": 0.35796275392102683,
      "learning_rate": 4.9348776155141876e-05,
      "loss": 0.5045382976531982,
      "step": 1680
    },
    {
      "epoch": 1.5888468809073724,
      "grad_norm": 0.3647230002659726,
      "learning_rate": 4.9296681556620207e-05,
      "loss": 0.5705474019050598,
      "step": 1681
    },
    {
      "epoch": 1.5897920604914932,
      "grad_norm": 0.35737792222180653,
      "learning_rate": 4.9244587721718174e-05,
      "loss": 0.5142084360122681,
      "step": 1682
    },
    {
      "epoch": 1.5907372400756143,
      "grad_norm": 0.4204770127610456,
      "learning_rate": 4.919249470699599e-05,
      "loss": 0.5522286891937256,
      "step": 1683
    },
    {
      "epoch": 1.5916824196597354,
      "grad_norm": 0.3267590099042977,
      "learning_rate": 4.914040256901309e-05,
      "loss": 0.4477725923061371,
      "step": 1684
    },
    {
      "epoch": 1.5926275992438563,
      "grad_norm": 0.3541952414977977,
      "learning_rate": 4.908831136432784e-05,
      "loss": 0.514543354511261,
      "step": 1685
    },
    {
      "epoch": 1.5935727788279772,
      "grad_norm": 0.37434597026880095,
      "learning_rate": 4.9036221149497684e-05,
      "loss": 0.5272210240364075,
      "step": 1686
    },
    {
      "epoch": 1.5945179584120983,
      "grad_norm": 0.38431800061984883,
      "learning_rate": 4.8984131981078915e-05,
      "loss": 0.5120689868927002,
      "step": 1687
    },
    {
      "epoch": 1.5954631379962194,
      "grad_norm": 0.3431956163715082,
      "learning_rate": 4.8932043915626776e-05,
      "loss": 0.5108009576797485,
      "step": 1688
    },
    {
      "epoch": 1.5964083175803403,
      "grad_norm": 0.34208005574313194,
      "learning_rate": 4.8879957009695216e-05,
      "loss": 0.4621889293193817,
      "step": 1689
    },
    {
      "epoch": 1.5973534971644612,
      "grad_norm": 0.37078539087193363,
      "learning_rate": 4.882787131983698e-05,
      "loss": 0.4795047342777252,
      "step": 1690
    },
    {
      "epoch": 1.5982986767485823,
      "grad_norm": 0.3786174979132033,
      "learning_rate": 4.877578690260351e-05,
      "loss": 0.5817227363586426,
      "step": 1691
    },
    {
      "epoch": 1.5992438563327032,
      "grad_norm": 0.34905034844908495,
      "learning_rate": 4.872370381454481e-05,
      "loss": 0.46846166253089905,
      "step": 1692
    },
    {
      "epoch": 1.600189035916824,
      "grad_norm": 0.3337225370513404,
      "learning_rate": 4.867162211220952e-05,
      "loss": 0.4905742406845093,
      "step": 1693
    },
    {
      "epoch": 1.6011342155009451,
      "grad_norm": 0.36724182285498064,
      "learning_rate": 4.861954185214468e-05,
      "loss": 0.5269461274147034,
      "step": 1694
    },
    {
      "epoch": 1.6020793950850662,
      "grad_norm": 0.38124318250583245,
      "learning_rate": 4.856746309089582e-05,
      "loss": 0.5217968225479126,
      "step": 1695
    },
    {
      "epoch": 1.6030245746691871,
      "grad_norm": 0.3680028183816457,
      "learning_rate": 4.851538588500687e-05,
      "loss": 0.5550822019577026,
      "step": 1696
    },
    {
      "epoch": 1.603969754253308,
      "grad_norm": 0.38025561034441335,
      "learning_rate": 4.846331029101998e-05,
      "loss": 0.5198801755905151,
      "step": 1697
    },
    {
      "epoch": 1.6049149338374291,
      "grad_norm": 0.3647700392854306,
      "learning_rate": 4.8411236365475655e-05,
      "loss": 0.45916375517845154,
      "step": 1698
    },
    {
      "epoch": 1.6058601134215502,
      "grad_norm": 0.3876935732368991,
      "learning_rate": 4.835916416491251e-05,
      "loss": 0.5376836061477661,
      "step": 1699
    },
    {
      "epoch": 1.606805293005671,
      "grad_norm": 0.34298404207626726,
      "learning_rate": 4.8307093745867335e-05,
      "loss": 0.4830905795097351,
      "step": 1700
    },
    {
      "epoch": 1.607750472589792,
      "grad_norm": 0.35174188258190625,
      "learning_rate": 4.825502516487497e-05,
      "loss": 0.48561912775039673,
      "step": 1701
    },
    {
      "epoch": 1.608695652173913,
      "grad_norm": 0.3557540847652119,
      "learning_rate": 4.820295847846822e-05,
      "loss": 0.4750676155090332,
      "step": 1702
    },
    {
      "epoch": 1.6096408317580342,
      "grad_norm": 0.34760204508717907,
      "learning_rate": 4.8150893743177906e-05,
      "loss": 0.48417994379997253,
      "step": 1703
    },
    {
      "epoch": 1.6105860113421548,
      "grad_norm": 0.3290050204839735,
      "learning_rate": 4.809883101553267e-05,
      "loss": 0.44669562578201294,
      "step": 1704
    },
    {
      "epoch": 1.611531190926276,
      "grad_norm": 0.35538760120522817,
      "learning_rate": 4.804677035205903e-05,
      "loss": 0.4814063012599945,
      "step": 1705
    },
    {
      "epoch": 1.612476370510397,
      "grad_norm": 0.4110920386818634,
      "learning_rate": 4.7994711809281224e-05,
      "loss": 0.5431903600692749,
      "step": 1706
    },
    {
      "epoch": 1.613421550094518,
      "grad_norm": 0.3773211402058403,
      "learning_rate": 4.7942655443721165e-05,
      "loss": 0.5561413764953613,
      "step": 1707
    },
    {
      "epoch": 1.6143667296786388,
      "grad_norm": 0.42059204261611316,
      "learning_rate": 4.789060131189845e-05,
      "loss": 0.5951234102249146,
      "step": 1708
    },
    {
      "epoch": 1.61531190926276,
      "grad_norm": 0.4043748891719786,
      "learning_rate": 4.783854947033023e-05,
      "loss": 0.5696234107017517,
      "step": 1709
    },
    {
      "epoch": 1.616257088846881,
      "grad_norm": 0.44797804606813335,
      "learning_rate": 4.778649997553119e-05,
      "loss": 0.5894062519073486,
      "step": 1710
    },
    {
      "epoch": 1.617202268431002,
      "grad_norm": 0.3436645382226002,
      "learning_rate": 4.773445288401343e-05,
      "loss": 0.47461044788360596,
      "step": 1711
    },
    {
      "epoch": 1.6181474480151228,
      "grad_norm": 0.33431641961512226,
      "learning_rate": 4.768240825228645e-05,
      "loss": 0.4334047734737396,
      "step": 1712
    },
    {
      "epoch": 1.6190926275992439,
      "grad_norm": 0.4050518088293065,
      "learning_rate": 4.763036613685711e-05,
      "loss": 0.5649898052215576,
      "step": 1713
    },
    {
      "epoch": 1.620037807183365,
      "grad_norm": 0.3740675808664451,
      "learning_rate": 4.75783265942295e-05,
      "loss": 0.509590744972229,
      "step": 1714
    },
    {
      "epoch": 1.6209829867674859,
      "grad_norm": 0.3484262085015192,
      "learning_rate": 4.752628968090496e-05,
      "loss": 0.46480345726013184,
      "step": 1715
    },
    {
      "epoch": 1.6219281663516067,
      "grad_norm": 0.36578429546831254,
      "learning_rate": 4.747425545338192e-05,
      "loss": 0.5355474352836609,
      "step": 1716
    },
    {
      "epoch": 1.6228733459357279,
      "grad_norm": 0.3797596208680713,
      "learning_rate": 4.742222396815593e-05,
      "loss": 0.516162633895874,
      "step": 1717
    },
    {
      "epoch": 1.6238185255198487,
      "grad_norm": 0.3981791176279372,
      "learning_rate": 4.737019528171958e-05,
      "loss": 0.5852369070053101,
      "step": 1718
    },
    {
      "epoch": 1.6247637051039696,
      "grad_norm": 0.4023232864852648,
      "learning_rate": 4.7318169450562357e-05,
      "loss": 0.5149006247520447,
      "step": 1719
    },
    {
      "epoch": 1.6257088846880907,
      "grad_norm": 0.3273306537950893,
      "learning_rate": 4.726614653117071e-05,
      "loss": 0.4662395119667053,
      "step": 1720
    },
    {
      "epoch": 1.6266540642722118,
      "grad_norm": 0.3761893532609091,
      "learning_rate": 4.72141265800279e-05,
      "loss": 0.5487098693847656,
      "step": 1721
    },
    {
      "epoch": 1.6275992438563327,
      "grad_norm": 0.3838143853433904,
      "learning_rate": 4.716210965361397e-05,
      "loss": 0.5561859607696533,
      "step": 1722
    },
    {
      "epoch": 1.6285444234404536,
      "grad_norm": 0.3906306225802898,
      "learning_rate": 4.711009580840569e-05,
      "loss": 0.5271644592285156,
      "step": 1723
    },
    {
      "epoch": 1.6294896030245747,
      "grad_norm": 0.36863443572143473,
      "learning_rate": 4.7058085100876435e-05,
      "loss": 0.5540220737457275,
      "step": 1724
    },
    {
      "epoch": 1.6304347826086958,
      "grad_norm": 0.35481164937117826,
      "learning_rate": 4.700607758749625e-05,
      "loss": 0.4823080897331238,
      "step": 1725
    },
    {
      "epoch": 1.6313799621928167,
      "grad_norm": 0.38670996029863175,
      "learning_rate": 4.6954073324731654e-05,
      "loss": 0.5391385555267334,
      "step": 1726
    },
    {
      "epoch": 1.6323251417769375,
      "grad_norm": 0.38732098628928036,
      "learning_rate": 4.6902072369045685e-05,
      "loss": 0.5078773498535156,
      "step": 1727
    },
    {
      "epoch": 1.6332703213610587,
      "grad_norm": 0.3437581771274246,
      "learning_rate": 4.68500747768977e-05,
      "loss": 0.4892716407775879,
      "step": 1728
    },
    {
      "epoch": 1.6342155009451795,
      "grad_norm": 0.35951243651091874,
      "learning_rate": 4.679808060474354e-05,
      "loss": 0.47137606143951416,
      "step": 1729
    },
    {
      "epoch": 1.6351606805293004,
      "grad_norm": 0.36465188478635097,
      "learning_rate": 4.674608990903521e-05,
      "loss": 0.4742051959037781,
      "step": 1730
    },
    {
      "epoch": 1.6361058601134215,
      "grad_norm": 0.40415961190626964,
      "learning_rate": 4.669410274622098e-05,
      "loss": 0.474221408367157,
      "step": 1731
    },
    {
      "epoch": 1.6370510396975426,
      "grad_norm": 0.386383867639003,
      "learning_rate": 4.6642119172745347e-05,
      "loss": 0.5045959949493408,
      "step": 1732
    },
    {
      "epoch": 1.6379962192816635,
      "grad_norm": 0.332274468993736,
      "learning_rate": 4.659013924504879e-05,
      "loss": 0.43866896629333496,
      "step": 1733
    },
    {
      "epoch": 1.6389413988657844,
      "grad_norm": 0.417357297703951,
      "learning_rate": 4.6538163019567945e-05,
      "loss": 0.5934369564056396,
      "step": 1734
    },
    {
      "epoch": 1.6398865784499055,
      "grad_norm": 0.39366713771854617,
      "learning_rate": 4.648619055273537e-05,
      "loss": 0.5522946119308472,
      "step": 1735
    },
    {
      "epoch": 1.6408317580340266,
      "grad_norm": 0.40656559925473273,
      "learning_rate": 4.6434221900979514e-05,
      "loss": 0.5317525267601013,
      "step": 1736
    },
    {
      "epoch": 1.6417769376181475,
      "grad_norm": 0.3550812093478995,
      "learning_rate": 4.638225712072478e-05,
      "loss": 0.5309265851974487,
      "step": 1737
    },
    {
      "epoch": 1.6427221172022684,
      "grad_norm": 0.3907885654194595,
      "learning_rate": 4.633029626839127e-05,
      "loss": 0.581211507320404,
      "step": 1738
    },
    {
      "epoch": 1.6436672967863895,
      "grad_norm": 0.39498528824348633,
      "learning_rate": 4.627833940039487e-05,
      "loss": 0.5910161137580872,
      "step": 1739
    },
    {
      "epoch": 1.6446124763705106,
      "grad_norm": 0.3822298284036613,
      "learning_rate": 4.622638657314716e-05,
      "loss": 0.502801775932312,
      "step": 1740
    },
    {
      "epoch": 1.6455576559546312,
      "grad_norm": 0.3783404723384362,
      "learning_rate": 4.617443784305524e-05,
      "loss": 0.558297872543335,
      "step": 1741
    },
    {
      "epoch": 1.6465028355387523,
      "grad_norm": 0.3834719808279742,
      "learning_rate": 4.6122493266521877e-05,
      "loss": 0.4790581464767456,
      "step": 1742
    },
    {
      "epoch": 1.6474480151228734,
      "grad_norm": 0.4019338307519857,
      "learning_rate": 4.6070552899945247e-05,
      "loss": 0.5655766725540161,
      "step": 1743
    },
    {
      "epoch": 1.6483931947069943,
      "grad_norm": 0.4011996339114603,
      "learning_rate": 4.601861679971901e-05,
      "loss": 0.5220062732696533,
      "step": 1744
    },
    {
      "epoch": 1.6493383742911152,
      "grad_norm": 0.3769215188258812,
      "learning_rate": 4.596668502223214e-05,
      "loss": 0.4920870363712311,
      "step": 1745
    },
    {
      "epoch": 1.6502835538752363,
      "grad_norm": 0.3649907238176595,
      "learning_rate": 4.591475762386899e-05,
      "loss": 0.4923795461654663,
      "step": 1746
    },
    {
      "epoch": 1.6512287334593574,
      "grad_norm": 0.3338434577833397,
      "learning_rate": 4.5862834661009066e-05,
      "loss": 0.4676467478275299,
      "step": 1747
    },
    {
      "epoch": 1.6521739130434783,
      "grad_norm": 0.3467951918404456,
      "learning_rate": 4.581091619002712e-05,
      "loss": 0.49651557207107544,
      "step": 1748
    },
    {
      "epoch": 1.6531190926275992,
      "grad_norm": 0.3728330223168932,
      "learning_rate": 4.5759002267293036e-05,
      "loss": 0.5564905405044556,
      "step": 1749
    },
    {
      "epoch": 1.6540642722117203,
      "grad_norm": 0.40226535369129807,
      "learning_rate": 4.5707092949171697e-05,
      "loss": 0.5682041049003601,
      "step": 1750
    },
    {
      "epoch": 1.6550094517958414,
      "grad_norm": 0.3773444029248531,
      "learning_rate": 4.5655188292023076e-05,
      "loss": 0.5294817686080933,
      "step": 1751
    },
    {
      "epoch": 1.6559546313799622,
      "grad_norm": 0.3970690698826226,
      "learning_rate": 4.5603288352202e-05,
      "loss": 0.5023767352104187,
      "step": 1752
    },
    {
      "epoch": 1.6568998109640831,
      "grad_norm": 0.38843953327564795,
      "learning_rate": 4.5551393186058213e-05,
      "loss": 0.5259071588516235,
      "step": 1753
    },
    {
      "epoch": 1.6578449905482042,
      "grad_norm": 0.35752522423442235,
      "learning_rate": 4.54995028499363e-05,
      "loss": 0.5002360939979553,
      "step": 1754
    },
    {
      "epoch": 1.658790170132325,
      "grad_norm": 0.3473587981449413,
      "learning_rate": 4.544761740017553e-05,
      "loss": 0.47136080265045166,
      "step": 1755
    },
    {
      "epoch": 1.659735349716446,
      "grad_norm": 0.3893357306904865,
      "learning_rate": 4.5395736893109956e-05,
      "loss": 0.47377949953079224,
      "step": 1756
    },
    {
      "epoch": 1.660680529300567,
      "grad_norm": 0.3415363294415931,
      "learning_rate": 4.534386138506819e-05,
      "loss": 0.4719523787498474,
      "step": 1757
    },
    {
      "epoch": 1.6616257088846882,
      "grad_norm": 0.3409563590830172,
      "learning_rate": 4.529199093237347e-05,
      "loss": 0.5213982462882996,
      "step": 1758
    },
    {
      "epoch": 1.662570888468809,
      "grad_norm": 0.37019928335491153,
      "learning_rate": 4.524012559134352e-05,
      "loss": 0.48131492733955383,
      "step": 1759
    },
    {
      "epoch": 1.66351606805293,
      "grad_norm": 0.3846539657027052,
      "learning_rate": 4.518826541829049e-05,
      "loss": 0.5350781679153442,
      "step": 1760
    },
    {
      "epoch": 1.664461247637051,
      "grad_norm": 0.3294530577071657,
      "learning_rate": 4.513641046952097e-05,
      "loss": 0.41405776143074036,
      "step": 1761
    },
    {
      "epoch": 1.6654064272211722,
      "grad_norm": 0.4294006223865104,
      "learning_rate": 4.508456080133584e-05,
      "loss": 0.5954400300979614,
      "step": 1762
    },
    {
      "epoch": 1.666351606805293,
      "grad_norm": 0.4057347497864783,
      "learning_rate": 4.503271647003028e-05,
      "loss": 0.5525850057601929,
      "step": 1763
    },
    {
      "epoch": 1.667296786389414,
      "grad_norm": 0.3864576170827602,
      "learning_rate": 4.498087753189364e-05,
      "loss": 0.5457656383514404,
      "step": 1764
    },
    {
      "epoch": 1.668241965973535,
      "grad_norm": 0.36254676768357946,
      "learning_rate": 4.492904404320942e-05,
      "loss": 0.4798280894756317,
      "step": 1765
    },
    {
      "epoch": 1.669187145557656,
      "grad_norm": 0.37305651728910266,
      "learning_rate": 4.487721606025522e-05,
      "loss": 0.45974570512771606,
      "step": 1766
    },
    {
      "epoch": 1.6701323251417768,
      "grad_norm": 0.304135280844034,
      "learning_rate": 4.482539363930266e-05,
      "loss": 0.35475054383277893,
      "step": 1767
    },
    {
      "epoch": 1.6710775047258979,
      "grad_norm": 0.3450324506483804,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.4710584878921509,
      "step": 1768
    },
    {
      "epoch": 1.672022684310019,
      "grad_norm": 0.410853127048933,
      "learning_rate": 4.472176570845869e-05,
      "loss": 0.5345973968505859,
      "step": 1769
    },
    {
      "epoch": 1.6729678638941399,
      "grad_norm": 0.3704108946319755,
      "learning_rate": 4.466996031108004e-05,
      "loss": 0.4712327718734741,
      "step": 1770
    },
    {
      "epoch": 1.6739130434782608,
      "grad_norm": 0.3427542278440402,
      "learning_rate": 4.461816070072851e-05,
      "loss": 0.46411386132240295,
      "step": 1771
    },
    {
      "epoch": 1.6748582230623819,
      "grad_norm": 0.3612171943256011,
      "learning_rate": 4.4566366933644845e-05,
      "loss": 0.5078140497207642,
      "step": 1772
    },
    {
      "epoch": 1.675803402646503,
      "grad_norm": 0.37701159560038827,
      "learning_rate": 4.451457906606361e-05,
      "loss": 0.5278763771057129,
      "step": 1773
    },
    {
      "epoch": 1.6767485822306238,
      "grad_norm": 0.3701123408650811,
      "learning_rate": 4.446279715421277e-05,
      "loss": 0.5303045511245728,
      "step": 1774
    },
    {
      "epoch": 1.6776937618147447,
      "grad_norm": 0.38064993761825755,
      "learning_rate": 4.441102125431398e-05,
      "loss": 0.49909886717796326,
      "step": 1775
    },
    {
      "epoch": 1.6786389413988658,
      "grad_norm": 0.3987180362270152,
      "learning_rate": 4.4359251422582284e-05,
      "loss": 0.5773625373840332,
      "step": 1776
    },
    {
      "epoch": 1.679584120982987,
      "grad_norm": 0.3731459366966239,
      "learning_rate": 4.430748771522615e-05,
      "loss": 0.5320460796356201,
      "step": 1777
    },
    {
      "epoch": 1.6805293005671076,
      "grad_norm": 0.3740995580349603,
      "learning_rate": 4.425573018844743e-05,
      "loss": 0.5272402763366699,
      "step": 1778
    },
    {
      "epoch": 1.6814744801512287,
      "grad_norm": 0.38831456750828436,
      "learning_rate": 4.420397889844121e-05,
      "loss": 0.5190265774726868,
      "step": 1779
    },
    {
      "epoch": 1.6824196597353498,
      "grad_norm": 0.38725992055167735,
      "learning_rate": 4.415223390139588e-05,
      "loss": 0.550054669380188,
      "step": 1780
    },
    {
      "epoch": 1.6833648393194707,
      "grad_norm": 0.37578845391396704,
      "learning_rate": 4.4100495253492935e-05,
      "loss": 0.5503447651863098,
      "step": 1781
    },
    {
      "epoch": 1.6843100189035916,
      "grad_norm": 0.4136160240014632,
      "learning_rate": 4.404876301090697e-05,
      "loss": 0.5607806444168091,
      "step": 1782
    },
    {
      "epoch": 1.6852551984877127,
      "grad_norm": 0.3286367463909985,
      "learning_rate": 4.399703722980569e-05,
      "loss": 0.48329243063926697,
      "step": 1783
    },
    {
      "epoch": 1.6862003780718338,
      "grad_norm": 0.34422752207659324,
      "learning_rate": 4.394531796634971e-05,
      "loss": 0.4394565224647522,
      "step": 1784
    },
    {
      "epoch": 1.6871455576559546,
      "grad_norm": 0.3800783688869203,
      "learning_rate": 4.3893605276692646e-05,
      "loss": 0.5124793648719788,
      "step": 1785
    },
    {
      "epoch": 1.6880907372400755,
      "grad_norm": 0.3743033974623709,
      "learning_rate": 4.3841899216980895e-05,
      "loss": 0.5051628351211548,
      "step": 1786
    },
    {
      "epoch": 1.6890359168241966,
      "grad_norm": 0.41775532377651975,
      "learning_rate": 4.379019984335371e-05,
      "loss": 0.5732215642929077,
      "step": 1787
    },
    {
      "epoch": 1.6899810964083177,
      "grad_norm": 0.38780689395508383,
      "learning_rate": 4.373850721194308e-05,
      "loss": 0.5571975708007812,
      "step": 1788
    },
    {
      "epoch": 1.6909262759924386,
      "grad_norm": 0.3616094134523778,
      "learning_rate": 4.368682137887365e-05,
      "loss": 0.5302836894989014,
      "step": 1789
    },
    {
      "epoch": 1.6918714555765595,
      "grad_norm": 0.3186004005227397,
      "learning_rate": 4.3635142400262715e-05,
      "loss": 0.38854703307151794,
      "step": 1790
    },
    {
      "epoch": 1.6928166351606806,
      "grad_norm": 0.37201127358038644,
      "learning_rate": 4.358347033222009e-05,
      "loss": 0.5141817331314087,
      "step": 1791
    },
    {
      "epoch": 1.6937618147448015,
      "grad_norm": 0.3520827532047788,
      "learning_rate": 4.353180523084812e-05,
      "loss": 0.49276185035705566,
      "step": 1792
    },
    {
      "epoch": 1.6947069943289224,
      "grad_norm": 0.34220059547424164,
      "learning_rate": 4.348014715224158e-05,
      "loss": 0.48996010422706604,
      "step": 1793
    },
    {
      "epoch": 1.6956521739130435,
      "grad_norm": 0.36563486013393687,
      "learning_rate": 4.3428496152487576e-05,
      "loss": 0.4645765423774719,
      "step": 1794
    },
    {
      "epoch": 1.6965973534971646,
      "grad_norm": 0.37167827714002943,
      "learning_rate": 4.3376852287665606e-05,
      "loss": 0.4711875915527344,
      "step": 1795
    },
    {
      "epoch": 1.6975425330812854,
      "grad_norm": 0.4124389182952709,
      "learning_rate": 4.332521561384736e-05,
      "loss": 0.5223144292831421,
      "step": 1796
    },
    {
      "epoch": 1.6984877126654063,
      "grad_norm": 0.39110045619047507,
      "learning_rate": 4.327358618709676e-05,
      "loss": 0.571606457233429,
      "step": 1797
    },
    {
      "epoch": 1.6994328922495274,
      "grad_norm": 0.3567219933983256,
      "learning_rate": 4.322196406346984e-05,
      "loss": 0.534321665763855,
      "step": 1798
    },
    {
      "epoch": 1.7003780718336485,
      "grad_norm": 0.3301767532867547,
      "learning_rate": 4.317034929901468e-05,
      "loss": 0.44493192434310913,
      "step": 1799
    },
    {
      "epoch": 1.7013232514177694,
      "grad_norm": 0.35827460117352095,
      "learning_rate": 4.311874194977141e-05,
      "loss": 0.47065895795822144,
      "step": 1800
    },
    {
      "epoch": 1.7022684310018903,
      "grad_norm": 0.354502239398913,
      "learning_rate": 4.306714207177209e-05,
      "loss": 0.5105794668197632,
      "step": 1801
    },
    {
      "epoch": 1.7032136105860114,
      "grad_norm": 0.3929436635782348,
      "learning_rate": 4.30155497210407e-05,
      "loss": 0.5646379590034485,
      "step": 1802
    },
    {
      "epoch": 1.7041587901701323,
      "grad_norm": 0.3591354124214967,
      "learning_rate": 4.296396495359298e-05,
      "loss": 0.5562634468078613,
      "step": 1803
    },
    {
      "epoch": 1.7051039697542532,
      "grad_norm": 0.40947306686688967,
      "learning_rate": 4.291238782543652e-05,
      "loss": 0.5704629421234131,
      "step": 1804
    },
    {
      "epoch": 1.7060491493383743,
      "grad_norm": 0.36041364809726645,
      "learning_rate": 4.2860818392570535e-05,
      "loss": 0.5308239459991455,
      "step": 1805
    },
    {
      "epoch": 1.7069943289224954,
      "grad_norm": 0.3467525202314211,
      "learning_rate": 4.280925671098592e-05,
      "loss": 0.4710659682750702,
      "step": 1806
    },
    {
      "epoch": 1.7079395085066162,
      "grad_norm": 0.37733509187554304,
      "learning_rate": 4.27577028366652e-05,
      "loss": 0.5382436513900757,
      "step": 1807
    },
    {
      "epoch": 1.7088846880907371,
      "grad_norm": 0.36071064716966245,
      "learning_rate": 4.270615682558234e-05,
      "loss": 0.5026828646659851,
      "step": 1808
    },
    {
      "epoch": 1.7098298676748582,
      "grad_norm": 0.39470775095172894,
      "learning_rate": 4.2654618733702825e-05,
      "loss": 0.5538691282272339,
      "step": 1809
    },
    {
      "epoch": 1.7107750472589793,
      "grad_norm": 0.3699280319440355,
      "learning_rate": 4.260308861698351e-05,
      "loss": 0.5432292222976685,
      "step": 1810
    },
    {
      "epoch": 1.7117202268431002,
      "grad_norm": 0.34857083388902127,
      "learning_rate": 4.255156653137261e-05,
      "loss": 0.4534902572631836,
      "step": 1811
    },
    {
      "epoch": 1.712665406427221,
      "grad_norm": 0.3791403030879001,
      "learning_rate": 4.250005253280965e-05,
      "loss": 0.5591314435005188,
      "step": 1812
    },
    {
      "epoch": 1.7136105860113422,
      "grad_norm": 0.3572188120774427,
      "learning_rate": 4.2448546677225275e-05,
      "loss": 0.44667914509773254,
      "step": 1813
    },
    {
      "epoch": 1.7145557655954633,
      "grad_norm": 0.396515161154098,
      "learning_rate": 4.239704902054141e-05,
      "loss": 0.5719116926193237,
      "step": 1814
    },
    {
      "epoch": 1.715500945179584,
      "grad_norm": 0.3908587479857424,
      "learning_rate": 4.234555961867099e-05,
      "loss": 0.5241976976394653,
      "step": 1815
    },
    {
      "epoch": 1.716446124763705,
      "grad_norm": 0.3584291650672085,
      "learning_rate": 4.229407852751806e-05,
      "loss": 0.5234099626541138,
      "step": 1816
    },
    {
      "epoch": 1.7173913043478262,
      "grad_norm": 0.37839800152342834,
      "learning_rate": 4.224260580297755e-05,
      "loss": 0.5431168079376221,
      "step": 1817
    },
    {
      "epoch": 1.718336483931947,
      "grad_norm": 0.39487362848462304,
      "learning_rate": 4.2191141500935375e-05,
      "loss": 0.5428977012634277,
      "step": 1818
    },
    {
      "epoch": 1.719281663516068,
      "grad_norm": 0.3723609667088894,
      "learning_rate": 4.2139685677268305e-05,
      "loss": 0.5216341018676758,
      "step": 1819
    },
    {
      "epoch": 1.720226843100189,
      "grad_norm": 0.41372445617440246,
      "learning_rate": 4.208823838784386e-05,
      "loss": 0.5985276103019714,
      "step": 1820
    },
    {
      "epoch": 1.7211720226843101,
      "grad_norm": 0.31603933032787707,
      "learning_rate": 4.2036799688520375e-05,
      "loss": 0.41243523359298706,
      "step": 1821
    },
    {
      "epoch": 1.722117202268431,
      "grad_norm": 0.3889846945925224,
      "learning_rate": 4.198536963514674e-05,
      "loss": 0.5533255338668823,
      "step": 1822
    },
    {
      "epoch": 1.723062381852552,
      "grad_norm": 0.391078440892704,
      "learning_rate": 4.1933948283562554e-05,
      "loss": 0.5623223781585693,
      "step": 1823
    },
    {
      "epoch": 1.724007561436673,
      "grad_norm": 0.33948571229996843,
      "learning_rate": 4.188253568959795e-05,
      "loss": 0.46665388345718384,
      "step": 1824
    },
    {
      "epoch": 1.724952741020794,
      "grad_norm": 0.32716788447158246,
      "learning_rate": 4.183113190907349e-05,
      "loss": 0.5015628933906555,
      "step": 1825
    },
    {
      "epoch": 1.725897920604915,
      "grad_norm": 0.35334858648269185,
      "learning_rate": 4.177973699780028e-05,
      "loss": 0.47734135389328003,
      "step": 1826
    },
    {
      "epoch": 1.7268431001890359,
      "grad_norm": 0.3716886563813251,
      "learning_rate": 4.172835101157969e-05,
      "loss": 0.5050278902053833,
      "step": 1827
    },
    {
      "epoch": 1.727788279773157,
      "grad_norm": 0.4123532749077132,
      "learning_rate": 4.167697400620345e-05,
      "loss": 0.564073920249939,
      "step": 1828
    },
    {
      "epoch": 1.7287334593572778,
      "grad_norm": 0.3302958202754204,
      "learning_rate": 4.1625606037453554e-05,
      "loss": 0.459759384393692,
      "step": 1829
    },
    {
      "epoch": 1.7296786389413987,
      "grad_norm": 0.36074526612427005,
      "learning_rate": 4.157424716110212e-05,
      "loss": 0.49070411920547485,
      "step": 1830
    },
    {
      "epoch": 1.7306238185255198,
      "grad_norm": 0.3798842544441128,
      "learning_rate": 4.152289743291148e-05,
      "loss": 0.6047898530960083,
      "step": 1831
    },
    {
      "epoch": 1.731568998109641,
      "grad_norm": 0.38066069098278243,
      "learning_rate": 4.147155690863395e-05,
      "loss": 0.574507474899292,
      "step": 1832
    },
    {
      "epoch": 1.7325141776937618,
      "grad_norm": 0.3409535816175999,
      "learning_rate": 4.142022564401193e-05,
      "loss": 0.47886812686920166,
      "step": 1833
    },
    {
      "epoch": 1.7334593572778827,
      "grad_norm": 0.3641491006094355,
      "learning_rate": 4.136890369477772e-05,
      "loss": 0.4815311133861542,
      "step": 1834
    },
    {
      "epoch": 1.7344045368620038,
      "grad_norm": 0.361243214585194,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.530300498008728,
      "step": 1835
    },
    {
      "epoch": 1.735349716446125,
      "grad_norm": 0.4064047687480022,
      "learning_rate": 4.126628796535128e-05,
      "loss": 0.5591949224472046,
      "step": 1836
    },
    {
      "epoch": 1.7362948960302458,
      "grad_norm": 0.36825344423008854,
      "learning_rate": 4.121499429657287e-05,
      "loss": 0.5020169019699097,
      "step": 1837
    },
    {
      "epoch": 1.7372400756143667,
      "grad_norm": 0.3703356122325229,
      "learning_rate": 4.116371016600977e-05,
      "loss": 0.5189231038093567,
      "step": 1838
    },
    {
      "epoch": 1.7381852551984878,
      "grad_norm": 0.39807573462593526,
      "learning_rate": 4.1112435629343086e-05,
      "loss": 0.4998430013656616,
      "step": 1839
    },
    {
      "epoch": 1.7391304347826086,
      "grad_norm": 0.41380244935906996,
      "learning_rate": 4.106117074224354e-05,
      "loss": 0.5191060304641724,
      "step": 1840
    },
    {
      "epoch": 1.7400756143667295,
      "grad_norm": 0.35303137255460126,
      "learning_rate": 4.100991556037138e-05,
      "loss": 0.4858666956424713,
      "step": 1841
    },
    {
      "epoch": 1.7410207939508506,
      "grad_norm": 0.38538214714250274,
      "learning_rate": 4.0958670139376306e-05,
      "loss": 0.5013076663017273,
      "step": 1842
    },
    {
      "epoch": 1.7419659735349717,
      "grad_norm": 0.35572187723043486,
      "learning_rate": 4.090743453489744e-05,
      "loss": 0.4707403779029846,
      "step": 1843
    },
    {
      "epoch": 1.7429111531190926,
      "grad_norm": 0.3563252887145492,
      "learning_rate": 4.0856208802563195e-05,
      "loss": 0.4417616128921509,
      "step": 1844
    },
    {
      "epoch": 1.7438563327032135,
      "grad_norm": 0.4143411008298073,
      "learning_rate": 4.080499299799133e-05,
      "loss": 0.5140938758850098,
      "step": 1845
    },
    {
      "epoch": 1.7448015122873346,
      "grad_norm": 0.40409765239705975,
      "learning_rate": 4.0753787176788794e-05,
      "loss": 0.5408952236175537,
      "step": 1846
    },
    {
      "epoch": 1.7457466918714557,
      "grad_norm": 0.4042159459701186,
      "learning_rate": 4.0702591394551654e-05,
      "loss": 0.568861722946167,
      "step": 1847
    },
    {
      "epoch": 1.7466918714555766,
      "grad_norm": 0.36072347459124127,
      "learning_rate": 4.0651405706865185e-05,
      "loss": 0.5176217555999756,
      "step": 1848
    },
    {
      "epoch": 1.7476370510396975,
      "grad_norm": 0.3710712386652127,
      "learning_rate": 4.060023016930359e-05,
      "loss": 0.5140879154205322,
      "step": 1849
    },
    {
      "epoch": 1.7485822306238186,
      "grad_norm": 0.3874401040008372,
      "learning_rate": 4.0549064837430124e-05,
      "loss": 0.5415709614753723,
      "step": 1850
    },
    {
      "epoch": 1.7495274102079397,
      "grad_norm": 0.37776154741241175,
      "learning_rate": 4.049790976679694e-05,
      "loss": 0.5539703965187073,
      "step": 1851
    },
    {
      "epoch": 1.7504725897920603,
      "grad_norm": 0.3714408638276484,
      "learning_rate": 4.0446765012945006e-05,
      "loss": 0.5227164030075073,
      "step": 1852
    },
    {
      "epoch": 1.7514177693761814,
      "grad_norm": 0.36794294139769224,
      "learning_rate": 4.039563063140418e-05,
      "loss": 0.533928632736206,
      "step": 1853
    },
    {
      "epoch": 1.7523629489603025,
      "grad_norm": 0.4066023196520573,
      "learning_rate": 4.0344506677692965e-05,
      "loss": 0.5687515735626221,
      "step": 1854
    },
    {
      "epoch": 1.7533081285444234,
      "grad_norm": 0.4014264331971704,
      "learning_rate": 4.029339320731862e-05,
      "loss": 0.5196583271026611,
      "step": 1855
    },
    {
      "epoch": 1.7542533081285443,
      "grad_norm": 0.3877705146282353,
      "learning_rate": 4.0242290275776986e-05,
      "loss": 0.4936404526233673,
      "step": 1856
    },
    {
      "epoch": 1.7551984877126654,
      "grad_norm": 0.35425653507071464,
      "learning_rate": 4.019119793855242e-05,
      "loss": 0.48280808329582214,
      "step": 1857
    },
    {
      "epoch": 1.7561436672967865,
      "grad_norm": 0.36115472193571335,
      "learning_rate": 4.014011625111787e-05,
      "loss": 0.450501412153244,
      "step": 1858
    },
    {
      "epoch": 1.7570888468809074,
      "grad_norm": 0.38224478869362244,
      "learning_rate": 4.008904526893462e-05,
      "loss": 0.5394353866577148,
      "step": 1859
    },
    {
      "epoch": 1.7580340264650283,
      "grad_norm": 0.37814045444908984,
      "learning_rate": 4.003798504745243e-05,
      "loss": 0.4764851927757263,
      "step": 1860
    },
    {
      "epoch": 1.7589792060491494,
      "grad_norm": 0.4158210872282914,
      "learning_rate": 3.998693564210929e-05,
      "loss": 0.5808486938476562,
      "step": 1861
    },
    {
      "epoch": 1.7599243856332705,
      "grad_norm": 0.35893911412272106,
      "learning_rate": 3.993589710833151e-05,
      "loss": 0.5279654860496521,
      "step": 1862
    },
    {
      "epoch": 1.7608695652173914,
      "grad_norm": 0.383276405178071,
      "learning_rate": 3.988486950153355e-05,
      "loss": 0.5639195442199707,
      "step": 1863
    },
    {
      "epoch": 1.7618147448015122,
      "grad_norm": 0.40438387370669954,
      "learning_rate": 3.9833852877118026e-05,
      "loss": 0.6004071831703186,
      "step": 1864
    },
    {
      "epoch": 1.7627599243856333,
      "grad_norm": 0.3545403990731361,
      "learning_rate": 3.978284729047567e-05,
      "loss": 0.49227917194366455,
      "step": 1865
    },
    {
      "epoch": 1.7637051039697542,
      "grad_norm": 0.377511339751389,
      "learning_rate": 3.973185279698512e-05,
      "loss": 0.515723466873169,
      "step": 1866
    },
    {
      "epoch": 1.764650283553875,
      "grad_norm": 0.3291208651155964,
      "learning_rate": 3.96808694520131e-05,
      "loss": 0.47791755199432373,
      "step": 1867
    },
    {
      "epoch": 1.7655954631379962,
      "grad_norm": 0.3671925594907195,
      "learning_rate": 3.962989731091415e-05,
      "loss": 0.4971652030944824,
      "step": 1868
    },
    {
      "epoch": 1.7665406427221173,
      "grad_norm": 0.3569368735856663,
      "learning_rate": 3.957893642903064e-05,
      "loss": 0.4350050091743469,
      "step": 1869
    },
    {
      "epoch": 1.7674858223062382,
      "grad_norm": 0.40061649332146926,
      "learning_rate": 3.952798686169279e-05,
      "loss": 0.5064713954925537,
      "step": 1870
    },
    {
      "epoch": 1.768431001890359,
      "grad_norm": 0.37243628321719824,
      "learning_rate": 3.947704866421844e-05,
      "loss": 0.4504944384098053,
      "step": 1871
    },
    {
      "epoch": 1.7693761814744802,
      "grad_norm": 0.38308029432146445,
      "learning_rate": 3.942612189191316e-05,
      "loss": 0.5629781484603882,
      "step": 1872
    },
    {
      "epoch": 1.7703213610586013,
      "grad_norm": 0.4063171034823784,
      "learning_rate": 3.937520660007009e-05,
      "loss": 0.5550225973129272,
      "step": 1873
    },
    {
      "epoch": 1.7712665406427222,
      "grad_norm": 0.3557586195226364,
      "learning_rate": 3.932430284396986e-05,
      "loss": 0.49868184328079224,
      "step": 1874
    },
    {
      "epoch": 1.772211720226843,
      "grad_norm": 0.35516774802893364,
      "learning_rate": 3.9273410678880654e-05,
      "loss": 0.5359030365943909,
      "step": 1875
    },
    {
      "epoch": 1.7731568998109641,
      "grad_norm": 0.39060331939109316,
      "learning_rate": 3.9222530160058024e-05,
      "loss": 0.4805881679058075,
      "step": 1876
    },
    {
      "epoch": 1.774102079395085,
      "grad_norm": 0.3591458818631869,
      "learning_rate": 3.917166134274489e-05,
      "loss": 0.5509016513824463,
      "step": 1877
    },
    {
      "epoch": 1.775047258979206,
      "grad_norm": 0.37300988196011586,
      "learning_rate": 3.912080428217146e-05,
      "loss": 0.5417537689208984,
      "step": 1878
    },
    {
      "epoch": 1.775992438563327,
      "grad_norm": 0.3615252837081343,
      "learning_rate": 3.9069959033555195e-05,
      "loss": 0.486447811126709,
      "step": 1879
    },
    {
      "epoch": 1.776937618147448,
      "grad_norm": 0.3818405682772109,
      "learning_rate": 3.901912565210071e-05,
      "loss": 0.5491993427276611,
      "step": 1880
    },
    {
      "epoch": 1.777882797731569,
      "grad_norm": 0.353826774666077,
      "learning_rate": 3.896830419299972e-05,
      "loss": 0.5140633583068848,
      "step": 1881
    },
    {
      "epoch": 1.7788279773156899,
      "grad_norm": 0.39235230350483813,
      "learning_rate": 3.891749471143106e-05,
      "loss": 0.5402032732963562,
      "step": 1882
    },
    {
      "epoch": 1.779773156899811,
      "grad_norm": 0.36702782508907084,
      "learning_rate": 3.886669726256048e-05,
      "loss": 0.5155764818191528,
      "step": 1883
    },
    {
      "epoch": 1.780718336483932,
      "grad_norm": 0.3967201165805106,
      "learning_rate": 3.8815911901540736e-05,
      "loss": 0.5403751134872437,
      "step": 1884
    },
    {
      "epoch": 1.781663516068053,
      "grad_norm": 0.37059841473770694,
      "learning_rate": 3.876513868351142e-05,
      "loss": 0.5415762662887573,
      "step": 1885
    },
    {
      "epoch": 1.7826086956521738,
      "grad_norm": 0.3774908349433492,
      "learning_rate": 3.871437766359893e-05,
      "loss": 0.5587754249572754,
      "step": 1886
    },
    {
      "epoch": 1.783553875236295,
      "grad_norm": 0.3514397369430103,
      "learning_rate": 3.866362889691647e-05,
      "loss": 0.5297296047210693,
      "step": 1887
    },
    {
      "epoch": 1.784499054820416,
      "grad_norm": 0.3663755136925369,
      "learning_rate": 3.861289243856388e-05,
      "loss": 0.5308074951171875,
      "step": 1888
    },
    {
      "epoch": 1.7854442344045367,
      "grad_norm": 0.39703855819895734,
      "learning_rate": 3.856216834362769e-05,
      "loss": 0.5443047285079956,
      "step": 1889
    },
    {
      "epoch": 1.7863894139886578,
      "grad_norm": 0.3469766791565685,
      "learning_rate": 3.851145666718095e-05,
      "loss": 0.4986753761768341,
      "step": 1890
    },
    {
      "epoch": 1.787334593572779,
      "grad_norm": 0.37905633757222684,
      "learning_rate": 3.84607574642833e-05,
      "loss": 0.5435605049133301,
      "step": 1891
    },
    {
      "epoch": 1.7882797731568998,
      "grad_norm": 0.36182109226411513,
      "learning_rate": 3.841007078998077e-05,
      "loss": 0.47318175435066223,
      "step": 1892
    },
    {
      "epoch": 1.7892249527410207,
      "grad_norm": 0.3843212733541874,
      "learning_rate": 3.835939669930581e-05,
      "loss": 0.5354818105697632,
      "step": 1893
    },
    {
      "epoch": 1.7901701323251418,
      "grad_norm": 0.38099875884344975,
      "learning_rate": 3.830873524727722e-05,
      "loss": 0.49623480439186096,
      "step": 1894
    },
    {
      "epoch": 1.7911153119092629,
      "grad_norm": 0.40289714581154934,
      "learning_rate": 3.825808648890005e-05,
      "loss": 0.5701708793640137,
      "step": 1895
    },
    {
      "epoch": 1.7920604914933838,
      "grad_norm": 0.3396892042036407,
      "learning_rate": 3.820745047916563e-05,
      "loss": 0.4532066583633423,
      "step": 1896
    },
    {
      "epoch": 1.7930056710775046,
      "grad_norm": 0.385874175512156,
      "learning_rate": 3.8156827273051365e-05,
      "loss": 0.5030742287635803,
      "step": 1897
    },
    {
      "epoch": 1.7939508506616257,
      "grad_norm": 0.4039021634883295,
      "learning_rate": 3.81062169255208e-05,
      "loss": 0.5029819011688232,
      "step": 1898
    },
    {
      "epoch": 1.7948960302457468,
      "grad_norm": 0.4037874400907923,
      "learning_rate": 3.8055619491523543e-05,
      "loss": 0.5149720311164856,
      "step": 1899
    },
    {
      "epoch": 1.7958412098298677,
      "grad_norm": 0.3908075896617147,
      "learning_rate": 3.8005035025995104e-05,
      "loss": 0.5275954604148865,
      "step": 1900
    },
    {
      "epoch": 1.7967863894139886,
      "grad_norm": 0.3897098159800907,
      "learning_rate": 3.7954463583857035e-05,
      "loss": 0.5385657548904419,
      "step": 1901
    },
    {
      "epoch": 1.7977315689981097,
      "grad_norm": 0.39506657991108074,
      "learning_rate": 3.790390522001662e-05,
      "loss": 0.5252353549003601,
      "step": 1902
    },
    {
      "epoch": 1.7986767485822306,
      "grad_norm": 0.35380268891106303,
      "learning_rate": 3.785335998936702e-05,
      "loss": 0.49155211448669434,
      "step": 1903
    },
    {
      "epoch": 1.7996219281663515,
      "grad_norm": 0.3757058460849153,
      "learning_rate": 3.7802827946787134e-05,
      "loss": 0.5049294233322144,
      "step": 1904
    },
    {
      "epoch": 1.8005671077504726,
      "grad_norm": 0.349884873983796,
      "learning_rate": 3.775230914714149e-05,
      "loss": 0.44136887788772583,
      "step": 1905
    },
    {
      "epoch": 1.8015122873345937,
      "grad_norm": 0.39810633962826625,
      "learning_rate": 3.770180364528029e-05,
      "loss": 0.5395088791847229,
      "step": 1906
    },
    {
      "epoch": 1.8024574669187146,
      "grad_norm": 0.40372349919576656,
      "learning_rate": 3.7651311496039285e-05,
      "loss": 0.5143775939941406,
      "step": 1907
    },
    {
      "epoch": 1.8034026465028354,
      "grad_norm": 0.3580267693619654,
      "learning_rate": 3.7600832754239724e-05,
      "loss": 0.4764373004436493,
      "step": 1908
    },
    {
      "epoch": 1.8043478260869565,
      "grad_norm": 0.3897373491275009,
      "learning_rate": 3.755036747468831e-05,
      "loss": 0.5054662823677063,
      "step": 1909
    },
    {
      "epoch": 1.8052930056710776,
      "grad_norm": 0.37643828613294383,
      "learning_rate": 3.7499915712177094e-05,
      "loss": 0.502275824546814,
      "step": 1910
    },
    {
      "epoch": 1.8062381852551985,
      "grad_norm": 0.34773633897706785,
      "learning_rate": 3.744947752148349e-05,
      "loss": 0.49097099900245667,
      "step": 1911
    },
    {
      "epoch": 1.8071833648393194,
      "grad_norm": 0.390110351607708,
      "learning_rate": 3.739905295737015e-05,
      "loss": 0.5609234571456909,
      "step": 1912
    },
    {
      "epoch": 1.8081285444234405,
      "grad_norm": 0.4404646707324305,
      "learning_rate": 3.734864207458497e-05,
      "loss": 0.5814790725708008,
      "step": 1913
    },
    {
      "epoch": 1.8090737240075614,
      "grad_norm": 0.3595006350094641,
      "learning_rate": 3.729824492786095e-05,
      "loss": 0.46478211879730225,
      "step": 1914
    },
    {
      "epoch": 1.8100189035916823,
      "grad_norm": 0.3894498445164366,
      "learning_rate": 3.7247861571916185e-05,
      "loss": 0.5252106189727783,
      "step": 1915
    },
    {
      "epoch": 1.8109640831758034,
      "grad_norm": 0.4115843226077555,
      "learning_rate": 3.7197492061453815e-05,
      "loss": 0.5381331443786621,
      "step": 1916
    },
    {
      "epoch": 1.8119092627599245,
      "grad_norm": 0.3811864410540255,
      "learning_rate": 3.714713645116192e-05,
      "loss": 0.5281251668930054,
      "step": 1917
    },
    {
      "epoch": 1.8128544423440454,
      "grad_norm": 0.4215782826674777,
      "learning_rate": 3.709679479571354e-05,
      "loss": 0.5507755279541016,
      "step": 1918
    },
    {
      "epoch": 1.8137996219281662,
      "grad_norm": 0.37121161510016026,
      "learning_rate": 3.7046467149766485e-05,
      "loss": 0.5750713348388672,
      "step": 1919
    },
    {
      "epoch": 1.8147448015122873,
      "grad_norm": 0.45168208753640726,
      "learning_rate": 3.699615356796342e-05,
      "loss": 0.6803639531135559,
      "step": 1920
    },
    {
      "epoch": 1.8156899810964084,
      "grad_norm": 0.3413773669091061,
      "learning_rate": 3.6945854104931725e-05,
      "loss": 0.4612014889717102,
      "step": 1921
    },
    {
      "epoch": 1.8166351606805293,
      "grad_norm": 0.387652311770123,
      "learning_rate": 3.68955688152834e-05,
      "loss": 0.5055382251739502,
      "step": 1922
    },
    {
      "epoch": 1.8175803402646502,
      "grad_norm": 0.37844032831460256,
      "learning_rate": 3.684529775361516e-05,
      "loss": 0.48959559202194214,
      "step": 1923
    },
    {
      "epoch": 1.8185255198487713,
      "grad_norm": 0.4105630139342899,
      "learning_rate": 3.679504097450816e-05,
      "loss": 0.5319094657897949,
      "step": 1924
    },
    {
      "epoch": 1.8194706994328924,
      "grad_norm": 0.3930680171450619,
      "learning_rate": 3.674479853252813e-05,
      "loss": 0.5142064690589905,
      "step": 1925
    },
    {
      "epoch": 1.820415879017013,
      "grad_norm": 0.37835977539607485,
      "learning_rate": 3.669457048222521e-05,
      "loss": 0.5136705636978149,
      "step": 1926
    },
    {
      "epoch": 1.8213610586011342,
      "grad_norm": 0.372950701960622,
      "learning_rate": 3.664435687813386e-05,
      "loss": 0.518903374671936,
      "step": 1927
    },
    {
      "epoch": 1.8223062381852553,
      "grad_norm": 0.3825870517920147,
      "learning_rate": 3.659415777477295e-05,
      "loss": 0.5177056789398193,
      "step": 1928
    },
    {
      "epoch": 1.8232514177693762,
      "grad_norm": 0.3615629582980017,
      "learning_rate": 3.6543973226645526e-05,
      "loss": 0.5147405862808228,
      "step": 1929
    },
    {
      "epoch": 1.824196597353497,
      "grad_norm": 0.3471157988183404,
      "learning_rate": 3.6493803288238896e-05,
      "loss": 0.4677199721336365,
      "step": 1930
    },
    {
      "epoch": 1.8251417769376181,
      "grad_norm": 0.38850614283427126,
      "learning_rate": 3.644364801402446e-05,
      "loss": 0.5410925149917603,
      "step": 1931
    },
    {
      "epoch": 1.8260869565217392,
      "grad_norm": 0.37696722000145594,
      "learning_rate": 3.639350745845769e-05,
      "loss": 0.5271806716918945,
      "step": 1932
    },
    {
      "epoch": 1.8270321361058601,
      "grad_norm": 0.3772362851929242,
      "learning_rate": 3.6343381675978116e-05,
      "loss": 0.5433135032653809,
      "step": 1933
    },
    {
      "epoch": 1.827977315689981,
      "grad_norm": 0.367565064253662,
      "learning_rate": 3.62932707210092e-05,
      "loss": 0.4963024854660034,
      "step": 1934
    },
    {
      "epoch": 1.8289224952741021,
      "grad_norm": 0.40954463259200485,
      "learning_rate": 3.624317464795834e-05,
      "loss": 0.5469388961791992,
      "step": 1935
    },
    {
      "epoch": 1.8298676748582232,
      "grad_norm": 0.33471658848393576,
      "learning_rate": 3.61930935112167e-05,
      "loss": 0.41202718019485474,
      "step": 1936
    },
    {
      "epoch": 1.830812854442344,
      "grad_norm": 0.37549816825209265,
      "learning_rate": 3.614302736515933e-05,
      "loss": 0.47775477170944214,
      "step": 1937
    },
    {
      "epoch": 1.831758034026465,
      "grad_norm": 0.35837760252931705,
      "learning_rate": 3.609297626414493e-05,
      "loss": 0.4989203214645386,
      "step": 1938
    },
    {
      "epoch": 1.832703213610586,
      "grad_norm": 0.399683251162192,
      "learning_rate": 3.604294026251587e-05,
      "loss": 0.5551716089248657,
      "step": 1939
    },
    {
      "epoch": 1.833648393194707,
      "grad_norm": 0.3653387669377067,
      "learning_rate": 3.599291941459818e-05,
      "loss": 0.4233468174934387,
      "step": 1940
    },
    {
      "epoch": 1.8345935727788278,
      "grad_norm": 0.43718141602016236,
      "learning_rate": 3.594291377470136e-05,
      "loss": 0.6098648309707642,
      "step": 1941
    },
    {
      "epoch": 1.835538752362949,
      "grad_norm": 0.5709741983606528,
      "learning_rate": 3.5892923397118474e-05,
      "loss": 0.5647149085998535,
      "step": 1942
    },
    {
      "epoch": 1.83648393194707,
      "grad_norm": 0.3634589621555628,
      "learning_rate": 3.584294833612596e-05,
      "loss": 0.5058320164680481,
      "step": 1943
    },
    {
      "epoch": 1.837429111531191,
      "grad_norm": 0.40697030158815667,
      "learning_rate": 3.579298864598362e-05,
      "loss": 0.5641506910324097,
      "step": 1944
    },
    {
      "epoch": 1.8383742911153118,
      "grad_norm": 0.3500659898292031,
      "learning_rate": 3.574304438093466e-05,
      "loss": 0.4629972577095032,
      "step": 1945
    },
    {
      "epoch": 1.839319470699433,
      "grad_norm": 0.3794521731692319,
      "learning_rate": 3.5693115595205397e-05,
      "loss": 0.47894200682640076,
      "step": 1946
    },
    {
      "epoch": 1.840264650283554,
      "grad_norm": 0.34048490792376335,
      "learning_rate": 3.564320234300545e-05,
      "loss": 0.49541962146759033,
      "step": 1947
    },
    {
      "epoch": 1.841209829867675,
      "grad_norm": 0.39136233694047096,
      "learning_rate": 3.559330467852755e-05,
      "loss": 0.48214370012283325,
      "step": 1948
    },
    {
      "epoch": 1.8421550094517958,
      "grad_norm": 0.3644537290736626,
      "learning_rate": 3.554342265594748e-05,
      "loss": 0.5075618028640747,
      "step": 1949
    },
    {
      "epoch": 1.8431001890359169,
      "grad_norm": 0.42289893731687656,
      "learning_rate": 3.549355632942405e-05,
      "loss": 0.5777016878128052,
      "step": 1950
    },
    {
      "epoch": 1.8440453686200378,
      "grad_norm": 0.34962992253569974,
      "learning_rate": 3.5443705753099014e-05,
      "loss": 0.4347735047340393,
      "step": 1951
    },
    {
      "epoch": 1.8449905482041586,
      "grad_norm": 0.3716510904097844,
      "learning_rate": 3.539387098109707e-05,
      "loss": 0.475225567817688,
      "step": 1952
    },
    {
      "epoch": 1.8459357277882797,
      "grad_norm": 0.4012886913214644,
      "learning_rate": 3.53440520675257e-05,
      "loss": 0.5743262767791748,
      "step": 1953
    },
    {
      "epoch": 1.8468809073724008,
      "grad_norm": 0.34005023233534387,
      "learning_rate": 3.5294249066475246e-05,
      "loss": 0.500430166721344,
      "step": 1954
    },
    {
      "epoch": 1.8478260869565217,
      "grad_norm": 0.3818488572591982,
      "learning_rate": 3.5244462032018666e-05,
      "loss": 0.5386780500411987,
      "step": 1955
    },
    {
      "epoch": 1.8487712665406426,
      "grad_norm": 0.3728378490632904,
      "learning_rate": 3.5194691018211655e-05,
      "loss": 0.5297654867172241,
      "step": 1956
    },
    {
      "epoch": 1.8497164461247637,
      "grad_norm": 0.3535809875837171,
      "learning_rate": 3.514493607909253e-05,
      "loss": 0.45356065034866333,
      "step": 1957
    },
    {
      "epoch": 1.8506616257088848,
      "grad_norm": 0.39534648416980045,
      "learning_rate": 3.509519726868207e-05,
      "loss": 0.4880695343017578,
      "step": 1958
    },
    {
      "epoch": 1.8516068052930057,
      "grad_norm": 0.369039953708234,
      "learning_rate": 3.504547464098364e-05,
      "loss": 0.5010110139846802,
      "step": 1959
    },
    {
      "epoch": 1.8525519848771266,
      "grad_norm": 0.37130949510818245,
      "learning_rate": 3.499576824998298e-05,
      "loss": 0.5087857246398926,
      "step": 1960
    },
    {
      "epoch": 1.8534971644612477,
      "grad_norm": 0.3536804557996416,
      "learning_rate": 3.4946078149648184e-05,
      "loss": 0.44036608934402466,
      "step": 1961
    },
    {
      "epoch": 1.8544423440453688,
      "grad_norm": 0.3522787315482795,
      "learning_rate": 3.489640439392974e-05,
      "loss": 0.44712895154953003,
      "step": 1962
    },
    {
      "epoch": 1.8553875236294894,
      "grad_norm": 0.36885932521545844,
      "learning_rate": 3.484674703676028e-05,
      "loss": 0.44689446687698364,
      "step": 1963
    },
    {
      "epoch": 1.8563327032136105,
      "grad_norm": 0.3848234583504505,
      "learning_rate": 3.4797106132054736e-05,
      "loss": 0.5131386518478394,
      "step": 1964
    },
    {
      "epoch": 1.8572778827977316,
      "grad_norm": 0.37529353795464393,
      "learning_rate": 3.474748173371008e-05,
      "loss": 0.49377599358558655,
      "step": 1965
    },
    {
      "epoch": 1.8582230623818525,
      "grad_norm": 0.37676261891666457,
      "learning_rate": 3.469787389560547e-05,
      "loss": 0.46099764108657837,
      "step": 1966
    },
    {
      "epoch": 1.8591682419659734,
      "grad_norm": 0.3357591863055808,
      "learning_rate": 3.464828267160199e-05,
      "loss": 0.4685811996459961,
      "step": 1967
    },
    {
      "epoch": 1.8601134215500945,
      "grad_norm": 0.38133935111171896,
      "learning_rate": 3.4598708115542714e-05,
      "loss": 0.5538822412490845,
      "step": 1968
    },
    {
      "epoch": 1.8610586011342156,
      "grad_norm": 0.3989743175786868,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.5005359649658203,
      "step": 1969
    },
    {
      "epoch": 1.8620037807183365,
      "grad_norm": 0.37008068706126274,
      "learning_rate": 3.4499609222538576e-05,
      "loss": 0.5179316997528076,
      "step": 1970
    },
    {
      "epoch": 1.8629489603024574,
      "grad_norm": 0.38101852856080126,
      "learning_rate": 3.445008499318918e-05,
      "loss": 0.5478320717811584,
      "step": 1971
    },
    {
      "epoch": 1.8638941398865785,
      "grad_norm": 0.348546733052073,
      "learning_rate": 3.4400577646974764e-05,
      "loss": 0.5230755805969238,
      "step": 1972
    },
    {
      "epoch": 1.8648393194706996,
      "grad_norm": 0.3659727475548959,
      "learning_rate": 3.435108723764732e-05,
      "loss": 0.5353562831878662,
      "step": 1973
    },
    {
      "epoch": 1.8657844990548205,
      "grad_norm": 0.36396607202042003,
      "learning_rate": 3.430161381894051e-05,
      "loss": 0.4596851170063019,
      "step": 1974
    },
    {
      "epoch": 1.8667296786389413,
      "grad_norm": 0.34524145843494,
      "learning_rate": 3.425215744456948e-05,
      "loss": 0.5208250284194946,
      "step": 1975
    },
    {
      "epoch": 1.8676748582230625,
      "grad_norm": 0.3876847819764721,
      "learning_rate": 3.420271816823093e-05,
      "loss": 0.52238929271698,
      "step": 1976
    },
    {
      "epoch": 1.8686200378071833,
      "grad_norm": 0.40582119093724167,
      "learning_rate": 3.415329604360293e-05,
      "loss": 0.5721954107284546,
      "step": 1977
    },
    {
      "epoch": 1.8695652173913042,
      "grad_norm": 0.3789995181508255,
      "learning_rate": 3.410389112434499e-05,
      "loss": 0.4861782491207123,
      "step": 1978
    },
    {
      "epoch": 1.8705103969754253,
      "grad_norm": 0.3580278497851857,
      "learning_rate": 3.405450346409791e-05,
      "loss": 0.5074586272239685,
      "step": 1979
    },
    {
      "epoch": 1.8714555765595464,
      "grad_norm": 0.389140221975222,
      "learning_rate": 3.400513311648372e-05,
      "loss": 0.548485279083252,
      "step": 1980
    },
    {
      "epoch": 1.8724007561436673,
      "grad_norm": 0.36669306145988323,
      "learning_rate": 3.3955780135105736e-05,
      "loss": 0.5294705629348755,
      "step": 1981
    },
    {
      "epoch": 1.8733459357277882,
      "grad_norm": 0.38935265119702006,
      "learning_rate": 3.390644457354833e-05,
      "loss": 0.5309770703315735,
      "step": 1982
    },
    {
      "epoch": 1.8742911153119093,
      "grad_norm": 0.38868859284047735,
      "learning_rate": 3.385712648537703e-05,
      "loss": 0.562687337398529,
      "step": 1983
    },
    {
      "epoch": 1.8752362948960304,
      "grad_norm": 0.3587298064339353,
      "learning_rate": 3.380782592413836e-05,
      "loss": 0.5167361497879028,
      "step": 1984
    },
    {
      "epoch": 1.8761814744801513,
      "grad_norm": 0.38422090382619134,
      "learning_rate": 3.375854294335977e-05,
      "loss": 0.5124733448028564,
      "step": 1985
    },
    {
      "epoch": 1.8771266540642721,
      "grad_norm": 0.36848191062018854,
      "learning_rate": 3.3709277596549736e-05,
      "loss": 0.49587374925613403,
      "step": 1986
    },
    {
      "epoch": 1.8780718336483933,
      "grad_norm": 0.3935372489361954,
      "learning_rate": 3.366002993719747e-05,
      "loss": 0.4826546013355255,
      "step": 1987
    },
    {
      "epoch": 1.8790170132325141,
      "grad_norm": 0.3767110218659958,
      "learning_rate": 3.361080001877308e-05,
      "loss": 0.4593847990036011,
      "step": 1988
    },
    {
      "epoch": 1.879962192816635,
      "grad_norm": 0.41357049741420365,
      "learning_rate": 3.356158789472732e-05,
      "loss": 0.5514704585075378,
      "step": 1989
    },
    {
      "epoch": 1.8809073724007561,
      "grad_norm": 0.3406845865701544,
      "learning_rate": 3.3512393618491676e-05,
      "loss": 0.49037793278694153,
      "step": 1990
    },
    {
      "epoch": 1.8818525519848772,
      "grad_norm": 0.43723170156871166,
      "learning_rate": 3.346321724347827e-05,
      "loss": 0.6320779323577881,
      "step": 1991
    },
    {
      "epoch": 1.882797731568998,
      "grad_norm": 0.36630149684930174,
      "learning_rate": 3.341405882307973e-05,
      "loss": 0.48896312713623047,
      "step": 1992
    },
    {
      "epoch": 1.883742911153119,
      "grad_norm": 0.3410745442997026,
      "learning_rate": 3.336491841066928e-05,
      "loss": 0.4347347021102905,
      "step": 1993
    },
    {
      "epoch": 1.88468809073724,
      "grad_norm": 0.39329153890851853,
      "learning_rate": 3.331579605960047e-05,
      "loss": 0.5370204448699951,
      "step": 1994
    },
    {
      "epoch": 1.8856332703213612,
      "grad_norm": 0.37550504205729923,
      "learning_rate": 3.326669182320736e-05,
      "loss": 0.46477124094963074,
      "step": 1995
    },
    {
      "epoch": 1.886578449905482,
      "grad_norm": 0.3522578042200325,
      "learning_rate": 3.321760575480427e-05,
      "loss": 0.4580044746398926,
      "step": 1996
    },
    {
      "epoch": 1.887523629489603,
      "grad_norm": 0.370177815554421,
      "learning_rate": 3.316853790768579e-05,
      "loss": 0.5711253881454468,
      "step": 1997
    },
    {
      "epoch": 1.888468809073724,
      "grad_norm": 0.4362293735605827,
      "learning_rate": 3.311948833512681e-05,
      "loss": 0.48161938786506653,
      "step": 1998
    },
    {
      "epoch": 1.8894139886578452,
      "grad_norm": 0.36310839489607916,
      "learning_rate": 3.307045709038226e-05,
      "loss": 0.507922351360321,
      "step": 1999
    },
    {
      "epoch": 1.8903591682419658,
      "grad_norm": 0.39930311515350314,
      "learning_rate": 3.302144422668726e-05,
      "loss": 0.5756919384002686,
      "step": 2000
    },
    {
      "epoch": 1.891304347826087,
      "grad_norm": 0.31616134030946663,
      "learning_rate": 3.297244979725695e-05,
      "loss": 0.39018499851226807,
      "step": 2001
    },
    {
      "epoch": 1.892249527410208,
      "grad_norm": 0.34684183008903924,
      "learning_rate": 3.2923473855286426e-05,
      "loss": 0.485091894865036,
      "step": 2002
    },
    {
      "epoch": 1.893194706994329,
      "grad_norm": 0.35884992663417997,
      "learning_rate": 3.287451645395074e-05,
      "loss": 0.525263786315918,
      "step": 2003
    },
    {
      "epoch": 1.8941398865784498,
      "grad_norm": 0.3130602970273822,
      "learning_rate": 3.28255776464048e-05,
      "loss": 0.45005786418914795,
      "step": 2004
    },
    {
      "epoch": 1.8950850661625709,
      "grad_norm": 0.36572316751165906,
      "learning_rate": 3.277665748578336e-05,
      "loss": 0.47230541706085205,
      "step": 2005
    },
    {
      "epoch": 1.896030245746692,
      "grad_norm": 0.3723367828589821,
      "learning_rate": 3.2727756025200865e-05,
      "loss": 0.5426722168922424,
      "step": 2006
    },
    {
      "epoch": 1.8969754253308129,
      "grad_norm": 0.36841151450991005,
      "learning_rate": 3.2678873317751545e-05,
      "loss": 0.5338841676712036,
      "step": 2007
    },
    {
      "epoch": 1.8979206049149338,
      "grad_norm": 0.3690364646660707,
      "learning_rate": 3.263000941650917e-05,
      "loss": 0.5176860094070435,
      "step": 2008
    },
    {
      "epoch": 1.8988657844990549,
      "grad_norm": 0.36043568959334377,
      "learning_rate": 3.258116437452714e-05,
      "loss": 0.5219160318374634,
      "step": 2009
    },
    {
      "epoch": 1.899810964083176,
      "grad_norm": 0.4218468422740774,
      "learning_rate": 3.2532338244838415e-05,
      "loss": 0.542986273765564,
      "step": 2010
    },
    {
      "epoch": 1.9007561436672968,
      "grad_norm": 0.3974647298039463,
      "learning_rate": 3.248353108045533e-05,
      "loss": 0.545242428779602,
      "step": 2011
    },
    {
      "epoch": 1.9017013232514177,
      "grad_norm": 0.37575784590057804,
      "learning_rate": 3.2434742934369747e-05,
      "loss": 0.4931345582008362,
      "step": 2012
    },
    {
      "epoch": 1.9026465028355388,
      "grad_norm": 0.36723974055104314,
      "learning_rate": 3.238597385955276e-05,
      "loss": 0.496492862701416,
      "step": 2013
    },
    {
      "epoch": 1.9035916824196597,
      "grad_norm": 0.3726693140794516,
      "learning_rate": 3.233722390895483e-05,
      "loss": 0.4901629090309143,
      "step": 2014
    },
    {
      "epoch": 1.9045368620037806,
      "grad_norm": 0.35663938426322067,
      "learning_rate": 3.228849313550566e-05,
      "loss": 0.4510194659233093,
      "step": 2015
    },
    {
      "epoch": 1.9054820415879017,
      "grad_norm": 0.3389197231257798,
      "learning_rate": 3.2239781592114056e-05,
      "loss": 0.5453932881355286,
      "step": 2016
    },
    {
      "epoch": 1.9064272211720228,
      "grad_norm": 0.36967623053280124,
      "learning_rate": 3.2191089331668037e-05,
      "loss": 0.5153694152832031,
      "step": 2017
    },
    {
      "epoch": 1.9073724007561437,
      "grad_norm": 0.3869311936232863,
      "learning_rate": 3.214241640703464e-05,
      "loss": 0.5199387073516846,
      "step": 2018
    },
    {
      "epoch": 1.9083175803402646,
      "grad_norm": 0.3814486355011513,
      "learning_rate": 3.209376287105986e-05,
      "loss": 0.5273022651672363,
      "step": 2019
    },
    {
      "epoch": 1.9092627599243857,
      "grad_norm": 0.39335689345203806,
      "learning_rate": 3.2045128776568784e-05,
      "loss": 0.5587946176528931,
      "step": 2020
    },
    {
      "epoch": 1.9102079395085068,
      "grad_norm": 0.3302910876007178,
      "learning_rate": 3.199651417636523e-05,
      "loss": 0.46270471811294556,
      "step": 2021
    },
    {
      "epoch": 1.9111531190926276,
      "grad_norm": 0.3890635479204398,
      "learning_rate": 3.1947919123231955e-05,
      "loss": 0.5729001760482788,
      "step": 2022
    },
    {
      "epoch": 1.9120982986767485,
      "grad_norm": 0.3480782916562638,
      "learning_rate": 3.1899343669930446e-05,
      "loss": 0.5100210905075073,
      "step": 2023
    },
    {
      "epoch": 1.9130434782608696,
      "grad_norm": 0.3647762922200471,
      "learning_rate": 3.185078786920094e-05,
      "loss": 0.535561203956604,
      "step": 2024
    },
    {
      "epoch": 1.9139886578449905,
      "grad_norm": 0.35689598360394104,
      "learning_rate": 3.180225177376229e-05,
      "loss": 0.47208288311958313,
      "step": 2025
    },
    {
      "epoch": 1.9149338374291114,
      "grad_norm": 0.41359516930040874,
      "learning_rate": 3.175373543631201e-05,
      "loss": 0.5496478080749512,
      "step": 2026
    },
    {
      "epoch": 1.9158790170132325,
      "grad_norm": 0.37659778244479836,
      "learning_rate": 3.1705238909526116e-05,
      "loss": 0.5342281460762024,
      "step": 2027
    },
    {
      "epoch": 1.9168241965973536,
      "grad_norm": 0.37618168005774133,
      "learning_rate": 3.1656762246059135e-05,
      "loss": 0.46914318203926086,
      "step": 2028
    },
    {
      "epoch": 1.9177693761814745,
      "grad_norm": 0.3810061308265302,
      "learning_rate": 3.160830549854406e-05,
      "loss": 0.4523455798625946,
      "step": 2029
    },
    {
      "epoch": 1.9187145557655954,
      "grad_norm": 0.3669052611504174,
      "learning_rate": 3.155986871959219e-05,
      "loss": 0.473121702671051,
      "step": 2030
    },
    {
      "epoch": 1.9196597353497165,
      "grad_norm": 0.44569186391189985,
      "learning_rate": 3.151145196179318e-05,
      "loss": 0.4946131408214569,
      "step": 2031
    },
    {
      "epoch": 1.9206049149338376,
      "grad_norm": 2.930328158428832,
      "learning_rate": 3.146305527771499e-05,
      "loss": 0.4136951267719269,
      "step": 2032
    },
    {
      "epoch": 1.9215500945179584,
      "grad_norm": 0.37018966424693156,
      "learning_rate": 3.141467871990368e-05,
      "loss": 0.4831899404525757,
      "step": 2033
    },
    {
      "epoch": 1.9224952741020793,
      "grad_norm": 0.4239390694834955,
      "learning_rate": 3.136632234088358e-05,
      "loss": 0.5125586986541748,
      "step": 2034
    },
    {
      "epoch": 1.9234404536862004,
      "grad_norm": 0.36997284279250403,
      "learning_rate": 3.131798619315702e-05,
      "loss": 0.5240697264671326,
      "step": 2035
    },
    {
      "epoch": 1.9243856332703215,
      "grad_norm": 0.36923038474024705,
      "learning_rate": 3.12696703292044e-05,
      "loss": 0.4923272132873535,
      "step": 2036
    },
    {
      "epoch": 1.9253308128544422,
      "grad_norm": 0.3793507299271429,
      "learning_rate": 3.122137480148412e-05,
      "loss": 0.5705890655517578,
      "step": 2037
    },
    {
      "epoch": 1.9262759924385633,
      "grad_norm": 0.37998074194232817,
      "learning_rate": 3.117309966243243e-05,
      "loss": 0.4913751780986786,
      "step": 2038
    },
    {
      "epoch": 1.9272211720226844,
      "grad_norm": 0.4078674190928192,
      "learning_rate": 3.1124844964463506e-05,
      "loss": 0.4564023017883301,
      "step": 2039
    },
    {
      "epoch": 1.9281663516068053,
      "grad_norm": 0.3647795848791362,
      "learning_rate": 3.107661075996932e-05,
      "loss": 0.4634438455104828,
      "step": 2040
    },
    {
      "epoch": 1.9291115311909262,
      "grad_norm": 0.3742554837792332,
      "learning_rate": 3.102839710131958e-05,
      "loss": 0.5122406482696533,
      "step": 2041
    },
    {
      "epoch": 1.9300567107750473,
      "grad_norm": 0.3762762567637356,
      "learning_rate": 3.098020404086171e-05,
      "loss": 0.5010645985603333,
      "step": 2042
    },
    {
      "epoch": 1.9310018903591684,
      "grad_norm": 0.41014388974419697,
      "learning_rate": 3.0932031630920717e-05,
      "loss": 0.5087373852729797,
      "step": 2043
    },
    {
      "epoch": 1.9319470699432892,
      "grad_norm": 0.3214608063677679,
      "learning_rate": 3.088387992379924e-05,
      "loss": 0.4572980999946594,
      "step": 2044
    },
    {
      "epoch": 1.9328922495274101,
      "grad_norm": 0.37244078883303927,
      "learning_rate": 3.083574897177741e-05,
      "loss": 0.5427142381668091,
      "step": 2045
    },
    {
      "epoch": 1.9338374291115312,
      "grad_norm": 0.3986182253087161,
      "learning_rate": 3.078763882711286e-05,
      "loss": 0.5119937062263489,
      "step": 2046
    },
    {
      "epoch": 1.9347826086956523,
      "grad_norm": 0.3864410430057464,
      "learning_rate": 3.073954954204058e-05,
      "loss": 0.567863941192627,
      "step": 2047
    },
    {
      "epoch": 1.9357277882797732,
      "grad_norm": 0.36567726863264816,
      "learning_rate": 3.069148116877294e-05,
      "loss": 0.4985615611076355,
      "step": 2048
    },
    {
      "epoch": 1.936672967863894,
      "grad_norm": 0.3719777220180968,
      "learning_rate": 3.064343375949962e-05,
      "loss": 0.5293709635734558,
      "step": 2049
    },
    {
      "epoch": 1.9376181474480152,
      "grad_norm": 0.334498101945676,
      "learning_rate": 3.0595407366387504e-05,
      "loss": 0.42135605216026306,
      "step": 2050
    },
    {
      "epoch": 1.938563327032136,
      "grad_norm": 0.3577050432292882,
      "learning_rate": 3.054740204158072e-05,
      "loss": 0.49873214960098267,
      "step": 2051
    },
    {
      "epoch": 1.939508506616257,
      "grad_norm": 0.3831729718697075,
      "learning_rate": 3.0499417837200407e-05,
      "loss": 0.5505514144897461,
      "step": 2052
    },
    {
      "epoch": 1.940453686200378,
      "grad_norm": 0.3505264193033531,
      "learning_rate": 3.045145480534489e-05,
      "loss": 0.4453883767127991,
      "step": 2053
    },
    {
      "epoch": 1.9413988657844992,
      "grad_norm": 0.3504282337040348,
      "learning_rate": 3.0403512998089466e-05,
      "loss": 0.5028552412986755,
      "step": 2054
    },
    {
      "epoch": 1.94234404536862,
      "grad_norm": 0.3709125856909244,
      "learning_rate": 3.035559246748635e-05,
      "loss": 0.47873741388320923,
      "step": 2055
    },
    {
      "epoch": 1.943289224952741,
      "grad_norm": 0.40647965877725567,
      "learning_rate": 3.0307693265564706e-05,
      "loss": 0.6110566854476929,
      "step": 2056
    },
    {
      "epoch": 1.944234404536862,
      "grad_norm": 0.3611638348582463,
      "learning_rate": 3.0259815444330508e-05,
      "loss": 0.49201807379722595,
      "step": 2057
    },
    {
      "epoch": 1.9451795841209831,
      "grad_norm": 0.3948152117251249,
      "learning_rate": 3.0211959055766548e-05,
      "loss": 0.5598848462104797,
      "step": 2058
    },
    {
      "epoch": 1.946124763705104,
      "grad_norm": 0.4071459598430652,
      "learning_rate": 3.016412415183233e-05,
      "loss": 0.597074031829834,
      "step": 2059
    },
    {
      "epoch": 1.947069943289225,
      "grad_norm": 0.3732035019835801,
      "learning_rate": 3.0116310784464008e-05,
      "loss": 0.5854983925819397,
      "step": 2060
    },
    {
      "epoch": 1.948015122873346,
      "grad_norm": 0.34170426071039905,
      "learning_rate": 3.0068519005574402e-05,
      "loss": 0.41852426528930664,
      "step": 2061
    },
    {
      "epoch": 1.9489603024574669,
      "grad_norm": 0.36678483562539504,
      "learning_rate": 3.002074886705284e-05,
      "loss": 0.5605223774909973,
      "step": 2062
    },
    {
      "epoch": 1.9499054820415878,
      "grad_norm": 0.3743752385032602,
      "learning_rate": 2.9973000420765218e-05,
      "loss": 0.4903303384780884,
      "step": 2063
    },
    {
      "epoch": 1.9508506616257089,
      "grad_norm": 0.3763818201701994,
      "learning_rate": 2.9925273718553813e-05,
      "loss": 0.4800161123275757,
      "step": 2064
    },
    {
      "epoch": 1.95179584120983,
      "grad_norm": 0.3587904112687578,
      "learning_rate": 2.987756881223732e-05,
      "loss": 0.46676957607269287,
      "step": 2065
    },
    {
      "epoch": 1.9527410207939508,
      "grad_norm": 0.38250479074612154,
      "learning_rate": 2.9829885753610796e-05,
      "loss": 0.5325554609298706,
      "step": 2066
    },
    {
      "epoch": 1.9536862003780717,
      "grad_norm": 0.3800795783133072,
      "learning_rate": 2.9782224594445534e-05,
      "loss": 0.5200610160827637,
      "step": 2067
    },
    {
      "epoch": 1.9546313799621928,
      "grad_norm": 0.3975645977806566,
      "learning_rate": 2.9734585386489093e-05,
      "loss": 0.5516006946563721,
      "step": 2068
    },
    {
      "epoch": 1.955576559546314,
      "grad_norm": 0.3347580669761802,
      "learning_rate": 2.968696818146515e-05,
      "loss": 0.4493429660797119,
      "step": 2069
    },
    {
      "epoch": 1.9565217391304348,
      "grad_norm": 0.3511636142480725,
      "learning_rate": 2.9639373031073525e-05,
      "loss": 0.4796178936958313,
      "step": 2070
    },
    {
      "epoch": 1.9574669187145557,
      "grad_norm": 0.42259050583929203,
      "learning_rate": 2.95917999869901e-05,
      "loss": 0.5211926102638245,
      "step": 2071
    },
    {
      "epoch": 1.9584120982986768,
      "grad_norm": 0.45058450514311554,
      "learning_rate": 2.9544249100866718e-05,
      "loss": 0.5943093299865723,
      "step": 2072
    },
    {
      "epoch": 1.9593572778827977,
      "grad_norm": 0.3792695568901925,
      "learning_rate": 2.9496720424331214e-05,
      "loss": 0.5014647841453552,
      "step": 2073
    },
    {
      "epoch": 1.9603024574669186,
      "grad_norm": 0.3815208689320114,
      "learning_rate": 2.9449214008987254e-05,
      "loss": 0.5703405141830444,
      "step": 2074
    },
    {
      "epoch": 1.9612476370510397,
      "grad_norm": 0.3890115761388147,
      "learning_rate": 2.940172990641438e-05,
      "loss": 0.4737856388092041,
      "step": 2075
    },
    {
      "epoch": 1.9621928166351608,
      "grad_norm": 0.42774353825835176,
      "learning_rate": 2.9354268168167902e-05,
      "loss": 0.543567419052124,
      "step": 2076
    },
    {
      "epoch": 1.9631379962192816,
      "grad_norm": 0.3764831657502124,
      "learning_rate": 2.93068288457788e-05,
      "loss": 0.4704555869102478,
      "step": 2077
    },
    {
      "epoch": 1.9640831758034025,
      "grad_norm": 0.39538960115986355,
      "learning_rate": 2.925941199075378e-05,
      "loss": 0.5060553550720215,
      "step": 2078
    },
    {
      "epoch": 1.9650283553875236,
      "grad_norm": 0.3412455027328821,
      "learning_rate": 2.9212017654575126e-05,
      "loss": 0.4439593553543091,
      "step": 2079
    },
    {
      "epoch": 1.9659735349716447,
      "grad_norm": 0.36321665393696256,
      "learning_rate": 2.916464588870067e-05,
      "loss": 0.47483986616134644,
      "step": 2080
    },
    {
      "epoch": 1.9669187145557656,
      "grad_norm": 0.3840320089073532,
      "learning_rate": 2.911729674456374e-05,
      "loss": 0.4818076491355896,
      "step": 2081
    },
    {
      "epoch": 1.9678638941398865,
      "grad_norm": 0.36610660853601956,
      "learning_rate": 2.906997027357312e-05,
      "loss": 0.47046786546707153,
      "step": 2082
    },
    {
      "epoch": 1.9688090737240076,
      "grad_norm": 0.3892554070411309,
      "learning_rate": 2.9022666527112953e-05,
      "loss": 0.5826300382614136,
      "step": 2083
    },
    {
      "epoch": 1.9697542533081287,
      "grad_norm": 0.3598363003578545,
      "learning_rate": 2.89753855565427e-05,
      "loss": 0.5027961134910583,
      "step": 2084
    },
    {
      "epoch": 1.9706994328922496,
      "grad_norm": 0.38624901379362414,
      "learning_rate": 2.8928127413197124e-05,
      "loss": 0.524280846118927,
      "step": 2085
    },
    {
      "epoch": 1.9716446124763705,
      "grad_norm": 0.3770895573971986,
      "learning_rate": 2.8880892148386198e-05,
      "loss": 0.5457723736763,
      "step": 2086
    },
    {
      "epoch": 1.9725897920604916,
      "grad_norm": 0.370480346478408,
      "learning_rate": 2.8833679813395054e-05,
      "loss": 0.40755945444107056,
      "step": 2087
    },
    {
      "epoch": 1.9735349716446124,
      "grad_norm": 0.32705101908153633,
      "learning_rate": 2.8786490459483917e-05,
      "loss": 0.4855905771255493,
      "step": 2088
    },
    {
      "epoch": 1.9744801512287333,
      "grad_norm": 0.3875267082034954,
      "learning_rate": 2.8739324137888046e-05,
      "loss": 0.5474836826324463,
      "step": 2089
    },
    {
      "epoch": 1.9754253308128544,
      "grad_norm": 0.4008619821748742,
      "learning_rate": 2.869218089981772e-05,
      "loss": 0.5018196702003479,
      "step": 2090
    },
    {
      "epoch": 1.9763705103969755,
      "grad_norm": 0.3951717538620143,
      "learning_rate": 2.8645060796458167e-05,
      "loss": 0.5765536427497864,
      "step": 2091
    },
    {
      "epoch": 1.9773156899810964,
      "grad_norm": 0.3725951891468992,
      "learning_rate": 2.8597963878969486e-05,
      "loss": 0.5289465188980103,
      "step": 2092
    },
    {
      "epoch": 1.9782608695652173,
      "grad_norm": 0.36714123157607637,
      "learning_rate": 2.855089019848657e-05,
      "loss": 0.4880030155181885,
      "step": 2093
    },
    {
      "epoch": 1.9792060491493384,
      "grad_norm": 0.3842658590408605,
      "learning_rate": 2.8503839806119094e-05,
      "loss": 0.5386883616447449,
      "step": 2094
    },
    {
      "epoch": 1.9801512287334595,
      "grad_norm": 0.3470147977950996,
      "learning_rate": 2.8456812752951485e-05,
      "loss": 0.5157870650291443,
      "step": 2095
    },
    {
      "epoch": 1.9810964083175804,
      "grad_norm": 0.35317846383250023,
      "learning_rate": 2.8409809090042792e-05,
      "loss": 0.515323281288147,
      "step": 2096
    },
    {
      "epoch": 1.9820415879017013,
      "grad_norm": 0.37289293519139693,
      "learning_rate": 2.8362828868426715e-05,
      "loss": 0.5057772397994995,
      "step": 2097
    },
    {
      "epoch": 1.9829867674858224,
      "grad_norm": 0.36343315469301063,
      "learning_rate": 2.831587213911142e-05,
      "loss": 0.5014412999153137,
      "step": 2098
    },
    {
      "epoch": 1.9839319470699432,
      "grad_norm": 0.39536918152552636,
      "learning_rate": 2.826893895307967e-05,
      "loss": 0.5856465101242065,
      "step": 2099
    },
    {
      "epoch": 1.9848771266540641,
      "grad_norm": 0.38895307858994416,
      "learning_rate": 2.8222029361288583e-05,
      "loss": 0.5359112024307251,
      "step": 2100
    },
    {
      "epoch": 1.9858223062381852,
      "grad_norm": 0.3749794063038171,
      "learning_rate": 2.8175143414669646e-05,
      "loss": 0.45887595415115356,
      "step": 2101
    },
    {
      "epoch": 1.9867674858223063,
      "grad_norm": 0.36081292704445234,
      "learning_rate": 2.81282811641288e-05,
      "loss": 0.5393571853637695,
      "step": 2102
    },
    {
      "epoch": 1.9877126654064272,
      "grad_norm": 0.36840620763523085,
      "learning_rate": 2.8081442660546125e-05,
      "loss": 0.5307824015617371,
      "step": 2103
    },
    {
      "epoch": 1.988657844990548,
      "grad_norm": 0.35012263900076734,
      "learning_rate": 2.8034627954775994e-05,
      "loss": 0.48886361718177795,
      "step": 2104
    },
    {
      "epoch": 1.9896030245746692,
      "grad_norm": 0.3517029668326387,
      "learning_rate": 2.7987837097646908e-05,
      "loss": 0.45998233556747437,
      "step": 2105
    },
    {
      "epoch": 1.9905482041587903,
      "grad_norm": 0.37978468219394784,
      "learning_rate": 2.7941070139961455e-05,
      "loss": 0.4645790457725525,
      "step": 2106
    },
    {
      "epoch": 1.9914933837429112,
      "grad_norm": 0.3741750633956033,
      "learning_rate": 2.7894327132496325e-05,
      "loss": 0.3941006660461426,
      "step": 2107
    },
    {
      "epoch": 1.992438563327032,
      "grad_norm": 0.409154799381066,
      "learning_rate": 2.7847608126002178e-05,
      "loss": 0.5339951515197754,
      "step": 2108
    },
    {
      "epoch": 1.9933837429111532,
      "grad_norm": 0.3305232249121578,
      "learning_rate": 2.7800913171203647e-05,
      "loss": 0.4103067219257355,
      "step": 2109
    },
    {
      "epoch": 1.994328922495274,
      "grad_norm": 0.34014640636635146,
      "learning_rate": 2.7754242318799174e-05,
      "loss": 0.5135060548782349,
      "step": 2110
    },
    {
      "epoch": 1.995274102079395,
      "grad_norm": 0.3983048439388205,
      "learning_rate": 2.7707595619461126e-05,
      "loss": 0.44879671931266785,
      "step": 2111
    },
    {
      "epoch": 1.996219281663516,
      "grad_norm": 0.38820430707661074,
      "learning_rate": 2.766097312383556e-05,
      "loss": 0.5163404941558838,
      "step": 2112
    },
    {
      "epoch": 1.9971644612476371,
      "grad_norm": 0.37570311660121003,
      "learning_rate": 2.7614374882542317e-05,
      "loss": 0.5073859691619873,
      "step": 2113
    },
    {
      "epoch": 1.998109640831758,
      "grad_norm": 0.4097918914227589,
      "learning_rate": 2.756780094617489e-05,
      "loss": 0.49228134751319885,
      "step": 2114
    },
    {
      "epoch": 1.999054820415879,
      "grad_norm": 0.37769864843267936,
      "learning_rate": 2.752125136530036e-05,
      "loss": 0.5284162163734436,
      "step": 2115
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3988449550560906,
      "learning_rate": 2.7474726190459398e-05,
      "loss": 0.5173879861831665,
      "step": 2116
    },
    {
      "epoch": 2.000945179584121,
      "grad_norm": 0.35180753101509926,
      "learning_rate": 2.7428225472166137e-05,
      "loss": 0.46776509284973145,
      "step": 2117
    },
    {
      "epoch": 2.0018903591682418,
      "grad_norm": 0.3575426727386266,
      "learning_rate": 2.7381749260908174e-05,
      "loss": 0.46384137868881226,
      "step": 2118
    },
    {
      "epoch": 2.002835538752363,
      "grad_norm": 0.3503814984814308,
      "learning_rate": 2.733529760714655e-05,
      "loss": 0.45158058404922485,
      "step": 2119
    },
    {
      "epoch": 2.003780718336484,
      "grad_norm": 0.34216067420876345,
      "learning_rate": 2.7288870561315527e-05,
      "loss": 0.43013182282447815,
      "step": 2120
    },
    {
      "epoch": 2.004725897920605,
      "grad_norm": 0.3548834257073154,
      "learning_rate": 2.7242468173822773e-05,
      "loss": 0.42817363142967224,
      "step": 2121
    },
    {
      "epoch": 2.0056710775047257,
      "grad_norm": 0.37609409384282066,
      "learning_rate": 2.719609049504911e-05,
      "loss": 0.4396292567253113,
      "step": 2122
    },
    {
      "epoch": 2.006616257088847,
      "grad_norm": 0.3997886322350689,
      "learning_rate": 2.7149737575348498e-05,
      "loss": 0.46783822774887085,
      "step": 2123
    },
    {
      "epoch": 2.007561436672968,
      "grad_norm": 0.3398151713315228,
      "learning_rate": 2.7103409465048152e-05,
      "loss": 0.4466320872306824,
      "step": 2124
    },
    {
      "epoch": 2.008506616257089,
      "grad_norm": 0.3169950484453944,
      "learning_rate": 2.7057106214448212e-05,
      "loss": 0.3584567606449127,
      "step": 2125
    },
    {
      "epoch": 2.0094517958412097,
      "grad_norm": 0.37380800926232643,
      "learning_rate": 2.701082787382192e-05,
      "loss": 0.49402543902397156,
      "step": 2126
    },
    {
      "epoch": 2.010396975425331,
      "grad_norm": 0.360521089022573,
      "learning_rate": 2.6964574493415386e-05,
      "loss": 0.37953054904937744,
      "step": 2127
    },
    {
      "epoch": 2.011342155009452,
      "grad_norm": 0.3732696612918124,
      "learning_rate": 2.691834612344771e-05,
      "loss": 0.48442554473876953,
      "step": 2128
    },
    {
      "epoch": 2.0122873345935726,
      "grad_norm": 0.3873800719976953,
      "learning_rate": 2.6872142814110757e-05,
      "loss": 0.4693063199520111,
      "step": 2129
    },
    {
      "epoch": 2.0132325141776937,
      "grad_norm": 0.4182103425602716,
      "learning_rate": 2.6825964615569233e-05,
      "loss": 0.44049128890037537,
      "step": 2130
    },
    {
      "epoch": 2.0141776937618148,
      "grad_norm": 0.3739109506748866,
      "learning_rate": 2.677981157796059e-05,
      "loss": 0.4427742660045624,
      "step": 2131
    },
    {
      "epoch": 2.015122873345936,
      "grad_norm": 0.41923304684445456,
      "learning_rate": 2.6733683751394906e-05,
      "loss": 0.464089035987854,
      "step": 2132
    },
    {
      "epoch": 2.0160680529300565,
      "grad_norm": 0.37232413467035835,
      "learning_rate": 2.6687581185954947e-05,
      "loss": 0.4314904510974884,
      "step": 2133
    },
    {
      "epoch": 2.0170132325141776,
      "grad_norm": 0.43809525607181465,
      "learning_rate": 2.6641503931695998e-05,
      "loss": 0.4239979088306427,
      "step": 2134
    },
    {
      "epoch": 2.0179584120982987,
      "grad_norm": 0.3761106131606495,
      "learning_rate": 2.6595452038645897e-05,
      "loss": 0.3814314305782318,
      "step": 2135
    },
    {
      "epoch": 2.01890359168242,
      "grad_norm": 0.3713377413074518,
      "learning_rate": 2.6549425556804964e-05,
      "loss": 0.4076201319694519,
      "step": 2136
    },
    {
      "epoch": 2.0198487712665405,
      "grad_norm": 0.41108696178322546,
      "learning_rate": 2.650342453614586e-05,
      "loss": 0.4476206302642822,
      "step": 2137
    },
    {
      "epoch": 2.0207939508506616,
      "grad_norm": 0.4051670426007556,
      "learning_rate": 2.645744902661369e-05,
      "loss": 0.44159430265426636,
      "step": 2138
    },
    {
      "epoch": 2.0217391304347827,
      "grad_norm": 0.4511690219067836,
      "learning_rate": 2.641149907812577e-05,
      "loss": 0.406975120306015,
      "step": 2139
    },
    {
      "epoch": 2.022684310018904,
      "grad_norm": 0.45035201720269885,
      "learning_rate": 2.636557474057173e-05,
      "loss": 0.4686965346336365,
      "step": 2140
    },
    {
      "epoch": 2.0236294896030245,
      "grad_norm": 0.4382850212700783,
      "learning_rate": 2.6319676063813405e-05,
      "loss": 0.4847043752670288,
      "step": 2141
    },
    {
      "epoch": 2.0245746691871456,
      "grad_norm": 0.3632940539021603,
      "learning_rate": 2.627380309768468e-05,
      "loss": 0.4052249491214752,
      "step": 2142
    },
    {
      "epoch": 2.0255198487712667,
      "grad_norm": 0.4020384113270473,
      "learning_rate": 2.6227955891991617e-05,
      "loss": 0.4137381911277771,
      "step": 2143
    },
    {
      "epoch": 2.0264650283553873,
      "grad_norm": 0.3992923293302969,
      "learning_rate": 2.6182134496512256e-05,
      "loss": 0.3994314968585968,
      "step": 2144
    },
    {
      "epoch": 2.0274102079395084,
      "grad_norm": 0.4437367907440134,
      "learning_rate": 2.6136338960996666e-05,
      "loss": 0.5059645175933838,
      "step": 2145
    },
    {
      "epoch": 2.0283553875236295,
      "grad_norm": 0.38870063938573585,
      "learning_rate": 2.609056933516675e-05,
      "loss": 0.4060082733631134,
      "step": 2146
    },
    {
      "epoch": 2.0293005671077506,
      "grad_norm": 0.4241646968443503,
      "learning_rate": 2.6044825668716354e-05,
      "loss": 0.47031843662261963,
      "step": 2147
    },
    {
      "epoch": 2.0302457466918713,
      "grad_norm": 0.3828025393115284,
      "learning_rate": 2.5999108011311158e-05,
      "loss": 0.43780744075775146,
      "step": 2148
    },
    {
      "epoch": 2.0311909262759924,
      "grad_norm": 0.3805143112958797,
      "learning_rate": 2.5953416412588503e-05,
      "loss": 0.42485904693603516,
      "step": 2149
    },
    {
      "epoch": 2.0321361058601135,
      "grad_norm": 0.3993772276248569,
      "learning_rate": 2.5907750922157552e-05,
      "loss": 0.4616858959197998,
      "step": 2150
    },
    {
      "epoch": 2.0330812854442346,
      "grad_norm": 0.4024038660409161,
      "learning_rate": 2.5862111589599026e-05,
      "loss": 0.4200604557991028,
      "step": 2151
    },
    {
      "epoch": 2.0340264650283553,
      "grad_norm": 0.40825218595271306,
      "learning_rate": 2.58164984644653e-05,
      "loss": 0.41966673731803894,
      "step": 2152
    },
    {
      "epoch": 2.0349716446124764,
      "grad_norm": 0.3855680766296279,
      "learning_rate": 2.57709115962803e-05,
      "loss": 0.4098477363586426,
      "step": 2153
    },
    {
      "epoch": 2.0359168241965975,
      "grad_norm": 0.3954474645559256,
      "learning_rate": 2.5725351034539404e-05,
      "loss": 0.43961507081985474,
      "step": 2154
    },
    {
      "epoch": 2.036862003780718,
      "grad_norm": 0.3572736435502417,
      "learning_rate": 2.5679816828709458e-05,
      "loss": 0.3825836479663849,
      "step": 2155
    },
    {
      "epoch": 2.0378071833648392,
      "grad_norm": 0.3948119654113459,
      "learning_rate": 2.5634309028228675e-05,
      "loss": 0.4019758999347687,
      "step": 2156
    },
    {
      "epoch": 2.0387523629489603,
      "grad_norm": 0.3849018380479272,
      "learning_rate": 2.55888276825066e-05,
      "loss": 0.4094269573688507,
      "step": 2157
    },
    {
      "epoch": 2.0396975425330814,
      "grad_norm": 0.40274302220427555,
      "learning_rate": 2.554337284092411e-05,
      "loss": 0.39837390184402466,
      "step": 2158
    },
    {
      "epoch": 2.040642722117202,
      "grad_norm": 0.45716094578779587,
      "learning_rate": 2.5497944552833186e-05,
      "loss": 0.4409695863723755,
      "step": 2159
    },
    {
      "epoch": 2.041587901701323,
      "grad_norm": 0.3915246264763467,
      "learning_rate": 2.5452542867557117e-05,
      "loss": 0.42825591564178467,
      "step": 2160
    },
    {
      "epoch": 2.0425330812854443,
      "grad_norm": 0.3832667848575064,
      "learning_rate": 2.5407167834390187e-05,
      "loss": 0.3718026280403137,
      "step": 2161
    },
    {
      "epoch": 2.0434782608695654,
      "grad_norm": 0.4066866100055429,
      "learning_rate": 2.5361819502597835e-05,
      "loss": 0.40043410658836365,
      "step": 2162
    },
    {
      "epoch": 2.044423440453686,
      "grad_norm": 0.4449421557596981,
      "learning_rate": 2.5316497921416473e-05,
      "loss": 0.46111804246902466,
      "step": 2163
    },
    {
      "epoch": 2.045368620037807,
      "grad_norm": 0.42191052493725656,
      "learning_rate": 2.5271203140053433e-05,
      "loss": 0.4474084973335266,
      "step": 2164
    },
    {
      "epoch": 2.0463137996219283,
      "grad_norm": 0.40650091758357476,
      "learning_rate": 2.5225935207687025e-05,
      "loss": 0.4404282569885254,
      "step": 2165
    },
    {
      "epoch": 2.047258979206049,
      "grad_norm": 0.39828942561648484,
      "learning_rate": 2.5180694173466335e-05,
      "loss": 0.4788551926612854,
      "step": 2166
    },
    {
      "epoch": 2.04820415879017,
      "grad_norm": 0.4240751989660345,
      "learning_rate": 2.5135480086511302e-05,
      "loss": 0.4734519124031067,
      "step": 2167
    },
    {
      "epoch": 2.049149338374291,
      "grad_norm": 0.3833394793939934,
      "learning_rate": 2.509029299591254e-05,
      "loss": 0.4561379551887512,
      "step": 2168
    },
    {
      "epoch": 2.0500945179584122,
      "grad_norm": 0.4249863526984204,
      "learning_rate": 2.5045132950731463e-05,
      "loss": 0.44410550594329834,
      "step": 2169
    },
    {
      "epoch": 2.051039697542533,
      "grad_norm": 0.3934021794262037,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.4470238983631134,
      "step": 2170
    },
    {
      "epoch": 2.051984877126654,
      "grad_norm": 0.45001804346553564,
      "learning_rate": 2.4954894192720735e-05,
      "loss": 0.4550589919090271,
      "step": 2171
    },
    {
      "epoch": 2.052930056710775,
      "grad_norm": 0.43124331040716146,
      "learning_rate": 2.4909815577866762e-05,
      "loss": 0.44697898626327515,
      "step": 2172
    },
    {
      "epoch": 2.053875236294896,
      "grad_norm": 0.41433529426705046,
      "learning_rate": 2.4864764204381624e-05,
      "loss": 0.4327826499938965,
      "step": 2173
    },
    {
      "epoch": 2.054820415879017,
      "grad_norm": 0.39817449605783733,
      "learning_rate": 2.481974012117933e-05,
      "loss": 0.41948264837265015,
      "step": 2174
    },
    {
      "epoch": 2.055765595463138,
      "grad_norm": 0.43154995395419143,
      "learning_rate": 2.4774743377144265e-05,
      "loss": 0.43815481662750244,
      "step": 2175
    },
    {
      "epoch": 2.056710775047259,
      "grad_norm": 0.4321809814570712,
      "learning_rate": 2.472977402113107e-05,
      "loss": 0.4852944612503052,
      "step": 2176
    },
    {
      "epoch": 2.05765595463138,
      "grad_norm": 0.4609701475223974,
      "learning_rate": 2.4684832101964724e-05,
      "loss": 0.5413817763328552,
      "step": 2177
    },
    {
      "epoch": 2.058601134215501,
      "grad_norm": 0.4250208209970256,
      "learning_rate": 2.4639917668440337e-05,
      "loss": 0.44342148303985596,
      "step": 2178
    },
    {
      "epoch": 2.059546313799622,
      "grad_norm": 0.43989029720848033,
      "learning_rate": 2.4595030769323247e-05,
      "loss": 0.4579635262489319,
      "step": 2179
    },
    {
      "epoch": 2.060491493383743,
      "grad_norm": 0.4295375580086489,
      "learning_rate": 2.4550171453348887e-05,
      "loss": 0.43907052278518677,
      "step": 2180
    },
    {
      "epoch": 2.0614366729678637,
      "grad_norm": 0.43309939735393055,
      "learning_rate": 2.4505339769222685e-05,
      "loss": 0.4476909339427948,
      "step": 2181
    },
    {
      "epoch": 2.062381852551985,
      "grad_norm": 0.430173638053452,
      "learning_rate": 2.4460535765620145e-05,
      "loss": 0.4591204822063446,
      "step": 2182
    },
    {
      "epoch": 2.063327032136106,
      "grad_norm": 0.41243723398278004,
      "learning_rate": 2.4415759491186646e-05,
      "loss": 0.48443731665611267,
      "step": 2183
    },
    {
      "epoch": 2.064272211720227,
      "grad_norm": 0.4001493881632056,
      "learning_rate": 2.4371010994537513e-05,
      "loss": 0.4340672194957733,
      "step": 2184
    },
    {
      "epoch": 2.0652173913043477,
      "grad_norm": 0.45308999759635094,
      "learning_rate": 2.4326290324257894e-05,
      "loss": 0.4881603717803955,
      "step": 2185
    },
    {
      "epoch": 2.0661625708884688,
      "grad_norm": 0.4164621181854967,
      "learning_rate": 2.4281597528902737e-05,
      "loss": 0.4622117280960083,
      "step": 2186
    },
    {
      "epoch": 2.06710775047259,
      "grad_norm": 0.4065112726189447,
      "learning_rate": 2.4236932656996697e-05,
      "loss": 0.43889665603637695,
      "step": 2187
    },
    {
      "epoch": 2.068052930056711,
      "grad_norm": 0.40412710137069346,
      "learning_rate": 2.4192295757034113e-05,
      "loss": 0.44030264019966125,
      "step": 2188
    },
    {
      "epoch": 2.0689981096408316,
      "grad_norm": 0.4305271417150185,
      "learning_rate": 2.4147686877478992e-05,
      "loss": 0.39753153920173645,
      "step": 2189
    },
    {
      "epoch": 2.0699432892249527,
      "grad_norm": 0.448537125273243,
      "learning_rate": 2.410310606676485e-05,
      "loss": 0.4880251884460449,
      "step": 2190
    },
    {
      "epoch": 2.070888468809074,
      "grad_norm": 0.49616016942446894,
      "learning_rate": 2.4058553373294845e-05,
      "loss": 0.47509896755218506,
      "step": 2191
    },
    {
      "epoch": 2.0718336483931945,
      "grad_norm": 0.3858429233828664,
      "learning_rate": 2.4014028845441488e-05,
      "loss": 0.4137875437736511,
      "step": 2192
    },
    {
      "epoch": 2.0727788279773156,
      "grad_norm": 0.45621822014663815,
      "learning_rate": 2.396953253154674e-05,
      "loss": 0.4933529794216156,
      "step": 2193
    },
    {
      "epoch": 2.0737240075614367,
      "grad_norm": 0.4153483682321523,
      "learning_rate": 2.3925064479921984e-05,
      "loss": 0.4279263913631439,
      "step": 2194
    },
    {
      "epoch": 2.074669187145558,
      "grad_norm": 0.38211541640392527,
      "learning_rate": 2.3880624738847835e-05,
      "loss": 0.4091939330101013,
      "step": 2195
    },
    {
      "epoch": 2.0756143667296785,
      "grad_norm": 0.4236179293196448,
      "learning_rate": 2.3836213356574233e-05,
      "loss": 0.45528119802474976,
      "step": 2196
    },
    {
      "epoch": 2.0765595463137996,
      "grad_norm": 0.45184544415755973,
      "learning_rate": 2.3791830381320308e-05,
      "loss": 0.471064031124115,
      "step": 2197
    },
    {
      "epoch": 2.0775047258979207,
      "grad_norm": 0.39391225719289563,
      "learning_rate": 2.3747475861274364e-05,
      "loss": 0.37518101930618286,
      "step": 2198
    },
    {
      "epoch": 2.078449905482042,
      "grad_norm": 0.42305447583191963,
      "learning_rate": 2.370314984459378e-05,
      "loss": 0.43955472111701965,
      "step": 2199
    },
    {
      "epoch": 2.0793950850661624,
      "grad_norm": 0.4033674972433396,
      "learning_rate": 2.3658852379404973e-05,
      "loss": 0.41392964124679565,
      "step": 2200
    },
    {
      "epoch": 2.0803402646502835,
      "grad_norm": 0.41704133306356245,
      "learning_rate": 2.3614583513803412e-05,
      "loss": 0.4711492657661438,
      "step": 2201
    },
    {
      "epoch": 2.0812854442344046,
      "grad_norm": 0.4073359503789675,
      "learning_rate": 2.3570343295853492e-05,
      "loss": 0.47906503081321716,
      "step": 2202
    },
    {
      "epoch": 2.0822306238185253,
      "grad_norm": 0.39252263395930664,
      "learning_rate": 2.3526131773588516e-05,
      "loss": 0.4007291793823242,
      "step": 2203
    },
    {
      "epoch": 2.0831758034026464,
      "grad_norm": 0.47624115215981383,
      "learning_rate": 2.34819489950106e-05,
      "loss": 0.47091785073280334,
      "step": 2204
    },
    {
      "epoch": 2.0841209829867675,
      "grad_norm": 0.4324889238008738,
      "learning_rate": 2.3437795008090656e-05,
      "loss": 0.4686664640903473,
      "step": 2205
    },
    {
      "epoch": 2.0850661625708886,
      "grad_norm": 0.42507415921964187,
      "learning_rate": 2.3393669860768364e-05,
      "loss": 0.45162689685821533,
      "step": 2206
    },
    {
      "epoch": 2.0860113421550093,
      "grad_norm": 0.41940068127584973,
      "learning_rate": 2.3349573600952075e-05,
      "loss": 0.46545669436454773,
      "step": 2207
    },
    {
      "epoch": 2.0869565217391304,
      "grad_norm": 0.43079418867640434,
      "learning_rate": 2.33055062765188e-05,
      "loss": 0.4841582775115967,
      "step": 2208
    },
    {
      "epoch": 2.0879017013232515,
      "grad_norm": 0.41498153257437065,
      "learning_rate": 2.3261467935314096e-05,
      "loss": 0.4612477123737335,
      "step": 2209
    },
    {
      "epoch": 2.0888468809073726,
      "grad_norm": 0.39017470752924444,
      "learning_rate": 2.3217458625152038e-05,
      "loss": 0.42734694480895996,
      "step": 2210
    },
    {
      "epoch": 2.0897920604914932,
      "grad_norm": 0.42053794458747656,
      "learning_rate": 2.317347839381526e-05,
      "loss": 0.4588688015937805,
      "step": 2211
    },
    {
      "epoch": 2.0907372400756143,
      "grad_norm": 0.4195925434270758,
      "learning_rate": 2.3129527289054713e-05,
      "loss": 0.44690102338790894,
      "step": 2212
    },
    {
      "epoch": 2.0916824196597354,
      "grad_norm": 0.39142158895191925,
      "learning_rate": 2.308560535858986e-05,
      "loss": 0.4090583920478821,
      "step": 2213
    },
    {
      "epoch": 2.0926275992438566,
      "grad_norm": 0.4254755429892476,
      "learning_rate": 2.304171265010835e-05,
      "loss": 0.43481290340423584,
      "step": 2214
    },
    {
      "epoch": 2.093572778827977,
      "grad_norm": 0.3979577019174357,
      "learning_rate": 2.2997849211266222e-05,
      "loss": 0.42950886487960815,
      "step": 2215
    },
    {
      "epoch": 2.0945179584120983,
      "grad_norm": 0.4157029180976103,
      "learning_rate": 2.295401508968765e-05,
      "loss": 0.4819897413253784,
      "step": 2216
    },
    {
      "epoch": 2.0954631379962194,
      "grad_norm": 0.38690543469860744,
      "learning_rate": 2.2910210332964993e-05,
      "loss": 0.38704901933670044,
      "step": 2217
    },
    {
      "epoch": 2.09640831758034,
      "grad_norm": 0.44231488374017913,
      "learning_rate": 2.286643498865877e-05,
      "loss": 0.44696998596191406,
      "step": 2218
    },
    {
      "epoch": 2.097353497164461,
      "grad_norm": 0.42001691510708566,
      "learning_rate": 2.282268910429753e-05,
      "loss": 0.45035320520401,
      "step": 2219
    },
    {
      "epoch": 2.0982986767485823,
      "grad_norm": 0.40913404533567815,
      "learning_rate": 2.2778972727377868e-05,
      "loss": 0.4146004915237427,
      "step": 2220
    },
    {
      "epoch": 2.0992438563327034,
      "grad_norm": 0.477115333053308,
      "learning_rate": 2.2735285905364306e-05,
      "loss": 0.43703269958496094,
      "step": 2221
    },
    {
      "epoch": 2.100189035916824,
      "grad_norm": 0.3794324810121396,
      "learning_rate": 2.2691628685689266e-05,
      "loss": 0.4069938659667969,
      "step": 2222
    },
    {
      "epoch": 2.101134215500945,
      "grad_norm": 0.4342266774072869,
      "learning_rate": 2.2648001115753076e-05,
      "loss": 0.44930171966552734,
      "step": 2223
    },
    {
      "epoch": 2.1020793950850662,
      "grad_norm": 0.4801431360409946,
      "learning_rate": 2.260440324292385e-05,
      "loss": 0.4542751610279083,
      "step": 2224
    },
    {
      "epoch": 2.1030245746691874,
      "grad_norm": 0.4100736750915237,
      "learning_rate": 2.256083511453747e-05,
      "loss": 0.3621795177459717,
      "step": 2225
    },
    {
      "epoch": 2.103969754253308,
      "grad_norm": 0.4292169195578733,
      "learning_rate": 2.2517296777897506e-05,
      "loss": 0.4276405870914459,
      "step": 2226
    },
    {
      "epoch": 2.104914933837429,
      "grad_norm": 0.43947627143968665,
      "learning_rate": 2.2473788280275164e-05,
      "loss": 0.4271991550922394,
      "step": 2227
    },
    {
      "epoch": 2.10586011342155,
      "grad_norm": 0.3817220549850932,
      "learning_rate": 2.2430309668909293e-05,
      "loss": 0.39904677867889404,
      "step": 2228
    },
    {
      "epoch": 2.106805293005671,
      "grad_norm": 0.4863793659587265,
      "learning_rate": 2.238686099100628e-05,
      "loss": 0.4764706492424011,
      "step": 2229
    },
    {
      "epoch": 2.107750472589792,
      "grad_norm": 0.44646679687611207,
      "learning_rate": 2.234344229374003e-05,
      "loss": 0.4663342237472534,
      "step": 2230
    },
    {
      "epoch": 2.108695652173913,
      "grad_norm": 0.46541202347411026,
      "learning_rate": 2.230005362425184e-05,
      "loss": 0.5142595767974854,
      "step": 2231
    },
    {
      "epoch": 2.109640831758034,
      "grad_norm": 0.4373037972124939,
      "learning_rate": 2.225669502965047e-05,
      "loss": 0.4519689381122589,
      "step": 2232
    },
    {
      "epoch": 2.110586011342155,
      "grad_norm": 0.44006975737616444,
      "learning_rate": 2.221336655701199e-05,
      "loss": 0.4358278512954712,
      "step": 2233
    },
    {
      "epoch": 2.111531190926276,
      "grad_norm": 0.46100841804697285,
      "learning_rate": 2.2170068253379744e-05,
      "loss": 0.4770128130912781,
      "step": 2234
    },
    {
      "epoch": 2.112476370510397,
      "grad_norm": 0.43284489457907904,
      "learning_rate": 2.2126800165764378e-05,
      "loss": 0.4546929597854614,
      "step": 2235
    },
    {
      "epoch": 2.113421550094518,
      "grad_norm": 0.40131263560301644,
      "learning_rate": 2.2083562341143694e-05,
      "loss": 0.3756088614463806,
      "step": 2236
    },
    {
      "epoch": 2.114366729678639,
      "grad_norm": 0.43018517757674946,
      "learning_rate": 2.2040354826462668e-05,
      "loss": 0.4500664472579956,
      "step": 2237
    },
    {
      "epoch": 2.11531190926276,
      "grad_norm": 0.39106664332167773,
      "learning_rate": 2.199717766863332e-05,
      "loss": 0.42121070623397827,
      "step": 2238
    },
    {
      "epoch": 2.116257088846881,
      "grad_norm": 0.43796821324414337,
      "learning_rate": 2.195403091453473e-05,
      "loss": 0.4715374708175659,
      "step": 2239
    },
    {
      "epoch": 2.1172022684310017,
      "grad_norm": 0.4139234316897558,
      "learning_rate": 2.191091461101298e-05,
      "loss": 0.4169420003890991,
      "step": 2240
    },
    {
      "epoch": 2.118147448015123,
      "grad_norm": 0.43278122438302874,
      "learning_rate": 2.186782880488108e-05,
      "loss": 0.49040645360946655,
      "step": 2241
    },
    {
      "epoch": 2.119092627599244,
      "grad_norm": 0.38253209422607987,
      "learning_rate": 2.182477354291896e-05,
      "loss": 0.37375566363334656,
      "step": 2242
    },
    {
      "epoch": 2.120037807183365,
      "grad_norm": 0.4975035635457089,
      "learning_rate": 2.178174887187331e-05,
      "loss": 0.4264434576034546,
      "step": 2243
    },
    {
      "epoch": 2.1209829867674856,
      "grad_norm": 0.43313474566757415,
      "learning_rate": 2.1738754838457698e-05,
      "loss": 0.4468883275985718,
      "step": 2244
    },
    {
      "epoch": 2.1219281663516067,
      "grad_norm": 0.3971850811147425,
      "learning_rate": 2.1695791489352345e-05,
      "loss": 0.416167289018631,
      "step": 2245
    },
    {
      "epoch": 2.122873345935728,
      "grad_norm": 0.4192776605833976,
      "learning_rate": 2.1652858871204214e-05,
      "loss": 0.474107027053833,
      "step": 2246
    },
    {
      "epoch": 2.123818525519849,
      "grad_norm": 0.4050465248947026,
      "learning_rate": 2.16099570306269e-05,
      "loss": 0.43649375438690186,
      "step": 2247
    },
    {
      "epoch": 2.1247637051039696,
      "grad_norm": 0.4531628102154596,
      "learning_rate": 2.156708601420053e-05,
      "loss": 0.433160662651062,
      "step": 2248
    },
    {
      "epoch": 2.1257088846880907,
      "grad_norm": 0.41261757628503176,
      "learning_rate": 2.1524245868471842e-05,
      "loss": 0.46831014752388,
      "step": 2249
    },
    {
      "epoch": 2.126654064272212,
      "grad_norm": 0.4089993143170721,
      "learning_rate": 2.1481436639953984e-05,
      "loss": 0.43478822708129883,
      "step": 2250
    },
    {
      "epoch": 2.1275992438563325,
      "grad_norm": 0.4123566996011664,
      "learning_rate": 2.1438658375126542e-05,
      "loss": 0.41455674171447754,
      "step": 2251
    },
    {
      "epoch": 2.1285444234404536,
      "grad_norm": 0.39033574516339226,
      "learning_rate": 2.139591112043557e-05,
      "loss": 0.4035297632217407,
      "step": 2252
    },
    {
      "epoch": 2.1294896030245747,
      "grad_norm": 0.4285053004068043,
      "learning_rate": 2.135319492229334e-05,
      "loss": 0.4527350664138794,
      "step": 2253
    },
    {
      "epoch": 2.130434782608696,
      "grad_norm": 0.4033109337753081,
      "learning_rate": 2.131050982707849e-05,
      "loss": 0.425895094871521,
      "step": 2254
    },
    {
      "epoch": 2.1313799621928164,
      "grad_norm": 0.392106649876054,
      "learning_rate": 2.126785588113584e-05,
      "loss": 0.3893361985683441,
      "step": 2255
    },
    {
      "epoch": 2.1323251417769375,
      "grad_norm": 0.4139524677688795,
      "learning_rate": 2.122523313077638e-05,
      "loss": 0.45040902495384216,
      "step": 2256
    },
    {
      "epoch": 2.1332703213610587,
      "grad_norm": 0.5339032100265028,
      "learning_rate": 2.1182641622277273e-05,
      "loss": 0.41967350244522095,
      "step": 2257
    },
    {
      "epoch": 2.1342155009451798,
      "grad_norm": 0.3726216442023657,
      "learning_rate": 2.1140081401881744e-05,
      "loss": 0.39142948389053345,
      "step": 2258
    },
    {
      "epoch": 2.1351606805293004,
      "grad_norm": 0.4296216481822586,
      "learning_rate": 2.109755251579906e-05,
      "loss": 0.4664303660392761,
      "step": 2259
    },
    {
      "epoch": 2.1361058601134215,
      "grad_norm": 0.40322912744555284,
      "learning_rate": 2.1055055010204427e-05,
      "loss": 0.42979103326797485,
      "step": 2260
    },
    {
      "epoch": 2.1370510396975426,
      "grad_norm": 0.47439692865241084,
      "learning_rate": 2.101258893123903e-05,
      "loss": 0.4591558277606964,
      "step": 2261
    },
    {
      "epoch": 2.1379962192816637,
      "grad_norm": 0.36657686584048055,
      "learning_rate": 2.097015432500987e-05,
      "loss": 0.3778771162033081,
      "step": 2262
    },
    {
      "epoch": 2.1389413988657844,
      "grad_norm": 0.39446256946064245,
      "learning_rate": 2.0927751237589847e-05,
      "loss": 0.4066808223724365,
      "step": 2263
    },
    {
      "epoch": 2.1398865784499055,
      "grad_norm": 0.3860301424386834,
      "learning_rate": 2.088537971501761e-05,
      "loss": 0.41214948892593384,
      "step": 2264
    },
    {
      "epoch": 2.1408317580340266,
      "grad_norm": 0.4873736222563265,
      "learning_rate": 2.0843039803297516e-05,
      "loss": 0.4342212378978729,
      "step": 2265
    },
    {
      "epoch": 2.1417769376181472,
      "grad_norm": 0.42338606661634565,
      "learning_rate": 2.0800731548399637e-05,
      "loss": 0.41329702734947205,
      "step": 2266
    },
    {
      "epoch": 2.1427221172022684,
      "grad_norm": 0.42576423618794873,
      "learning_rate": 2.075845499625964e-05,
      "loss": 0.45578914880752563,
      "step": 2267
    },
    {
      "epoch": 2.1436672967863895,
      "grad_norm": 0.48096479655814467,
      "learning_rate": 2.071621019277879e-05,
      "loss": 0.479977548122406,
      "step": 2268
    },
    {
      "epoch": 2.1446124763705106,
      "grad_norm": 1.425931222760771,
      "learning_rate": 2.06739971838239e-05,
      "loss": 0.3961474597454071,
      "step": 2269
    },
    {
      "epoch": 2.145557655954631,
      "grad_norm": 0.3982605172126214,
      "learning_rate": 2.0631816015227218e-05,
      "loss": 0.393368661403656,
      "step": 2270
    },
    {
      "epoch": 2.1465028355387523,
      "grad_norm": 0.475068312386899,
      "learning_rate": 2.0589666732786465e-05,
      "loss": 0.44606107473373413,
      "step": 2271
    },
    {
      "epoch": 2.1474480151228734,
      "grad_norm": 0.42434267264609055,
      "learning_rate": 2.05475493822647e-05,
      "loss": 0.48994940519332886,
      "step": 2272
    },
    {
      "epoch": 2.1483931947069945,
      "grad_norm": 0.4463051694990245,
      "learning_rate": 2.050546400939034e-05,
      "loss": 0.4604531526565552,
      "step": 2273
    },
    {
      "epoch": 2.149338374291115,
      "grad_norm": 0.44224613408699925,
      "learning_rate": 2.0463410659857106e-05,
      "loss": 0.45417752861976624,
      "step": 2274
    },
    {
      "epoch": 2.1502835538752363,
      "grad_norm": 0.41579801733828237,
      "learning_rate": 2.042138937932388e-05,
      "loss": 0.4545689523220062,
      "step": 2275
    },
    {
      "epoch": 2.1512287334593574,
      "grad_norm": 0.41413758088725666,
      "learning_rate": 2.03794002134148e-05,
      "loss": 0.4444429874420166,
      "step": 2276
    },
    {
      "epoch": 2.1521739130434785,
      "grad_norm": 0.4567390023019892,
      "learning_rate": 2.0337443207719076e-05,
      "loss": 0.4728860855102539,
      "step": 2277
    },
    {
      "epoch": 2.153119092627599,
      "grad_norm": 0.43222141606482617,
      "learning_rate": 2.0295518407791055e-05,
      "loss": 0.4377979040145874,
      "step": 2278
    },
    {
      "epoch": 2.1540642722117203,
      "grad_norm": 0.4039325122026116,
      "learning_rate": 2.0253625859150056e-05,
      "loss": 0.44620129466056824,
      "step": 2279
    },
    {
      "epoch": 2.1550094517958414,
      "grad_norm": 0.43174500937447,
      "learning_rate": 2.021176560728043e-05,
      "loss": 0.45099252462387085,
      "step": 2280
    },
    {
      "epoch": 2.155954631379962,
      "grad_norm": 0.41965877636922777,
      "learning_rate": 2.016993769763147e-05,
      "loss": 0.4309593737125397,
      "step": 2281
    },
    {
      "epoch": 2.156899810964083,
      "grad_norm": 0.386571946025387,
      "learning_rate": 2.012814217561729e-05,
      "loss": 0.41073453426361084,
      "step": 2282
    },
    {
      "epoch": 2.1578449905482042,
      "grad_norm": 0.3868908061672237,
      "learning_rate": 2.0086379086616912e-05,
      "loss": 0.4384550452232361,
      "step": 2283
    },
    {
      "epoch": 2.1587901701323253,
      "grad_norm": 0.41992524833026734,
      "learning_rate": 2.0044648475974094e-05,
      "loss": 0.4707615375518799,
      "step": 2284
    },
    {
      "epoch": 2.159735349716446,
      "grad_norm": 0.395533940194917,
      "learning_rate": 2.0002950388997345e-05,
      "loss": 0.40836620330810547,
      "step": 2285
    },
    {
      "epoch": 2.160680529300567,
      "grad_norm": 0.4380620241810634,
      "learning_rate": 1.99612848709599e-05,
      "loss": 0.4329989552497864,
      "step": 2286
    },
    {
      "epoch": 2.161625708884688,
      "grad_norm": 0.4579829172133857,
      "learning_rate": 1.9919651967099557e-05,
      "loss": 0.5095086693763733,
      "step": 2287
    },
    {
      "epoch": 2.1625708884688093,
      "grad_norm": 0.42919485283854075,
      "learning_rate": 1.9878051722618784e-05,
      "loss": 0.4321185350418091,
      "step": 2288
    },
    {
      "epoch": 2.16351606805293,
      "grad_norm": 0.44335218228789963,
      "learning_rate": 1.983648418268451e-05,
      "loss": 0.44879651069641113,
      "step": 2289
    },
    {
      "epoch": 2.164461247637051,
      "grad_norm": 0.4178982050941664,
      "learning_rate": 1.979494939242822e-05,
      "loss": 0.4254727363586426,
      "step": 2290
    },
    {
      "epoch": 2.165406427221172,
      "grad_norm": 0.37820964735080864,
      "learning_rate": 1.9753447396945828e-05,
      "loss": 0.41968098282814026,
      "step": 2291
    },
    {
      "epoch": 2.166351606805293,
      "grad_norm": 0.39653361046787194,
      "learning_rate": 1.9711978241297603e-05,
      "loss": 0.40162724256515503,
      "step": 2292
    },
    {
      "epoch": 2.167296786389414,
      "grad_norm": 0.4208245218096597,
      "learning_rate": 1.9670541970508223e-05,
      "loss": 0.3912498950958252,
      "step": 2293
    },
    {
      "epoch": 2.168241965973535,
      "grad_norm": 0.4469050848609377,
      "learning_rate": 1.9629138629566578e-05,
      "loss": 0.4836313724517822,
      "step": 2294
    },
    {
      "epoch": 2.169187145557656,
      "grad_norm": 0.447220723408727,
      "learning_rate": 1.9587768263425886e-05,
      "loss": 0.43168407678604126,
      "step": 2295
    },
    {
      "epoch": 2.170132325141777,
      "grad_norm": 0.4489161679067302,
      "learning_rate": 1.954643091700354e-05,
      "loss": 0.4685268998146057,
      "step": 2296
    },
    {
      "epoch": 2.171077504725898,
      "grad_norm": 0.4081582345308543,
      "learning_rate": 1.950512663518103e-05,
      "loss": 0.43070903420448303,
      "step": 2297
    },
    {
      "epoch": 2.172022684310019,
      "grad_norm": 0.3717359114363842,
      "learning_rate": 1.9463855462804036e-05,
      "loss": 0.364361047744751,
      "step": 2298
    },
    {
      "epoch": 2.17296786389414,
      "grad_norm": 0.4085670051891673,
      "learning_rate": 1.9422617444682196e-05,
      "loss": 0.4269847869873047,
      "step": 2299
    },
    {
      "epoch": 2.1739130434782608,
      "grad_norm": 0.42110700461373735,
      "learning_rate": 1.9381412625589234e-05,
      "loss": 0.4063974618911743,
      "step": 2300
    },
    {
      "epoch": 2.174858223062382,
      "grad_norm": 0.41837609116240226,
      "learning_rate": 1.934024105026276e-05,
      "loss": 0.4069247841835022,
      "step": 2301
    },
    {
      "epoch": 2.175803402646503,
      "grad_norm": 0.3977725697479642,
      "learning_rate": 1.9299102763404336e-05,
      "loss": 0.40714317560195923,
      "step": 2302
    },
    {
      "epoch": 2.1767485822306236,
      "grad_norm": 0.5041295720476653,
      "learning_rate": 1.9257997809679383e-05,
      "loss": 0.5283827781677246,
      "step": 2303
    },
    {
      "epoch": 2.1776937618147447,
      "grad_norm": 0.3861366910122589,
      "learning_rate": 1.9216926233717085e-05,
      "loss": 0.34970659017562866,
      "step": 2304
    },
    {
      "epoch": 2.178638941398866,
      "grad_norm": 0.4100925331546122,
      "learning_rate": 1.917588808011045e-05,
      "loss": 0.4114328622817993,
      "step": 2305
    },
    {
      "epoch": 2.179584120982987,
      "grad_norm": 0.4229903812623607,
      "learning_rate": 1.9134883393416127e-05,
      "loss": 0.43312329053878784,
      "step": 2306
    },
    {
      "epoch": 2.1805293005671076,
      "grad_norm": 0.3932802886099624,
      "learning_rate": 1.9093912218154487e-05,
      "loss": 0.4203762412071228,
      "step": 2307
    },
    {
      "epoch": 2.1814744801512287,
      "grad_norm": 0.44197366318039044,
      "learning_rate": 1.9052974598809503e-05,
      "loss": 0.4257298707962036,
      "step": 2308
    },
    {
      "epoch": 2.18241965973535,
      "grad_norm": 0.41339458428225256,
      "learning_rate": 1.9012070579828684e-05,
      "loss": 0.45401421189308167,
      "step": 2309
    },
    {
      "epoch": 2.183364839319471,
      "grad_norm": 0.449519549984608,
      "learning_rate": 1.897120020562311e-05,
      "loss": 0.43737614154815674,
      "step": 2310
    },
    {
      "epoch": 2.1843100189035916,
      "grad_norm": 0.46749663036867994,
      "learning_rate": 1.893036352056728e-05,
      "loss": 0.44166988134384155,
      "step": 2311
    },
    {
      "epoch": 2.1852551984877127,
      "grad_norm": 0.4165850252189561,
      "learning_rate": 1.8889560568999136e-05,
      "loss": 0.43168431520462036,
      "step": 2312
    },
    {
      "epoch": 2.1862003780718338,
      "grad_norm": 0.4135340918402514,
      "learning_rate": 1.8848791395220027e-05,
      "loss": 0.41234874725341797,
      "step": 2313
    },
    {
      "epoch": 2.1871455576559544,
      "grad_norm": 0.39561166937915,
      "learning_rate": 1.880805604349456e-05,
      "loss": 0.37359338998794556,
      "step": 2314
    },
    {
      "epoch": 2.1880907372400755,
      "grad_norm": 0.45304412779688935,
      "learning_rate": 1.8767354558050693e-05,
      "loss": 0.4899611473083496,
      "step": 2315
    },
    {
      "epoch": 2.1890359168241966,
      "grad_norm": 0.4063683103983855,
      "learning_rate": 1.8726686983079544e-05,
      "loss": 0.4226369857788086,
      "step": 2316
    },
    {
      "epoch": 2.1899810964083177,
      "grad_norm": 0.4338564969353195,
      "learning_rate": 1.8686053362735468e-05,
      "loss": 0.4228600263595581,
      "step": 2317
    },
    {
      "epoch": 2.1909262759924384,
      "grad_norm": 0.4446417825818023,
      "learning_rate": 1.864545374113593e-05,
      "loss": 0.46487075090408325,
      "step": 2318
    },
    {
      "epoch": 2.1918714555765595,
      "grad_norm": 0.4507501006058236,
      "learning_rate": 1.8604888162361507e-05,
      "loss": 0.5081264972686768,
      "step": 2319
    },
    {
      "epoch": 2.1928166351606806,
      "grad_norm": 0.4240587680525219,
      "learning_rate": 1.8564356670455767e-05,
      "loss": 0.43096303939819336,
      "step": 2320
    },
    {
      "epoch": 2.1937618147448017,
      "grad_norm": 0.4447883049490616,
      "learning_rate": 1.852385930942529e-05,
      "loss": 0.4773816168308258,
      "step": 2321
    },
    {
      "epoch": 2.1947069943289224,
      "grad_norm": 0.3841961940450887,
      "learning_rate": 1.8483396123239627e-05,
      "loss": 0.4011695981025696,
      "step": 2322
    },
    {
      "epoch": 2.1956521739130435,
      "grad_norm": 0.4394356109192051,
      "learning_rate": 1.844296715583117e-05,
      "loss": 0.4160184860229492,
      "step": 2323
    },
    {
      "epoch": 2.1965973534971646,
      "grad_norm": 0.4733193400917841,
      "learning_rate": 1.8402572451095208e-05,
      "loss": 0.4811866879463196,
      "step": 2324
    },
    {
      "epoch": 2.197542533081285,
      "grad_norm": 0.4375125136183248,
      "learning_rate": 1.8362212052889826e-05,
      "loss": 0.4381261467933655,
      "step": 2325
    },
    {
      "epoch": 2.1984877126654063,
      "grad_norm": 0.46705583510771204,
      "learning_rate": 1.832188600503581e-05,
      "loss": 0.4454651474952698,
      "step": 2326
    },
    {
      "epoch": 2.1994328922495274,
      "grad_norm": 0.4931334078087394,
      "learning_rate": 1.828159435131672e-05,
      "loss": 0.5129187107086182,
      "step": 2327
    },
    {
      "epoch": 2.2003780718336485,
      "grad_norm": 0.4234719283825417,
      "learning_rate": 1.8241337135478707e-05,
      "loss": 0.40480366349220276,
      "step": 2328
    },
    {
      "epoch": 2.201323251417769,
      "grad_norm": 0.3946334007884592,
      "learning_rate": 1.8201114401230585e-05,
      "loss": 0.35328763723373413,
      "step": 2329
    },
    {
      "epoch": 2.2022684310018903,
      "grad_norm": 0.41242146124738216,
      "learning_rate": 1.8160926192243698e-05,
      "loss": 0.39752280712127686,
      "step": 2330
    },
    {
      "epoch": 2.2032136105860114,
      "grad_norm": 0.3973368844393759,
      "learning_rate": 1.8120772552151936e-05,
      "loss": 0.3845710754394531,
      "step": 2331
    },
    {
      "epoch": 2.2041587901701325,
      "grad_norm": 0.40382673395708957,
      "learning_rate": 1.8080653524551623e-05,
      "loss": 0.40469080209732056,
      "step": 2332
    },
    {
      "epoch": 2.205103969754253,
      "grad_norm": 0.49131892898853274,
      "learning_rate": 1.8040569153001496e-05,
      "loss": 0.4733401834964752,
      "step": 2333
    },
    {
      "epoch": 2.2060491493383743,
      "grad_norm": 0.4475770605382421,
      "learning_rate": 1.80005194810227e-05,
      "loss": 0.49803367257118225,
      "step": 2334
    },
    {
      "epoch": 2.2069943289224954,
      "grad_norm": 0.4339260354499076,
      "learning_rate": 1.796050455209869e-05,
      "loss": 0.4618058502674103,
      "step": 2335
    },
    {
      "epoch": 2.207939508506616,
      "grad_norm": 0.4542955145711345,
      "learning_rate": 1.7920524409675215e-05,
      "loss": 0.39934080839157104,
      "step": 2336
    },
    {
      "epoch": 2.208884688090737,
      "grad_norm": 0.452360564806109,
      "learning_rate": 1.7880579097160217e-05,
      "loss": 0.4692729115486145,
      "step": 2337
    },
    {
      "epoch": 2.2098298676748582,
      "grad_norm": 0.46993215282612044,
      "learning_rate": 1.7840668657923835e-05,
      "loss": 0.4678827226161957,
      "step": 2338
    },
    {
      "epoch": 2.2107750472589793,
      "grad_norm": 0.46390275483060417,
      "learning_rate": 1.7800793135298384e-05,
      "loss": 0.46104639768600464,
      "step": 2339
    },
    {
      "epoch": 2.2117202268431,
      "grad_norm": 0.3746589432727812,
      "learning_rate": 1.7760952572578182e-05,
      "loss": 0.33611860871315,
      "step": 2340
    },
    {
      "epoch": 2.212665406427221,
      "grad_norm": 0.4027499892247036,
      "learning_rate": 1.7721147013019717e-05,
      "loss": 0.41428297758102417,
      "step": 2341
    },
    {
      "epoch": 2.213610586011342,
      "grad_norm": 0.4632215481965366,
      "learning_rate": 1.7681376499841363e-05,
      "loss": 0.48332479596138,
      "step": 2342
    },
    {
      "epoch": 2.2145557655954633,
      "grad_norm": 0.4434543351919962,
      "learning_rate": 1.7641641076223475e-05,
      "loss": 0.4360387325286865,
      "step": 2343
    },
    {
      "epoch": 2.215500945179584,
      "grad_norm": 0.4968809786700685,
      "learning_rate": 1.7601940785308336e-05,
      "loss": 0.480716347694397,
      "step": 2344
    },
    {
      "epoch": 2.216446124763705,
      "grad_norm": 0.4618743487554191,
      "learning_rate": 1.756227567020004e-05,
      "loss": 0.4529496729373932,
      "step": 2345
    },
    {
      "epoch": 2.217391304347826,
      "grad_norm": 0.43565699958593723,
      "learning_rate": 1.7522645773964525e-05,
      "loss": 0.4529123902320862,
      "step": 2346
    },
    {
      "epoch": 2.2183364839319473,
      "grad_norm": 0.4328309830381412,
      "learning_rate": 1.748305113962948e-05,
      "loss": 0.4332658648490906,
      "step": 2347
    },
    {
      "epoch": 2.219281663516068,
      "grad_norm": 0.4410950447569124,
      "learning_rate": 1.7443491810184326e-05,
      "loss": 0.4707214832305908,
      "step": 2348
    },
    {
      "epoch": 2.220226843100189,
      "grad_norm": 0.4299571970171896,
      "learning_rate": 1.740396782858012e-05,
      "loss": 0.42847269773483276,
      "step": 2349
    },
    {
      "epoch": 2.22117202268431,
      "grad_norm": 0.41114973686018974,
      "learning_rate": 1.7364479237729526e-05,
      "loss": 0.42042067646980286,
      "step": 2350
    },
    {
      "epoch": 2.2221172022684312,
      "grad_norm": 0.4216256928563654,
      "learning_rate": 1.732502608050684e-05,
      "loss": 0.42964428663253784,
      "step": 2351
    },
    {
      "epoch": 2.223062381852552,
      "grad_norm": 0.41444444742103714,
      "learning_rate": 1.728560839974785e-05,
      "loss": 0.3795408606529236,
      "step": 2352
    },
    {
      "epoch": 2.224007561436673,
      "grad_norm": 0.4630787057314805,
      "learning_rate": 1.724622623824985e-05,
      "loss": 0.4509872794151306,
      "step": 2353
    },
    {
      "epoch": 2.224952741020794,
      "grad_norm": 0.4922013989334339,
      "learning_rate": 1.7206879638771546e-05,
      "loss": 0.4575501084327698,
      "step": 2354
    },
    {
      "epoch": 2.2258979206049148,
      "grad_norm": 0.46475619827180026,
      "learning_rate": 1.7167568644033005e-05,
      "loss": 0.427565336227417,
      "step": 2355
    },
    {
      "epoch": 2.226843100189036,
      "grad_norm": 0.4385085822258331,
      "learning_rate": 1.7128293296715703e-05,
      "loss": 0.4047184884548187,
      "step": 2356
    },
    {
      "epoch": 2.227788279773157,
      "grad_norm": 0.47762784863708785,
      "learning_rate": 1.7089053639462377e-05,
      "loss": 0.48844975233078003,
      "step": 2357
    },
    {
      "epoch": 2.228733459357278,
      "grad_norm": 0.4310706877154517,
      "learning_rate": 1.7049849714877035e-05,
      "loss": 0.47752678394317627,
      "step": 2358
    },
    {
      "epoch": 2.2296786389413987,
      "grad_norm": 0.41159704666528557,
      "learning_rate": 1.7010681565524835e-05,
      "loss": 0.44635480642318726,
      "step": 2359
    },
    {
      "epoch": 2.23062381852552,
      "grad_norm": 0.4225648681779558,
      "learning_rate": 1.697154923393216e-05,
      "loss": 0.45756715536117554,
      "step": 2360
    },
    {
      "epoch": 2.231568998109641,
      "grad_norm": 0.4666116941252373,
      "learning_rate": 1.693245276258646e-05,
      "loss": 0.4498025178909302,
      "step": 2361
    },
    {
      "epoch": 2.232514177693762,
      "grad_norm": 0.4621073010625924,
      "learning_rate": 1.689339219393623e-05,
      "loss": 0.48739975690841675,
      "step": 2362
    },
    {
      "epoch": 2.2334593572778827,
      "grad_norm": 0.40443815740956474,
      "learning_rate": 1.6854367570391077e-05,
      "loss": 0.4202251434326172,
      "step": 2363
    },
    {
      "epoch": 2.234404536862004,
      "grad_norm": 0.4303101910135521,
      "learning_rate": 1.6815378934321486e-05,
      "loss": 0.4223949611186981,
      "step": 2364
    },
    {
      "epoch": 2.235349716446125,
      "grad_norm": 0.4687451407349286,
      "learning_rate": 1.677642632805892e-05,
      "loss": 0.47713661193847656,
      "step": 2365
    },
    {
      "epoch": 2.2362948960302456,
      "grad_norm": 0.46079834869243563,
      "learning_rate": 1.67375097938957e-05,
      "loss": 0.43819722533226013,
      "step": 2366
    },
    {
      "epoch": 2.2372400756143667,
      "grad_norm": 0.4430399869967913,
      "learning_rate": 1.669862937408497e-05,
      "loss": 0.4758162200450897,
      "step": 2367
    },
    {
      "epoch": 2.2381852551984878,
      "grad_norm": 0.4454288835191172,
      "learning_rate": 1.6659785110840704e-05,
      "loss": 0.488650918006897,
      "step": 2368
    },
    {
      "epoch": 2.239130434782609,
      "grad_norm": 0.41921659202556183,
      "learning_rate": 1.6620977046337594e-05,
      "loss": 0.41195666790008545,
      "step": 2369
    },
    {
      "epoch": 2.2400756143667295,
      "grad_norm": 0.4491142746246426,
      "learning_rate": 1.658220522271105e-05,
      "loss": 0.47994378209114075,
      "step": 2370
    },
    {
      "epoch": 2.2410207939508506,
      "grad_norm": 0.3964153980692464,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.3978731036186218,
      "step": 2371
    },
    {
      "epoch": 2.2419659735349717,
      "grad_norm": 0.458779079494377,
      "learning_rate": 1.650477046643239e-05,
      "loss": 0.4511764645576477,
      "step": 2372
    },
    {
      "epoch": 2.242911153119093,
      "grad_norm": 0.40312946024745905,
      "learning_rate": 1.6466107617854133e-05,
      "loss": 0.41725730895996094,
      "step": 2373
    },
    {
      "epoch": 2.2438563327032135,
      "grad_norm": 0.4212987068469432,
      "learning_rate": 1.6427481178300062e-05,
      "loss": 0.442128449678421,
      "step": 2374
    },
    {
      "epoch": 2.2448015122873346,
      "grad_norm": 0.4330345696834623,
      "learning_rate": 1.63888911897084e-05,
      "loss": 0.41599318385124207,
      "step": 2375
    },
    {
      "epoch": 2.2457466918714557,
      "grad_norm": 0.4741359317653155,
      "learning_rate": 1.635033769397772e-05,
      "loss": 0.4591531455516815,
      "step": 2376
    },
    {
      "epoch": 2.2466918714555764,
      "grad_norm": 0.4411957717894831,
      "learning_rate": 1.631182073296706e-05,
      "loss": 0.45284515619277954,
      "step": 2377
    },
    {
      "epoch": 2.2476370510396975,
      "grad_norm": 0.37107023909911707,
      "learning_rate": 1.627334034849572e-05,
      "loss": 0.39474397897720337,
      "step": 2378
    },
    {
      "epoch": 2.2485822306238186,
      "grad_norm": 0.4521101001578554,
      "learning_rate": 1.623489658234333e-05,
      "loss": 0.46627819538116455,
      "step": 2379
    },
    {
      "epoch": 2.2495274102079397,
      "grad_norm": 0.4191772124517357,
      "learning_rate": 1.6196489476249777e-05,
      "loss": 0.4094266891479492,
      "step": 2380
    },
    {
      "epoch": 2.2504725897920603,
      "grad_norm": 0.4210694214200305,
      "learning_rate": 1.6158119071915075e-05,
      "loss": 0.4247094392776489,
      "step": 2381
    },
    {
      "epoch": 2.2514177693761814,
      "grad_norm": 0.4307494852627218,
      "learning_rate": 1.6119785410999478e-05,
      "loss": 0.38465046882629395,
      "step": 2382
    },
    {
      "epoch": 2.2523629489603025,
      "grad_norm": 0.43965451125949,
      "learning_rate": 1.6081488535123275e-05,
      "loss": 0.4609178304672241,
      "step": 2383
    },
    {
      "epoch": 2.2533081285444236,
      "grad_norm": 0.40661328736903335,
      "learning_rate": 1.6043228485866812e-05,
      "loss": 0.39696764945983887,
      "step": 2384
    },
    {
      "epoch": 2.2542533081285443,
      "grad_norm": 0.4301320393471178,
      "learning_rate": 1.6005005304770552e-05,
      "loss": 0.4097920358181,
      "step": 2385
    },
    {
      "epoch": 2.2551984877126654,
      "grad_norm": 0.49520898155587917,
      "learning_rate": 1.5966819033334805e-05,
      "loss": 0.48843851685523987,
      "step": 2386
    },
    {
      "epoch": 2.2561436672967865,
      "grad_norm": 0.4276463177441401,
      "learning_rate": 1.59286697130199e-05,
      "loss": 0.37779200077056885,
      "step": 2387
    },
    {
      "epoch": 2.257088846880907,
      "grad_norm": 0.3953002665258101,
      "learning_rate": 1.5890557385245986e-05,
      "loss": 0.40672942996025085,
      "step": 2388
    },
    {
      "epoch": 2.2580340264650283,
      "grad_norm": 0.4540322680241016,
      "learning_rate": 1.5852482091393044e-05,
      "loss": 0.46225443482398987,
      "step": 2389
    },
    {
      "epoch": 2.2589792060491494,
      "grad_norm": 0.39898542187101227,
      "learning_rate": 1.5814443872800906e-05,
      "loss": 0.39813950657844543,
      "step": 2390
    },
    {
      "epoch": 2.2599243856332705,
      "grad_norm": 0.427136320890409,
      "learning_rate": 1.5776442770769094e-05,
      "loss": 0.4056336581707001,
      "step": 2391
    },
    {
      "epoch": 2.260869565217391,
      "grad_norm": 0.43481539969525373,
      "learning_rate": 1.5738478826556886e-05,
      "loss": 0.454784631729126,
      "step": 2392
    },
    {
      "epoch": 2.2618147448015122,
      "grad_norm": 0.4173062283439494,
      "learning_rate": 1.5700552081383134e-05,
      "loss": 0.4608730375766754,
      "step": 2393
    },
    {
      "epoch": 2.2627599243856333,
      "grad_norm": 0.4535778214167699,
      "learning_rate": 1.5662662576426396e-05,
      "loss": 0.4625069499015808,
      "step": 2394
    },
    {
      "epoch": 2.2637051039697544,
      "grad_norm": 0.4592686183320719,
      "learning_rate": 1.562481035282471e-05,
      "loss": 0.4256003201007843,
      "step": 2395
    },
    {
      "epoch": 2.264650283553875,
      "grad_norm": 0.4052447580306782,
      "learning_rate": 1.5586995451675696e-05,
      "loss": 0.4257810115814209,
      "step": 2396
    },
    {
      "epoch": 2.265595463137996,
      "grad_norm": 0.47496385602556457,
      "learning_rate": 1.5549217914036445e-05,
      "loss": 0.4764748215675354,
      "step": 2397
    },
    {
      "epoch": 2.2665406427221173,
      "grad_norm": 0.3966562249064388,
      "learning_rate": 1.5511477780923438e-05,
      "loss": 0.4114837050437927,
      "step": 2398
    },
    {
      "epoch": 2.267485822306238,
      "grad_norm": 0.44132121519716005,
      "learning_rate": 1.5473775093312615e-05,
      "loss": 0.43253713846206665,
      "step": 2399
    },
    {
      "epoch": 2.268431001890359,
      "grad_norm": 0.3612071111601426,
      "learning_rate": 1.5436109892139177e-05,
      "loss": 0.37437403202056885,
      "step": 2400
    },
    {
      "epoch": 2.26937618147448,
      "grad_norm": 0.3827402584319773,
      "learning_rate": 1.539848221829769e-05,
      "loss": 0.4056958556175232,
      "step": 2401
    },
    {
      "epoch": 2.2703213610586013,
      "grad_norm": 0.4767300674868224,
      "learning_rate": 1.536089211264196e-05,
      "loss": 0.5118268728256226,
      "step": 2402
    },
    {
      "epoch": 2.271266540642722,
      "grad_norm": 0.47204740639455406,
      "learning_rate": 1.5323339615984978e-05,
      "loss": 0.4731203317642212,
      "step": 2403
    },
    {
      "epoch": 2.272211720226843,
      "grad_norm": 0.4628334852369142,
      "learning_rate": 1.5285824769098937e-05,
      "loss": 0.43852555751800537,
      "step": 2404
    },
    {
      "epoch": 2.273156899810964,
      "grad_norm": 0.4010029743728272,
      "learning_rate": 1.5248347612715119e-05,
      "loss": 0.4300031065940857,
      "step": 2405
    },
    {
      "epoch": 2.2741020793950852,
      "grad_norm": 0.4580410165548389,
      "learning_rate": 1.521090818752392e-05,
      "loss": 0.45832961797714233,
      "step": 2406
    },
    {
      "epoch": 2.275047258979206,
      "grad_norm": 0.45161930275793266,
      "learning_rate": 1.5173506534174719e-05,
      "loss": 0.44092440605163574,
      "step": 2407
    },
    {
      "epoch": 2.275992438563327,
      "grad_norm": 0.4756943041847318,
      "learning_rate": 1.5136142693275934e-05,
      "loss": 0.4377001225948334,
      "step": 2408
    },
    {
      "epoch": 2.276937618147448,
      "grad_norm": 0.462919011125249,
      "learning_rate": 1.5098816705394931e-05,
      "loss": 0.47647956013679504,
      "step": 2409
    },
    {
      "epoch": 2.2778827977315688,
      "grad_norm": 0.45622498689474666,
      "learning_rate": 1.5061528611057918e-05,
      "loss": 0.4255296587944031,
      "step": 2410
    },
    {
      "epoch": 2.27882797731569,
      "grad_norm": 0.43581684863660025,
      "learning_rate": 1.5024278450750034e-05,
      "loss": 0.4405800700187683,
      "step": 2411
    },
    {
      "epoch": 2.279773156899811,
      "grad_norm": 0.40366862505023976,
      "learning_rate": 1.4987066264915168e-05,
      "loss": 0.4109494090080261,
      "step": 2412
    },
    {
      "epoch": 2.280718336483932,
      "grad_norm": 0.4107010228195561,
      "learning_rate": 1.4949892093956014e-05,
      "loss": 0.4141641855239868,
      "step": 2413
    },
    {
      "epoch": 2.281663516068053,
      "grad_norm": 0.45422293901729954,
      "learning_rate": 1.4912755978234005e-05,
      "loss": 0.4772476553916931,
      "step": 2414
    },
    {
      "epoch": 2.282608695652174,
      "grad_norm": 0.4568817148615888,
      "learning_rate": 1.4875657958069212e-05,
      "loss": 0.44183483719825745,
      "step": 2415
    },
    {
      "epoch": 2.283553875236295,
      "grad_norm": 0.45015360866518167,
      "learning_rate": 1.4838598073740395e-05,
      "loss": 0.44287723302841187,
      "step": 2416
    },
    {
      "epoch": 2.284499054820416,
      "grad_norm": 0.42816056889923076,
      "learning_rate": 1.4801576365484842e-05,
      "loss": 0.43063685297966003,
      "step": 2417
    },
    {
      "epoch": 2.2854442344045367,
      "grad_norm": 0.3955214275139481,
      "learning_rate": 1.4764592873498462e-05,
      "loss": 0.40761715173721313,
      "step": 2418
    },
    {
      "epoch": 2.286389413988658,
      "grad_norm": 0.3929979908258227,
      "learning_rate": 1.4727647637935649e-05,
      "loss": 0.39713233709335327,
      "step": 2419
    },
    {
      "epoch": 2.287334593572779,
      "grad_norm": 0.45314037900932075,
      "learning_rate": 1.4690740698909222e-05,
      "loss": 0.49254024028778076,
      "step": 2420
    },
    {
      "epoch": 2.2882797731568996,
      "grad_norm": 0.4423880684303732,
      "learning_rate": 1.4653872096490484e-05,
      "loss": 0.442094087600708,
      "step": 2421
    },
    {
      "epoch": 2.2892249527410207,
      "grad_norm": 0.4359593810430339,
      "learning_rate": 1.4617041870709042e-05,
      "loss": 0.42633795738220215,
      "step": 2422
    },
    {
      "epoch": 2.2901701323251418,
      "grad_norm": 0.41831124044760215,
      "learning_rate": 1.4580250061552903e-05,
      "loss": 0.4457341432571411,
      "step": 2423
    },
    {
      "epoch": 2.291115311909263,
      "grad_norm": 0.38393245247425,
      "learning_rate": 1.4543496708968335e-05,
      "loss": 0.3952561616897583,
      "step": 2424
    },
    {
      "epoch": 2.292060491493384,
      "grad_norm": 0.47228447507937243,
      "learning_rate": 1.4506781852859835e-05,
      "loss": 0.4780524969100952,
      "step": 2425
    },
    {
      "epoch": 2.2930056710775046,
      "grad_norm": 0.4096528646417563,
      "learning_rate": 1.447010553309015e-05,
      "loss": 0.4137040376663208,
      "step": 2426
    },
    {
      "epoch": 2.2939508506616257,
      "grad_norm": 0.47742446469507144,
      "learning_rate": 1.4433467789480121e-05,
      "loss": 0.4278452396392822,
      "step": 2427
    },
    {
      "epoch": 2.294896030245747,
      "grad_norm": 0.4434500894686418,
      "learning_rate": 1.4396868661808777e-05,
      "loss": 0.4565833508968353,
      "step": 2428
    },
    {
      "epoch": 2.2958412098298675,
      "grad_norm": 0.42268742881766397,
      "learning_rate": 1.4360308189813149e-05,
      "loss": 0.4276459217071533,
      "step": 2429
    },
    {
      "epoch": 2.2967863894139886,
      "grad_norm": 0.39743971748916945,
      "learning_rate": 1.432378641318835e-05,
      "loss": 0.39310890436172485,
      "step": 2430
    },
    {
      "epoch": 2.2977315689981097,
      "grad_norm": 0.46143708749072404,
      "learning_rate": 1.428730337158749e-05,
      "loss": 0.4736470580101013,
      "step": 2431
    },
    {
      "epoch": 2.298676748582231,
      "grad_norm": 0.4232656073801379,
      "learning_rate": 1.425085910462155e-05,
      "loss": 0.4380645751953125,
      "step": 2432
    },
    {
      "epoch": 2.2996219281663515,
      "grad_norm": 0.4613348631394816,
      "learning_rate": 1.4214453651859504e-05,
      "loss": 0.45357757806777954,
      "step": 2433
    },
    {
      "epoch": 2.3005671077504726,
      "grad_norm": 0.4084033299551755,
      "learning_rate": 1.4178087052828098e-05,
      "loss": 0.3644987940788269,
      "step": 2434
    },
    {
      "epoch": 2.3015122873345937,
      "grad_norm": 0.40469729624893996,
      "learning_rate": 1.4141759347011952e-05,
      "loss": 0.4076794385910034,
      "step": 2435
    },
    {
      "epoch": 2.302457466918715,
      "grad_norm": 0.4476486117177362,
      "learning_rate": 1.4105470573853458e-05,
      "loss": 0.4453621208667755,
      "step": 2436
    },
    {
      "epoch": 2.3034026465028354,
      "grad_norm": 0.42024637541574533,
      "learning_rate": 1.4069220772752683e-05,
      "loss": 0.4414924681186676,
      "step": 2437
    },
    {
      "epoch": 2.3043478260869565,
      "grad_norm": 0.4778753834983097,
      "learning_rate": 1.4033009983067452e-05,
      "loss": 0.4696066677570343,
      "step": 2438
    },
    {
      "epoch": 2.3052930056710776,
      "grad_norm": 0.42982831985253384,
      "learning_rate": 1.399683824411317e-05,
      "loss": 0.46618133783340454,
      "step": 2439
    },
    {
      "epoch": 2.3062381852551983,
      "grad_norm": 0.48255052315024083,
      "learning_rate": 1.3960705595162876e-05,
      "loss": 0.44435468316078186,
      "step": 2440
    },
    {
      "epoch": 2.3071833648393194,
      "grad_norm": 0.4560844968305203,
      "learning_rate": 1.3924612075447191e-05,
      "loss": 0.4540443420410156,
      "step": 2441
    },
    {
      "epoch": 2.3081285444234405,
      "grad_norm": 0.4090359018739647,
      "learning_rate": 1.388855772415419e-05,
      "loss": 0.3859800696372986,
      "step": 2442
    },
    {
      "epoch": 2.3090737240075616,
      "grad_norm": 0.3940145306491719,
      "learning_rate": 1.3852542580429478e-05,
      "loss": 0.3556947112083435,
      "step": 2443
    },
    {
      "epoch": 2.3100189035916823,
      "grad_norm": 0.4518935773301727,
      "learning_rate": 1.3816566683376047e-05,
      "loss": 0.4415906071662903,
      "step": 2444
    },
    {
      "epoch": 2.3109640831758034,
      "grad_norm": 0.42692762713297394,
      "learning_rate": 1.3780630072054313e-05,
      "loss": 0.439871609210968,
      "step": 2445
    },
    {
      "epoch": 2.3119092627599245,
      "grad_norm": 0.3903315366969254,
      "learning_rate": 1.3744732785482034e-05,
      "loss": 0.3954285681247711,
      "step": 2446
    },
    {
      "epoch": 2.3128544423440456,
      "grad_norm": 0.3807845789248114,
      "learning_rate": 1.3708874862634236e-05,
      "loss": 0.40317508578300476,
      "step": 2447
    },
    {
      "epoch": 2.3137996219281662,
      "grad_norm": 0.4654969358040096,
      "learning_rate": 1.3673056342443258e-05,
      "loss": 0.44394397735595703,
      "step": 2448
    },
    {
      "epoch": 2.3147448015122873,
      "grad_norm": 0.4518008795215234,
      "learning_rate": 1.3637277263798604e-05,
      "loss": 0.42594730854034424,
      "step": 2449
    },
    {
      "epoch": 2.3156899810964084,
      "grad_norm": 0.4539406173219504,
      "learning_rate": 1.3601537665547009e-05,
      "loss": 0.47868651151657104,
      "step": 2450
    },
    {
      "epoch": 2.316635160680529,
      "grad_norm": 0.4580058869377305,
      "learning_rate": 1.356583758649228e-05,
      "loss": 0.45534342527389526,
      "step": 2451
    },
    {
      "epoch": 2.31758034026465,
      "grad_norm": 0.4286749557717541,
      "learning_rate": 1.3530177065395399e-05,
      "loss": 0.4642760157585144,
      "step": 2452
    },
    {
      "epoch": 2.3185255198487713,
      "grad_norm": 0.41530857295459367,
      "learning_rate": 1.3494556140974335e-05,
      "loss": 0.4196838140487671,
      "step": 2453
    },
    {
      "epoch": 2.3194706994328924,
      "grad_norm": 0.4240614151601316,
      "learning_rate": 1.3458974851904054e-05,
      "loss": 0.4481913447380066,
      "step": 2454
    },
    {
      "epoch": 2.320415879017013,
      "grad_norm": 0.38247522841795506,
      "learning_rate": 1.3423433236816563e-05,
      "loss": 0.34103724360466003,
      "step": 2455
    },
    {
      "epoch": 2.321361058601134,
      "grad_norm": 0.45911990057784635,
      "learning_rate": 1.3387931334300701e-05,
      "loss": 0.48332253098487854,
      "step": 2456
    },
    {
      "epoch": 2.3223062381852553,
      "grad_norm": 0.43256015766270345,
      "learning_rate": 1.3352469182902267e-05,
      "loss": 0.40125519037246704,
      "step": 2457
    },
    {
      "epoch": 2.3232514177693764,
      "grad_norm": 0.43589903947250114,
      "learning_rate": 1.3317046821123868e-05,
      "loss": 0.47374463081359863,
      "step": 2458
    },
    {
      "epoch": 2.324196597353497,
      "grad_norm": 0.37742313608281103,
      "learning_rate": 1.328166428742489e-05,
      "loss": 0.31759509444236755,
      "step": 2459
    },
    {
      "epoch": 2.325141776937618,
      "grad_norm": 0.4630786212600488,
      "learning_rate": 1.324632162022153e-05,
      "loss": 0.4759445786476135,
      "step": 2460
    },
    {
      "epoch": 2.3260869565217392,
      "grad_norm": 0.47366058649024584,
      "learning_rate": 1.3211018857886631e-05,
      "loss": 0.4794473350048065,
      "step": 2461
    },
    {
      "epoch": 2.32703213610586,
      "grad_norm": 0.3956255761510245,
      "learning_rate": 1.3175756038749765e-05,
      "loss": 0.41842085123062134,
      "step": 2462
    },
    {
      "epoch": 2.327977315689981,
      "grad_norm": 0.4324917495048681,
      "learning_rate": 1.3140533201097122e-05,
      "loss": 0.426818311214447,
      "step": 2463
    },
    {
      "epoch": 2.328922495274102,
      "grad_norm": 0.4326820426684949,
      "learning_rate": 1.3105350383171483e-05,
      "loss": 0.3942812979221344,
      "step": 2464
    },
    {
      "epoch": 2.329867674858223,
      "grad_norm": 0.4358486762043232,
      "learning_rate": 1.307020762317217e-05,
      "loss": 0.41964468359947205,
      "step": 2465
    },
    {
      "epoch": 2.330812854442344,
      "grad_norm": 0.43298003256296036,
      "learning_rate": 1.303510495925499e-05,
      "loss": 0.41287916898727417,
      "step": 2466
    },
    {
      "epoch": 2.331758034026465,
      "grad_norm": 0.4767556752809987,
      "learning_rate": 1.3000042429532267e-05,
      "loss": 0.43288087844848633,
      "step": 2467
    },
    {
      "epoch": 2.332703213610586,
      "grad_norm": 0.4804254609027467,
      "learning_rate": 1.296502007207272e-05,
      "loss": 0.45004260540008545,
      "step": 2468
    },
    {
      "epoch": 2.333648393194707,
      "grad_norm": 0.4305034303761276,
      "learning_rate": 1.2930037924901473e-05,
      "loss": 0.43385353684425354,
      "step": 2469
    },
    {
      "epoch": 2.334593572778828,
      "grad_norm": 0.45513338843934603,
      "learning_rate": 1.289509602599996e-05,
      "loss": 0.4614972472190857,
      "step": 2470
    },
    {
      "epoch": 2.335538752362949,
      "grad_norm": 0.4638026418667257,
      "learning_rate": 1.2860194413305915e-05,
      "loss": 0.4254574179649353,
      "step": 2471
    },
    {
      "epoch": 2.33648393194707,
      "grad_norm": 0.4491427854888822,
      "learning_rate": 1.2825333124713385e-05,
      "loss": 0.4177279472351074,
      "step": 2472
    },
    {
      "epoch": 2.3374291115311907,
      "grad_norm": 0.39924705977708164,
      "learning_rate": 1.2790512198072558e-05,
      "loss": 0.4053025543689728,
      "step": 2473
    },
    {
      "epoch": 2.338374291115312,
      "grad_norm": 0.41300892428887065,
      "learning_rate": 1.2755731671189858e-05,
      "loss": 0.40452665090560913,
      "step": 2474
    },
    {
      "epoch": 2.339319470699433,
      "grad_norm": 0.40327950678323754,
      "learning_rate": 1.272099158182785e-05,
      "loss": 0.4232214093208313,
      "step": 2475
    },
    {
      "epoch": 2.340264650283554,
      "grad_norm": 0.44918374445118114,
      "learning_rate": 1.2686291967705138e-05,
      "loss": 0.3905335068702698,
      "step": 2476
    },
    {
      "epoch": 2.3412098298676747,
      "grad_norm": 0.4497973384031674,
      "learning_rate": 1.2651632866496443e-05,
      "loss": 0.4445609450340271,
      "step": 2477
    },
    {
      "epoch": 2.3421550094517958,
      "grad_norm": 0.44735092123803466,
      "learning_rate": 1.2617014315832443e-05,
      "loss": 0.4643290638923645,
      "step": 2478
    },
    {
      "epoch": 2.343100189035917,
      "grad_norm": 0.4054424532755233,
      "learning_rate": 1.2582436353299831e-05,
      "loss": 0.3687050938606262,
      "step": 2479
    },
    {
      "epoch": 2.344045368620038,
      "grad_norm": 0.47425158580032534,
      "learning_rate": 1.2547899016441222e-05,
      "loss": 0.3915455937385559,
      "step": 2480
    },
    {
      "epoch": 2.3449905482041586,
      "grad_norm": 0.49681507151303084,
      "learning_rate": 1.2513402342755126e-05,
      "loss": 0.4717596173286438,
      "step": 2481
    },
    {
      "epoch": 2.3459357277882797,
      "grad_norm": 0.41881458969209373,
      "learning_rate": 1.2478946369695882e-05,
      "loss": 0.40554314851760864,
      "step": 2482
    },
    {
      "epoch": 2.346880907372401,
      "grad_norm": 0.41763925643367,
      "learning_rate": 1.2444531134673643e-05,
      "loss": 0.3734094500541687,
      "step": 2483
    },
    {
      "epoch": 2.3478260869565215,
      "grad_norm": 0.5159617867642786,
      "learning_rate": 1.2410156675054352e-05,
      "loss": 0.45290905237197876,
      "step": 2484
    },
    {
      "epoch": 2.3487712665406426,
      "grad_norm": 0.4227198262337864,
      "learning_rate": 1.2375823028159667e-05,
      "loss": 0.41950783133506775,
      "step": 2485
    },
    {
      "epoch": 2.3497164461247637,
      "grad_norm": 0.41434432034896873,
      "learning_rate": 1.2341530231266951e-05,
      "loss": 0.4138210415840149,
      "step": 2486
    },
    {
      "epoch": 2.350661625708885,
      "grad_norm": 0.45236011780575636,
      "learning_rate": 1.2307278321609189e-05,
      "loss": 0.43032950162887573,
      "step": 2487
    },
    {
      "epoch": 2.351606805293006,
      "grad_norm": 0.46970271376239675,
      "learning_rate": 1.2273067336374972e-05,
      "loss": 0.42929309606552124,
      "step": 2488
    },
    {
      "epoch": 2.3525519848771266,
      "grad_norm": 0.4464808788023047,
      "learning_rate": 1.2238897312708475e-05,
      "loss": 0.4561523199081421,
      "step": 2489
    },
    {
      "epoch": 2.3534971644612477,
      "grad_norm": 0.41348755154033073,
      "learning_rate": 1.2204768287709395e-05,
      "loss": 0.43162161111831665,
      "step": 2490
    },
    {
      "epoch": 2.354442344045369,
      "grad_norm": 0.42038393286134096,
      "learning_rate": 1.2170680298432935e-05,
      "loss": 0.42629334330558777,
      "step": 2491
    },
    {
      "epoch": 2.3553875236294894,
      "grad_norm": 0.43705209089183034,
      "learning_rate": 1.2136633381889695e-05,
      "loss": 0.4600503444671631,
      "step": 2492
    },
    {
      "epoch": 2.3563327032136105,
      "grad_norm": 0.48136651265719704,
      "learning_rate": 1.2102627575045734e-05,
      "loss": 0.4690813422203064,
      "step": 2493
    },
    {
      "epoch": 2.3572778827977316,
      "grad_norm": 0.3921520497176565,
      "learning_rate": 1.2068662914822432e-05,
      "loss": 0.4318865239620209,
      "step": 2494
    },
    {
      "epoch": 2.3582230623818523,
      "grad_norm": 0.45205536512394184,
      "learning_rate": 1.203473943809651e-05,
      "loss": 0.46268802881240845,
      "step": 2495
    },
    {
      "epoch": 2.3591682419659734,
      "grad_norm": 0.4265078687917648,
      "learning_rate": 1.2000857181699986e-05,
      "loss": 0.3977741003036499,
      "step": 2496
    },
    {
      "epoch": 2.3601134215500945,
      "grad_norm": 0.41453397239043593,
      "learning_rate": 1.196701618242012e-05,
      "loss": 0.3943415880203247,
      "step": 2497
    },
    {
      "epoch": 2.3610586011342156,
      "grad_norm": 0.41430249988441015,
      "learning_rate": 1.1933216476999393e-05,
      "loss": 0.39513489603996277,
      "step": 2498
    },
    {
      "epoch": 2.3620037807183367,
      "grad_norm": 0.4523488506880437,
      "learning_rate": 1.1899458102135408e-05,
      "loss": 0.4266386032104492,
      "step": 2499
    },
    {
      "epoch": 2.3629489603024574,
      "grad_norm": 0.39399795235034724,
      "learning_rate": 1.1865741094480909e-05,
      "loss": 0.39472663402557373,
      "step": 2500
    },
    {
      "epoch": 2.3638941398865785,
      "grad_norm": 0.42181444547760133,
      "learning_rate": 1.1832065490643746e-05,
      "loss": 0.4521563947200775,
      "step": 2501
    },
    {
      "epoch": 2.3648393194706996,
      "grad_norm": 0.4462793663737836,
      "learning_rate": 1.1798431327186815e-05,
      "loss": 0.44949212670326233,
      "step": 2502
    },
    {
      "epoch": 2.3657844990548202,
      "grad_norm": 0.46393428015820204,
      "learning_rate": 1.1764838640628011e-05,
      "loss": 0.5045512914657593,
      "step": 2503
    },
    {
      "epoch": 2.3667296786389413,
      "grad_norm": 0.4696434769706429,
      "learning_rate": 1.1731287467440184e-05,
      "loss": 0.4626625180244446,
      "step": 2504
    },
    {
      "epoch": 2.3676748582230625,
      "grad_norm": 0.4358216203745698,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 0.4464097321033478,
      "step": 2505
    },
    {
      "epoch": 2.3686200378071836,
      "grad_norm": 0.46383313329746273,
      "learning_rate": 1.1664309806843466e-05,
      "loss": 0.44989997148513794,
      "step": 2506
    },
    {
      "epoch": 2.369565217391304,
      "grad_norm": 0.42775118190662714,
      "learning_rate": 1.1630883392154785e-05,
      "loss": 0.4156571626663208,
      "step": 2507
    },
    {
      "epoch": 2.3705103969754253,
      "grad_norm": 0.4575513945138914,
      "learning_rate": 1.159749863627741e-05,
      "loss": 0.40896475315093994,
      "step": 2508
    },
    {
      "epoch": 2.3714555765595464,
      "grad_norm": 0.4318290018581302,
      "learning_rate": 1.1564155575458413e-05,
      "loss": 0.44691967964172363,
      "step": 2509
    },
    {
      "epoch": 2.3724007561436675,
      "grad_norm": 0.4499667694858054,
      "learning_rate": 1.1530854245899659e-05,
      "loss": 0.41095632314682007,
      "step": 2510
    },
    {
      "epoch": 2.373345935727788,
      "grad_norm": 0.5083914770505531,
      "learning_rate": 1.1497594683757645e-05,
      "loss": 0.44392499327659607,
      "step": 2511
    },
    {
      "epoch": 2.3742911153119093,
      "grad_norm": 0.45299495006096097,
      "learning_rate": 1.1464376925143528e-05,
      "loss": 0.45617222785949707,
      "step": 2512
    },
    {
      "epoch": 2.3752362948960304,
      "grad_norm": 0.3989380529231217,
      "learning_rate": 1.1431201006123138e-05,
      "loss": 0.4266203045845032,
      "step": 2513
    },
    {
      "epoch": 2.376181474480151,
      "grad_norm": 0.4564953155177954,
      "learning_rate": 1.1398066962716803e-05,
      "loss": 0.42492350935935974,
      "step": 2514
    },
    {
      "epoch": 2.377126654064272,
      "grad_norm": 0.40288922744312794,
      "learning_rate": 1.1364974830899439e-05,
      "loss": 0.397478848695755,
      "step": 2515
    },
    {
      "epoch": 2.3780718336483933,
      "grad_norm": 0.4844894599413972,
      "learning_rate": 1.1331924646600416e-05,
      "loss": 0.45784473419189453,
      "step": 2516
    },
    {
      "epoch": 2.3790170132325144,
      "grad_norm": 0.48507467705035945,
      "learning_rate": 1.1298916445703567e-05,
      "loss": 0.47298771142959595,
      "step": 2517
    },
    {
      "epoch": 2.379962192816635,
      "grad_norm": 0.45223692621081163,
      "learning_rate": 1.126595026404717e-05,
      "loss": 0.45207783579826355,
      "step": 2518
    },
    {
      "epoch": 2.380907372400756,
      "grad_norm": 0.48873266529603143,
      "learning_rate": 1.1233026137423857e-05,
      "loss": 0.4741474390029907,
      "step": 2519
    },
    {
      "epoch": 2.381852551984877,
      "grad_norm": 0.4880128533200077,
      "learning_rate": 1.1200144101580635e-05,
      "loss": 0.45112478733062744,
      "step": 2520
    },
    {
      "epoch": 2.3827977315689983,
      "grad_norm": 0.4548301969541449,
      "learning_rate": 1.1167304192218737e-05,
      "loss": 0.4040190279483795,
      "step": 2521
    },
    {
      "epoch": 2.383742911153119,
      "grad_norm": 0.4206777926338536,
      "learning_rate": 1.1134506444993736e-05,
      "loss": 0.40665534138679504,
      "step": 2522
    },
    {
      "epoch": 2.38468809073724,
      "grad_norm": 0.3924768551719016,
      "learning_rate": 1.1101750895515378e-05,
      "loss": 0.3692183494567871,
      "step": 2523
    },
    {
      "epoch": 2.385633270321361,
      "grad_norm": 0.6002882133100554,
      "learning_rate": 1.1069037579347613e-05,
      "loss": 0.4863740801811218,
      "step": 2524
    },
    {
      "epoch": 2.386578449905482,
      "grad_norm": 0.42731334046704955,
      "learning_rate": 1.1036366532008552e-05,
      "loss": 0.4032970666885376,
      "step": 2525
    },
    {
      "epoch": 2.387523629489603,
      "grad_norm": 0.4315655139899895,
      "learning_rate": 1.1003737788970364e-05,
      "loss": 0.44499537348747253,
      "step": 2526
    },
    {
      "epoch": 2.388468809073724,
      "grad_norm": 0.4796031056479257,
      "learning_rate": 1.0971151385659356e-05,
      "loss": 0.462598979473114,
      "step": 2527
    },
    {
      "epoch": 2.389413988657845,
      "grad_norm": 0.4458377774198329,
      "learning_rate": 1.0938607357455788e-05,
      "loss": 0.4133928418159485,
      "step": 2528
    },
    {
      "epoch": 2.390359168241966,
      "grad_norm": 0.46269245570771395,
      "learning_rate": 1.0906105739693979e-05,
      "loss": 0.49965405464172363,
      "step": 2529
    },
    {
      "epoch": 2.391304347826087,
      "grad_norm": 0.4602308562593866,
      "learning_rate": 1.0873646567662165e-05,
      "loss": 0.43670451641082764,
      "step": 2530
    },
    {
      "epoch": 2.392249527410208,
      "grad_norm": 0.4991823549799953,
      "learning_rate": 1.08412298766025e-05,
      "loss": 0.46852701902389526,
      "step": 2531
    },
    {
      "epoch": 2.393194706994329,
      "grad_norm": 0.42429557112917565,
      "learning_rate": 1.0808855701711034e-05,
      "loss": 0.40196526050567627,
      "step": 2532
    },
    {
      "epoch": 2.39413988657845,
      "grad_norm": 0.5252227474161788,
      "learning_rate": 1.0776524078137639e-05,
      "loss": 0.5169253349304199,
      "step": 2533
    },
    {
      "epoch": 2.395085066162571,
      "grad_norm": 0.44932786133913344,
      "learning_rate": 1.0744235040985951e-05,
      "loss": 0.3844372630119324,
      "step": 2534
    },
    {
      "epoch": 2.396030245746692,
      "grad_norm": 0.4898811254051634,
      "learning_rate": 1.0711988625313468e-05,
      "loss": 0.4286476671695709,
      "step": 2535
    },
    {
      "epoch": 2.3969754253308126,
      "grad_norm": 0.44228286222260305,
      "learning_rate": 1.067978486613131e-05,
      "loss": 0.4120205044746399,
      "step": 2536
    },
    {
      "epoch": 2.3979206049149338,
      "grad_norm": 0.4277399618797466,
      "learning_rate": 1.0647623798404355e-05,
      "loss": 0.4504888653755188,
      "step": 2537
    },
    {
      "epoch": 2.398865784499055,
      "grad_norm": 0.4560315338304475,
      "learning_rate": 1.0615505457051068e-05,
      "loss": 0.4553142488002777,
      "step": 2538
    },
    {
      "epoch": 2.399810964083176,
      "grad_norm": 0.49411580929166465,
      "learning_rate": 1.0583429876943585e-05,
      "loss": 0.4516581892967224,
      "step": 2539
    },
    {
      "epoch": 2.4007561436672966,
      "grad_norm": 0.4483994521918096,
      "learning_rate": 1.055139709290755e-05,
      "loss": 0.4017055034637451,
      "step": 2540
    },
    {
      "epoch": 2.4017013232514177,
      "grad_norm": 0.4360720676581422,
      "learning_rate": 1.0519407139722203e-05,
      "loss": 0.39867103099823,
      "step": 2541
    },
    {
      "epoch": 2.402646502835539,
      "grad_norm": 0.44598903853061145,
      "learning_rate": 1.0487460052120263e-05,
      "loss": 0.4603021740913391,
      "step": 2542
    },
    {
      "epoch": 2.40359168241966,
      "grad_norm": 0.42054095133857455,
      "learning_rate": 1.045555586478787e-05,
      "loss": 0.437349796295166,
      "step": 2543
    },
    {
      "epoch": 2.4045368620037806,
      "grad_norm": 0.4143387245751995,
      "learning_rate": 1.0423694612364643e-05,
      "loss": 0.4156915843486786,
      "step": 2544
    },
    {
      "epoch": 2.4054820415879017,
      "grad_norm": 0.48673775989718104,
      "learning_rate": 1.0391876329443533e-05,
      "loss": 0.48974254727363586,
      "step": 2545
    },
    {
      "epoch": 2.406427221172023,
      "grad_norm": 0.47039693180773356,
      "learning_rate": 1.0360101050570874e-05,
      "loss": 0.46263766288757324,
      "step": 2546
    },
    {
      "epoch": 2.4073724007561434,
      "grad_norm": 0.3905236924268833,
      "learning_rate": 1.0328368810246313e-05,
      "loss": 0.40238115191459656,
      "step": 2547
    },
    {
      "epoch": 2.4083175803402646,
      "grad_norm": 0.4763326783655039,
      "learning_rate": 1.0296679642922718e-05,
      "loss": 0.4557226896286011,
      "step": 2548
    },
    {
      "epoch": 2.4092627599243857,
      "grad_norm": 0.46467069675264844,
      "learning_rate": 1.0265033583006257e-05,
      "loss": 0.4383893609046936,
      "step": 2549
    },
    {
      "epoch": 2.4102079395085068,
      "grad_norm": 0.5156277863758657,
      "learning_rate": 1.0233430664856236e-05,
      "loss": 0.4730386435985565,
      "step": 2550
    },
    {
      "epoch": 2.4111531190926274,
      "grad_norm": 0.44243583566339134,
      "learning_rate": 1.0201870922785156e-05,
      "loss": 0.4468458294868469,
      "step": 2551
    },
    {
      "epoch": 2.4120982986767485,
      "grad_norm": 0.47225504616588004,
      "learning_rate": 1.0170354391058645e-05,
      "loss": 0.48150384426116943,
      "step": 2552
    },
    {
      "epoch": 2.4130434782608696,
      "grad_norm": 0.41400761915218287,
      "learning_rate": 1.0138881103895381e-05,
      "loss": 0.41469883918762207,
      "step": 2553
    },
    {
      "epoch": 2.4139886578449907,
      "grad_norm": 0.4188200384768831,
      "learning_rate": 1.010745109546713e-05,
      "loss": 0.3869956135749817,
      "step": 2554
    },
    {
      "epoch": 2.4149338374291114,
      "grad_norm": 0.4160425198307869,
      "learning_rate": 1.0076064399898627e-05,
      "loss": 0.4127315580844879,
      "step": 2555
    },
    {
      "epoch": 2.4158790170132325,
      "grad_norm": 0.4474506367672899,
      "learning_rate": 1.004472105126762e-05,
      "loss": 0.42889004945755005,
      "step": 2556
    },
    {
      "epoch": 2.4168241965973536,
      "grad_norm": 0.5176777172613548,
      "learning_rate": 1.001342108360478e-05,
      "loss": 0.5232925415039062,
      "step": 2557
    },
    {
      "epoch": 2.4177693761814743,
      "grad_norm": 0.47847857065466504,
      "learning_rate": 9.98216453089365e-06,
      "loss": 0.48676374554634094,
      "step": 2558
    },
    {
      "epoch": 2.4187145557655954,
      "grad_norm": 0.4395026221631063,
      "learning_rate": 9.950951427070693e-06,
      "loss": 0.4198583960533142,
      "step": 2559
    },
    {
      "epoch": 2.4196597353497165,
      "grad_norm": 0.4316300778575391,
      "learning_rate": 9.919781806025135e-06,
      "loss": 0.41687077283859253,
      "step": 2560
    },
    {
      "epoch": 2.4206049149338376,
      "grad_norm": 0.45304900776892104,
      "learning_rate": 9.888655701599043e-06,
      "loss": 0.43156060576438904,
      "step": 2561
    },
    {
      "epoch": 2.421550094517958,
      "grad_norm": 0.3611182986502264,
      "learning_rate": 9.85757314758719e-06,
      "loss": 0.3627498149871826,
      "step": 2562
    },
    {
      "epoch": 2.4224952741020793,
      "grad_norm": 0.43010123852603455,
      "learning_rate": 9.826534177737107e-06,
      "loss": 0.4558180868625641,
      "step": 2563
    },
    {
      "epoch": 2.4234404536862004,
      "grad_norm": 0.4362446124384006,
      "learning_rate": 9.79553882574899e-06,
      "loss": 0.3797916769981384,
      "step": 2564
    },
    {
      "epoch": 2.4243856332703215,
      "grad_norm": 0.4358970697060926,
      "learning_rate": 9.764587125275654e-06,
      "loss": 0.41577786207199097,
      "step": 2565
    },
    {
      "epoch": 2.425330812854442,
      "grad_norm": 0.4113387996919505,
      "learning_rate": 9.733679109922566e-06,
      "loss": 0.42365044355392456,
      "step": 2566
    },
    {
      "epoch": 2.4262759924385633,
      "grad_norm": 0.4136719370844545,
      "learning_rate": 9.702814813247701e-06,
      "loss": 0.41137993335723877,
      "step": 2567
    },
    {
      "epoch": 2.4272211720226844,
      "grad_norm": 0.42816784331340374,
      "learning_rate": 9.67199426876162e-06,
      "loss": 0.4439784288406372,
      "step": 2568
    },
    {
      "epoch": 2.428166351606805,
      "grad_norm": 0.4611621811198385,
      "learning_rate": 9.641217509927375e-06,
      "loss": 0.45181918144226074,
      "step": 2569
    },
    {
      "epoch": 2.429111531190926,
      "grad_norm": 0.42952147629082105,
      "learning_rate": 9.610484570160444e-06,
      "loss": 0.4164160490036011,
      "step": 2570
    },
    {
      "epoch": 2.4300567107750473,
      "grad_norm": 0.3965064861380075,
      "learning_rate": 9.579795482828769e-06,
      "loss": 0.43270406126976013,
      "step": 2571
    },
    {
      "epoch": 2.4310018903591684,
      "grad_norm": 0.46225835156580614,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.4643319249153137,
      "step": 2572
    },
    {
      "epoch": 2.4319470699432895,
      "grad_norm": 0.46583667897996606,
      "learning_rate": 9.518548998704718e-06,
      "loss": 0.433761864900589,
      "step": 2573
    },
    {
      "epoch": 2.43289224952741,
      "grad_norm": 0.4192963034028492,
      "learning_rate": 9.487991668410013e-06,
      "loss": 0.44192877411842346,
      "step": 2574
    },
    {
      "epoch": 2.4338374291115312,
      "grad_norm": 0.48089004722191353,
      "learning_rate": 9.45747832354575e-06,
      "loss": 0.48351460695266724,
      "step": 2575
    },
    {
      "epoch": 2.4347826086956523,
      "grad_norm": 0.4493981880063315,
      "learning_rate": 9.427008997241454e-06,
      "loss": 0.4567515254020691,
      "step": 2576
    },
    {
      "epoch": 2.435727788279773,
      "grad_norm": 0.44608235621395753,
      "learning_rate": 9.396583722578806e-06,
      "loss": 0.39823341369628906,
      "step": 2577
    },
    {
      "epoch": 2.436672967863894,
      "grad_norm": 0.4379118511297267,
      "learning_rate": 9.366202532591717e-06,
      "loss": 0.4236559271812439,
      "step": 2578
    },
    {
      "epoch": 2.437618147448015,
      "grad_norm": 0.4481778053645614,
      "learning_rate": 9.33586546026617e-06,
      "loss": 0.40283092856407166,
      "step": 2579
    },
    {
      "epoch": 2.4385633270321363,
      "grad_norm": 0.4951799443634283,
      "learning_rate": 9.305572538540296e-06,
      "loss": 0.46971654891967773,
      "step": 2580
    },
    {
      "epoch": 2.439508506616257,
      "grad_norm": 0.44173191956520125,
      "learning_rate": 9.275323800304286e-06,
      "loss": 0.43956634402275085,
      "step": 2581
    },
    {
      "epoch": 2.440453686200378,
      "grad_norm": 0.4625311966958024,
      "learning_rate": 9.245119278400322e-06,
      "loss": 0.44087350368499756,
      "step": 2582
    },
    {
      "epoch": 2.441398865784499,
      "grad_norm": 0.47198929849037147,
      "learning_rate": 9.21495900562263e-06,
      "loss": 0.4982827305793762,
      "step": 2583
    },
    {
      "epoch": 2.4423440453686203,
      "grad_norm": 0.43491132266342164,
      "learning_rate": 9.184843014717336e-06,
      "loss": 0.44456422328948975,
      "step": 2584
    },
    {
      "epoch": 2.443289224952741,
      "grad_norm": 0.42438192259773705,
      "learning_rate": 9.154771338382545e-06,
      "loss": 0.39886099100112915,
      "step": 2585
    },
    {
      "epoch": 2.444234404536862,
      "grad_norm": 0.44562268560180895,
      "learning_rate": 9.124744009268215e-06,
      "loss": 0.439552903175354,
      "step": 2586
    },
    {
      "epoch": 2.445179584120983,
      "grad_norm": 0.42123397081362546,
      "learning_rate": 9.094761059976153e-06,
      "loss": 0.37529420852661133,
      "step": 2587
    },
    {
      "epoch": 2.446124763705104,
      "grad_norm": 0.44699183213207444,
      "learning_rate": 9.064822523060001e-06,
      "loss": 0.4363275170326233,
      "step": 2588
    },
    {
      "epoch": 2.447069943289225,
      "grad_norm": 0.3971730081011595,
      "learning_rate": 9.034928431025152e-06,
      "loss": 0.40923503041267395,
      "step": 2589
    },
    {
      "epoch": 2.448015122873346,
      "grad_norm": 0.43105971412727545,
      "learning_rate": 9.005078816328771e-06,
      "loss": 0.3967304825782776,
      "step": 2590
    },
    {
      "epoch": 2.448960302457467,
      "grad_norm": 0.4006478231750413,
      "learning_rate": 8.975273711379728e-06,
      "loss": 0.4045651853084564,
      "step": 2591
    },
    {
      "epoch": 2.4499054820415878,
      "grad_norm": 0.4109848159464914,
      "learning_rate": 8.945513148538549e-06,
      "loss": 0.4154146909713745,
      "step": 2592
    },
    {
      "epoch": 2.450850661625709,
      "grad_norm": 0.45632419387975154,
      "learning_rate": 8.915797160117423e-06,
      "loss": 0.43437790870666504,
      "step": 2593
    },
    {
      "epoch": 2.45179584120983,
      "grad_norm": 0.4626548004222342,
      "learning_rate": 8.886125778380105e-06,
      "loss": 0.43379783630371094,
      "step": 2594
    },
    {
      "epoch": 2.452741020793951,
      "grad_norm": 0.4866948769141694,
      "learning_rate": 8.856499035541971e-06,
      "loss": 0.39678800106048584,
      "step": 2595
    },
    {
      "epoch": 2.4536862003780717,
      "grad_norm": 0.4808733887690744,
      "learning_rate": 8.826916963769889e-06,
      "loss": 0.4438195526599884,
      "step": 2596
    },
    {
      "epoch": 2.454631379962193,
      "grad_norm": 0.4891722042054317,
      "learning_rate": 8.797379595182258e-06,
      "loss": 0.4386444091796875,
      "step": 2597
    },
    {
      "epoch": 2.455576559546314,
      "grad_norm": 0.40110085616624935,
      "learning_rate": 8.767886961848915e-06,
      "loss": 0.4146635830402374,
      "step": 2598
    },
    {
      "epoch": 2.4565217391304346,
      "grad_norm": 0.45200229254390073,
      "learning_rate": 8.738439095791123e-06,
      "loss": 0.44533270597457886,
      "step": 2599
    },
    {
      "epoch": 2.4574669187145557,
      "grad_norm": 0.47313066820986416,
      "learning_rate": 8.70903602898157e-06,
      "loss": 0.49036669731140137,
      "step": 2600
    },
    {
      "epoch": 2.458412098298677,
      "grad_norm": 0.4737756642933239,
      "learning_rate": 8.679677793344255e-06,
      "loss": 0.4623938202857971,
      "step": 2601
    },
    {
      "epoch": 2.459357277882798,
      "grad_norm": 0.4077691787990242,
      "learning_rate": 8.650364420754581e-06,
      "loss": 0.37974125146865845,
      "step": 2602
    },
    {
      "epoch": 2.4603024574669186,
      "grad_norm": 0.4707910597657286,
      "learning_rate": 8.621095943039164e-06,
      "loss": 0.4679224491119385,
      "step": 2603
    },
    {
      "epoch": 2.4612476370510397,
      "grad_norm": 0.5046383881314049,
      "learning_rate": 8.591872391975897e-06,
      "loss": 0.47223836183547974,
      "step": 2604
    },
    {
      "epoch": 2.4621928166351608,
      "grad_norm": 0.44720030830241014,
      "learning_rate": 8.562693799293931e-06,
      "loss": 0.46277138590812683,
      "step": 2605
    },
    {
      "epoch": 2.463137996219282,
      "grad_norm": 0.48118051731057027,
      "learning_rate": 8.53356019667354e-06,
      "loss": 0.48143911361694336,
      "step": 2606
    },
    {
      "epoch": 2.4640831758034025,
      "grad_norm": 0.4453829806353652,
      "learning_rate": 8.504471615746207e-06,
      "loss": 0.4306177794933319,
      "step": 2607
    },
    {
      "epoch": 2.4650283553875236,
      "grad_norm": 0.47322878658002027,
      "learning_rate": 8.475428088094517e-06,
      "loss": 0.47815778851509094,
      "step": 2608
    },
    {
      "epoch": 2.4659735349716447,
      "grad_norm": 0.4678009891674314,
      "learning_rate": 8.446429645252119e-06,
      "loss": 0.4090172052383423,
      "step": 2609
    },
    {
      "epoch": 2.4669187145557654,
      "grad_norm": 0.4154172148876046,
      "learning_rate": 8.417476318703744e-06,
      "loss": 0.3855113387107849,
      "step": 2610
    },
    {
      "epoch": 2.4678638941398865,
      "grad_norm": 0.41893618403746335,
      "learning_rate": 8.3885681398851e-06,
      "loss": 0.42560142278671265,
      "step": 2611
    },
    {
      "epoch": 2.4688090737240076,
      "grad_norm": 0.45512718641062977,
      "learning_rate": 8.35970514018291e-06,
      "loss": 0.4255632758140564,
      "step": 2612
    },
    {
      "epoch": 2.4697542533081287,
      "grad_norm": 0.4095880629026154,
      "learning_rate": 8.330887350934835e-06,
      "loss": 0.38368529081344604,
      "step": 2613
    },
    {
      "epoch": 2.4706994328922494,
      "grad_norm": 0.3952665957784336,
      "learning_rate": 8.302114803429451e-06,
      "loss": 0.3604421019554138,
      "step": 2614
    },
    {
      "epoch": 2.4716446124763705,
      "grad_norm": 0.4166012186309722,
      "learning_rate": 8.2733875289062e-06,
      "loss": 0.43247365951538086,
      "step": 2615
    },
    {
      "epoch": 2.4725897920604916,
      "grad_norm": 0.4318837020887975,
      "learning_rate": 8.244705558555365e-06,
      "loss": 0.43523842096328735,
      "step": 2616
    },
    {
      "epoch": 2.4735349716446127,
      "grad_norm": 0.4315292549705706,
      "learning_rate": 8.216068923518072e-06,
      "loss": 0.46313077211380005,
      "step": 2617
    },
    {
      "epoch": 2.4744801512287333,
      "grad_norm": 0.45423143804917665,
      "learning_rate": 8.187477654886206e-06,
      "loss": 0.4348974823951721,
      "step": 2618
    },
    {
      "epoch": 2.4754253308128544,
      "grad_norm": 0.3785517205291096,
      "learning_rate": 8.158931783702395e-06,
      "loss": 0.3561185598373413,
      "step": 2619
    },
    {
      "epoch": 2.4763705103969755,
      "grad_norm": 0.47240683114223414,
      "learning_rate": 8.130431340959981e-06,
      "loss": 0.4659004807472229,
      "step": 2620
    },
    {
      "epoch": 2.477315689981096,
      "grad_norm": 0.4669319832570199,
      "learning_rate": 8.101976357602959e-06,
      "loss": 0.4646679162979126,
      "step": 2621
    },
    {
      "epoch": 2.4782608695652173,
      "grad_norm": 0.4663081028368912,
      "learning_rate": 8.073566864526023e-06,
      "loss": 0.4107164144515991,
      "step": 2622
    },
    {
      "epoch": 2.4792060491493384,
      "grad_norm": 0.47073555333053824,
      "learning_rate": 8.045202892574395e-06,
      "loss": 0.4389885663986206,
      "step": 2623
    },
    {
      "epoch": 2.4801512287334595,
      "grad_norm": 0.4735209185765353,
      "learning_rate": 8.016884472543968e-06,
      "loss": 0.44780439138412476,
      "step": 2624
    },
    {
      "epoch": 2.48109640831758,
      "grad_norm": 0.45485486427999183,
      "learning_rate": 7.988611635181098e-06,
      "loss": 0.48579350113868713,
      "step": 2625
    },
    {
      "epoch": 2.4820415879017013,
      "grad_norm": 0.47452620732338985,
      "learning_rate": 7.960384411182709e-06,
      "loss": 0.44850462675094604,
      "step": 2626
    },
    {
      "epoch": 2.4829867674858224,
      "grad_norm": 0.506237041053555,
      "learning_rate": 7.93220283119615e-06,
      "loss": 0.5200462341308594,
      "step": 2627
    },
    {
      "epoch": 2.4839319470699435,
      "grad_norm": 0.4217095615545372,
      "learning_rate": 7.904066925819237e-06,
      "loss": 0.44232404232025146,
      "step": 2628
    },
    {
      "epoch": 2.484877126654064,
      "grad_norm": 0.46222863077063703,
      "learning_rate": 7.875976725600193e-06,
      "loss": 0.43318891525268555,
      "step": 2629
    },
    {
      "epoch": 2.4858223062381852,
      "grad_norm": 0.4635537503899768,
      "learning_rate": 7.847932261037627e-06,
      "loss": 0.48837122321128845,
      "step": 2630
    },
    {
      "epoch": 2.4867674858223063,
      "grad_norm": 0.4630264799666236,
      "learning_rate": 7.819933562580494e-06,
      "loss": 0.4493846893310547,
      "step": 2631
    },
    {
      "epoch": 2.487712665406427,
      "grad_norm": 0.4462331796366202,
      "learning_rate": 7.791980660628029e-06,
      "loss": 0.42123717069625854,
      "step": 2632
    },
    {
      "epoch": 2.488657844990548,
      "grad_norm": 0.42528587602354667,
      "learning_rate": 7.764073585529758e-06,
      "loss": 0.41616547107696533,
      "step": 2633
    },
    {
      "epoch": 2.489603024574669,
      "grad_norm": 0.48826360205853336,
      "learning_rate": 7.73621236758546e-06,
      "loss": 0.4244873523712158,
      "step": 2634
    },
    {
      "epoch": 2.4905482041587903,
      "grad_norm": 0.46998863652529277,
      "learning_rate": 7.708397037045129e-06,
      "loss": 0.46491849422454834,
      "step": 2635
    },
    {
      "epoch": 2.491493383742911,
      "grad_norm": 0.46871419246859525,
      "learning_rate": 7.680627624108932e-06,
      "loss": 0.44228434562683105,
      "step": 2636
    },
    {
      "epoch": 2.492438563327032,
      "grad_norm": 0.4665910948232383,
      "learning_rate": 7.65290415892717e-06,
      "loss": 0.44241905212402344,
      "step": 2637
    },
    {
      "epoch": 2.493383742911153,
      "grad_norm": 0.40130830657123995,
      "learning_rate": 7.625226671600255e-06,
      "loss": 0.42087477445602417,
      "step": 2638
    },
    {
      "epoch": 2.4943289224952743,
      "grad_norm": 0.49308972291310205,
      "learning_rate": 7.597595192178702e-06,
      "loss": 0.48422396183013916,
      "step": 2639
    },
    {
      "epoch": 2.495274102079395,
      "grad_norm": 0.46059730224972023,
      "learning_rate": 7.570009750663054e-06,
      "loss": 0.47090959548950195,
      "step": 2640
    },
    {
      "epoch": 2.496219281663516,
      "grad_norm": 0.44119261715578917,
      "learning_rate": 7.542470377003896e-06,
      "loss": 0.46922236680984497,
      "step": 2641
    },
    {
      "epoch": 2.497164461247637,
      "grad_norm": 0.4625531936497539,
      "learning_rate": 7.514977101101745e-06,
      "loss": 0.4429711103439331,
      "step": 2642
    },
    {
      "epoch": 2.498109640831758,
      "grad_norm": 0.4234319768027277,
      "learning_rate": 7.487529952807121e-06,
      "loss": 0.42234528064727783,
      "step": 2643
    },
    {
      "epoch": 2.499054820415879,
      "grad_norm": 0.48659560217227926,
      "learning_rate": 7.460128961920432e-06,
      "loss": 0.45744359493255615,
      "step": 2644
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.39244374561452827,
      "learning_rate": 7.432774158191946e-06,
      "loss": 0.3678399622440338,
      "step": 2645
    },
    {
      "epoch": 2.500945179584121,
      "grad_norm": 0.38907668060262135,
      "learning_rate": 7.405465571321868e-06,
      "loss": 0.4147511124610901,
      "step": 2646
    },
    {
      "epoch": 2.501890359168242,
      "grad_norm": 0.45034322816360267,
      "learning_rate": 7.37820323096014e-06,
      "loss": 0.4430926442146301,
      "step": 2647
    },
    {
      "epoch": 2.502835538752363,
      "grad_norm": 0.4249858042990229,
      "learning_rate": 7.350987166706535e-06,
      "loss": 0.407281756401062,
      "step": 2648
    },
    {
      "epoch": 2.503780718336484,
      "grad_norm": 0.5064860743125797,
      "learning_rate": 7.323817408110567e-06,
      "loss": 0.4693644046783447,
      "step": 2649
    },
    {
      "epoch": 2.504725897920605,
      "grad_norm": 0.43459563038662424,
      "learning_rate": 7.296693984671465e-06,
      "loss": 0.43162742257118225,
      "step": 2650
    },
    {
      "epoch": 2.5056710775047257,
      "grad_norm": 0.4602193881566698,
      "learning_rate": 7.269616925838174e-06,
      "loss": 0.4252548813819885,
      "step": 2651
    },
    {
      "epoch": 2.506616257088847,
      "grad_norm": 0.4345333271388414,
      "learning_rate": 7.242586261009288e-06,
      "loss": 0.4443492591381073,
      "step": 2652
    },
    {
      "epoch": 2.507561436672968,
      "grad_norm": 0.49590055288497653,
      "learning_rate": 7.215602019533041e-06,
      "loss": 0.48875176906585693,
      "step": 2653
    },
    {
      "epoch": 2.5085066162570886,
      "grad_norm": 0.4810413194315104,
      "learning_rate": 7.188664230707231e-06,
      "loss": 0.4404040575027466,
      "step": 2654
    },
    {
      "epoch": 2.5094517958412097,
      "grad_norm": 0.4285583652757214,
      "learning_rate": 7.161772923779258e-06,
      "loss": 0.4388620853424072,
      "step": 2655
    },
    {
      "epoch": 2.510396975425331,
      "grad_norm": 0.5056606068964503,
      "learning_rate": 7.1349281279460165e-06,
      "loss": 0.4910208582878113,
      "step": 2656
    },
    {
      "epoch": 2.511342155009452,
      "grad_norm": 0.49196751834488023,
      "learning_rate": 7.108129872353942e-06,
      "loss": 0.4454636573791504,
      "step": 2657
    },
    {
      "epoch": 2.512287334593573,
      "grad_norm": 0.44662943761037327,
      "learning_rate": 7.08137818609893e-06,
      "loss": 0.45096346735954285,
      "step": 2658
    },
    {
      "epoch": 2.5132325141776937,
      "grad_norm": 0.41734172778253587,
      "learning_rate": 7.054673098226277e-06,
      "loss": 0.4113657474517822,
      "step": 2659
    },
    {
      "epoch": 2.5141776937618148,
      "grad_norm": 0.48197407221023,
      "learning_rate": 7.0280146377307395e-06,
      "loss": 0.4132530689239502,
      "step": 2660
    },
    {
      "epoch": 2.515122873345936,
      "grad_norm": 0.4401028444943267,
      "learning_rate": 7.001402833556408e-06,
      "loss": 0.42114394903182983,
      "step": 2661
    },
    {
      "epoch": 2.5160680529300565,
      "grad_norm": 0.47250545606780653,
      "learning_rate": 6.974837714596733e-06,
      "loss": 0.496654212474823,
      "step": 2662
    },
    {
      "epoch": 2.5170132325141776,
      "grad_norm": 0.40775726665856166,
      "learning_rate": 6.94831930969449e-06,
      "loss": 0.4488027095794678,
      "step": 2663
    },
    {
      "epoch": 2.5179584120982987,
      "grad_norm": 0.40888603033928084,
      "learning_rate": 6.921847647641705e-06,
      "loss": 0.41656219959259033,
      "step": 2664
    },
    {
      "epoch": 2.5189035916824194,
      "grad_norm": 0.41151462827360036,
      "learning_rate": 6.8954227571796815e-06,
      "loss": 0.4025575518608093,
      "step": 2665
    },
    {
      "epoch": 2.5198487712665405,
      "grad_norm": 0.4403670336827633,
      "learning_rate": 6.869044666998925e-06,
      "loss": 0.43734365701675415,
      "step": 2666
    },
    {
      "epoch": 2.5207939508506616,
      "grad_norm": 0.4545312275539096,
      "learning_rate": 6.842713405739121e-06,
      "loss": 0.42452532052993774,
      "step": 2667
    },
    {
      "epoch": 2.5217391304347827,
      "grad_norm": 0.48789606965084154,
      "learning_rate": 6.816429001989133e-06,
      "loss": 0.4417220652103424,
      "step": 2668
    },
    {
      "epoch": 2.522684310018904,
      "grad_norm": 0.4390759419363336,
      "learning_rate": 6.790191484286945e-06,
      "loss": 0.43401265144348145,
      "step": 2669
    },
    {
      "epoch": 2.5236294896030245,
      "grad_norm": 0.47970970579519423,
      "learning_rate": 6.764000881119631e-06,
      "loss": 0.47399312257766724,
      "step": 2670
    },
    {
      "epoch": 2.5245746691871456,
      "grad_norm": 0.40216092659640085,
      "learning_rate": 6.737857220923305e-06,
      "loss": 0.4066655933856964,
      "step": 2671
    },
    {
      "epoch": 2.5255198487712667,
      "grad_norm": 0.5186790538441524,
      "learning_rate": 6.711760532083167e-06,
      "loss": 0.4418320953845978,
      "step": 2672
    },
    {
      "epoch": 2.5264650283553873,
      "grad_norm": 0.41211834593275154,
      "learning_rate": 6.685710842933352e-06,
      "loss": 0.4546205401420593,
      "step": 2673
    },
    {
      "epoch": 2.5274102079395084,
      "grad_norm": 0.4779513574051482,
      "learning_rate": 6.659708181757013e-06,
      "loss": 0.4525682330131531,
      "step": 2674
    },
    {
      "epoch": 2.5283553875236295,
      "grad_norm": 0.43331317249219364,
      "learning_rate": 6.6337525767862505e-06,
      "loss": 0.4140746295452118,
      "step": 2675
    },
    {
      "epoch": 2.52930056710775,
      "grad_norm": 0.42617100745807357,
      "learning_rate": 6.607844056202023e-06,
      "loss": 0.41409894824028015,
      "step": 2676
    },
    {
      "epoch": 2.5302457466918713,
      "grad_norm": 0.4276421314122907,
      "learning_rate": 6.581982648134216e-06,
      "loss": 0.41667455434799194,
      "step": 2677
    },
    {
      "epoch": 2.5311909262759924,
      "grad_norm": 0.4857193609396329,
      "learning_rate": 6.556168380661531e-06,
      "loss": 0.453121155500412,
      "step": 2678
    },
    {
      "epoch": 2.5321361058601135,
      "grad_norm": 0.433123061986026,
      "learning_rate": 6.53040128181151e-06,
      "loss": 0.384280800819397,
      "step": 2679
    },
    {
      "epoch": 2.5330812854442346,
      "grad_norm": 0.4261255016376589,
      "learning_rate": 6.50468137956049e-06,
      "loss": 0.38342708349227905,
      "step": 2680
    },
    {
      "epoch": 2.5340264650283553,
      "grad_norm": 0.45320856415939104,
      "learning_rate": 6.479008701833516e-06,
      "loss": 0.4470060467720032,
      "step": 2681
    },
    {
      "epoch": 2.5349716446124764,
      "grad_norm": 0.5373547661026264,
      "learning_rate": 6.4533832765044215e-06,
      "loss": 0.5268712639808655,
      "step": 2682
    },
    {
      "epoch": 2.5359168241965975,
      "grad_norm": 0.5037108495949054,
      "learning_rate": 6.427805131395681e-06,
      "loss": 0.4441536068916321,
      "step": 2683
    },
    {
      "epoch": 2.536862003780718,
      "grad_norm": 0.45641778279647743,
      "learning_rate": 6.402274294278471e-06,
      "loss": 0.439646452665329,
      "step": 2684
    },
    {
      "epoch": 2.5378071833648392,
      "grad_norm": 0.45698709041900737,
      "learning_rate": 6.376790792872611e-06,
      "loss": 0.4801362156867981,
      "step": 2685
    },
    {
      "epoch": 2.5387523629489603,
      "grad_norm": 0.43507066387673465,
      "learning_rate": 6.351354654846481e-06,
      "loss": 0.44434112310409546,
      "step": 2686
    },
    {
      "epoch": 2.5396975425330814,
      "grad_norm": 0.449125619416677,
      "learning_rate": 6.325965907817089e-06,
      "loss": 0.4738543629646301,
      "step": 2687
    },
    {
      "epoch": 2.5406427221172025,
      "grad_norm": 0.4186341340710809,
      "learning_rate": 6.3006245793499396e-06,
      "loss": 0.4310359060764313,
      "step": 2688
    },
    {
      "epoch": 2.541587901701323,
      "grad_norm": 0.4442257923047555,
      "learning_rate": 6.275330696959108e-06,
      "loss": 0.44933223724365234,
      "step": 2689
    },
    {
      "epoch": 2.5425330812854443,
      "grad_norm": 0.4445868074893592,
      "learning_rate": 6.2500842881071e-06,
      "loss": 0.4426150918006897,
      "step": 2690
    },
    {
      "epoch": 2.5434782608695654,
      "grad_norm": 0.44143400121651927,
      "learning_rate": 6.224885380204909e-06,
      "loss": 0.4261302351951599,
      "step": 2691
    },
    {
      "epoch": 2.544423440453686,
      "grad_norm": 0.46223339488466253,
      "learning_rate": 6.1997340006119675e-06,
      "loss": 0.47669488191604614,
      "step": 2692
    },
    {
      "epoch": 2.545368620037807,
      "grad_norm": 0.4456473931514921,
      "learning_rate": 6.174630176636059e-06,
      "loss": 0.4537107050418854,
      "step": 2693
    },
    {
      "epoch": 2.5463137996219283,
      "grad_norm": 0.4559743256909207,
      "learning_rate": 6.149573935533387e-06,
      "loss": 0.4708673655986786,
      "step": 2694
    },
    {
      "epoch": 2.547258979206049,
      "grad_norm": 0.47681604184536946,
      "learning_rate": 6.124565304508439e-06,
      "loss": 0.45791491866111755,
      "step": 2695
    },
    {
      "epoch": 2.54820415879017,
      "grad_norm": 0.47305837718064453,
      "learning_rate": 6.099604310714058e-06,
      "loss": 0.4677482843399048,
      "step": 2696
    },
    {
      "epoch": 2.549149338374291,
      "grad_norm": 0.4474894738196079,
      "learning_rate": 6.074690981251352e-06,
      "loss": 0.49106845259666443,
      "step": 2697
    },
    {
      "epoch": 2.5500945179584122,
      "grad_norm": 0.44171274498498364,
      "learning_rate": 6.049825343169652e-06,
      "loss": 0.4353407621383667,
      "step": 2698
    },
    {
      "epoch": 2.5510396975425333,
      "grad_norm": 0.4792271270344955,
      "learning_rate": 6.025007423466544e-06,
      "loss": 0.470819890499115,
      "step": 2699
    },
    {
      "epoch": 2.551984877126654,
      "grad_norm": 0.48128861872838863,
      "learning_rate": 6.000237249087776e-06,
      "loss": 0.4682561159133911,
      "step": 2700
    },
    {
      "epoch": 2.552930056710775,
      "grad_norm": 0.4346809421632012,
      "learning_rate": 5.975514846927271e-06,
      "loss": 0.4059310555458069,
      "step": 2701
    },
    {
      "epoch": 2.553875236294896,
      "grad_norm": 0.41382401067681274,
      "learning_rate": 5.950840243827099e-06,
      "loss": 0.4246930480003357,
      "step": 2702
    },
    {
      "epoch": 2.554820415879017,
      "grad_norm": 0.3848420927039842,
      "learning_rate": 5.926213466577396e-06,
      "loss": 0.3856430947780609,
      "step": 2703
    },
    {
      "epoch": 2.555765595463138,
      "grad_norm": 0.4587236086108724,
      "learning_rate": 5.901634541916407e-06,
      "loss": 0.4687715172767639,
      "step": 2704
    },
    {
      "epoch": 2.556710775047259,
      "grad_norm": 0.4735055703987423,
      "learning_rate": 5.877103496530395e-06,
      "loss": 0.4397061765193939,
      "step": 2705
    },
    {
      "epoch": 2.5576559546313797,
      "grad_norm": 0.44093674305011915,
      "learning_rate": 5.852620357053651e-06,
      "loss": 0.44436126947402954,
      "step": 2706
    },
    {
      "epoch": 2.558601134215501,
      "grad_norm": 0.45822598330056835,
      "learning_rate": 5.828185150068472e-06,
      "loss": 0.4681771993637085,
      "step": 2707
    },
    {
      "epoch": 2.559546313799622,
      "grad_norm": 0.46618656086653126,
      "learning_rate": 5.803797902105062e-06,
      "loss": 0.46965163946151733,
      "step": 2708
    },
    {
      "epoch": 2.560491493383743,
      "grad_norm": 0.4449227222840232,
      "learning_rate": 5.779458639641616e-06,
      "loss": 0.41730132699012756,
      "step": 2709
    },
    {
      "epoch": 2.561436672967864,
      "grad_norm": 0.4312713832075517,
      "learning_rate": 5.755167389104166e-06,
      "loss": 0.39455363154411316,
      "step": 2710
    },
    {
      "epoch": 2.562381852551985,
      "grad_norm": 0.513947380403203,
      "learning_rate": 5.730924176866681e-06,
      "loss": 0.4212377667427063,
      "step": 2711
    },
    {
      "epoch": 2.563327032136106,
      "grad_norm": 0.4419174985928206,
      "learning_rate": 5.706729029250901e-06,
      "loss": 0.4226003587245941,
      "step": 2712
    },
    {
      "epoch": 2.564272211720227,
      "grad_norm": 0.4971959766244271,
      "learning_rate": 5.682581972526463e-06,
      "loss": 0.5147312879562378,
      "step": 2713
    },
    {
      "epoch": 2.5652173913043477,
      "grad_norm": 0.44126326880389605,
      "learning_rate": 5.658483032910728e-06,
      "loss": 0.4536789655685425,
      "step": 2714
    },
    {
      "epoch": 2.5661625708884688,
      "grad_norm": 0.4261508590351866,
      "learning_rate": 5.634432236568815e-06,
      "loss": 0.3790372610092163,
      "step": 2715
    },
    {
      "epoch": 2.56710775047259,
      "grad_norm": 0.4369194917109048,
      "learning_rate": 5.610429609613616e-06,
      "loss": 0.41400840878486633,
      "step": 2716
    },
    {
      "epoch": 2.5680529300567105,
      "grad_norm": 0.40417914529154936,
      "learning_rate": 5.586475178105677e-06,
      "loss": 0.36876100301742554,
      "step": 2717
    },
    {
      "epoch": 2.5689981096408316,
      "grad_norm": 0.47194767061957654,
      "learning_rate": 5.5625689680532445e-06,
      "loss": 0.4095423221588135,
      "step": 2718
    },
    {
      "epoch": 2.5699432892249527,
      "grad_norm": 0.43872464493472974,
      "learning_rate": 5.538711005412211e-06,
      "loss": 0.45714226365089417,
      "step": 2719
    },
    {
      "epoch": 2.570888468809074,
      "grad_norm": 0.4842499410776815,
      "learning_rate": 5.514901316086057e-06,
      "loss": 0.4609593152999878,
      "step": 2720
    },
    {
      "epoch": 2.571833648393195,
      "grad_norm": 0.4107886684934995,
      "learning_rate": 5.491139925925886e-06,
      "loss": 0.4106951057910919,
      "step": 2721
    },
    {
      "epoch": 2.5727788279773156,
      "grad_norm": 0.4274747287702828,
      "learning_rate": 5.467426860730335e-06,
      "loss": 0.4489469826221466,
      "step": 2722
    },
    {
      "epoch": 2.5737240075614367,
      "grad_norm": 0.42421848393790657,
      "learning_rate": 5.4437621462455735e-06,
      "loss": 0.41439759731292725,
      "step": 2723
    },
    {
      "epoch": 2.574669187145558,
      "grad_norm": 0.44077517199684885,
      "learning_rate": 5.420145808165311e-06,
      "loss": 0.44382625818252563,
      "step": 2724
    },
    {
      "epoch": 2.5756143667296785,
      "grad_norm": 0.4325439838888057,
      "learning_rate": 5.3965778721306755e-06,
      "loss": 0.42274853587150574,
      "step": 2725
    },
    {
      "epoch": 2.5765595463137996,
      "grad_norm": 0.46177987713368096,
      "learning_rate": 5.373058363730293e-06,
      "loss": 0.43304091691970825,
      "step": 2726
    },
    {
      "epoch": 2.5775047258979207,
      "grad_norm": 0.48119937767268833,
      "learning_rate": 5.349587308500176e-06,
      "loss": 0.46720239520072937,
      "step": 2727
    },
    {
      "epoch": 2.5784499054820413,
      "grad_norm": 0.432469295046283,
      "learning_rate": 5.3261647319237496e-06,
      "loss": 0.4199672341346741,
      "step": 2728
    },
    {
      "epoch": 2.5793950850661624,
      "grad_norm": 0.4445750300421496,
      "learning_rate": 5.302790659431789e-06,
      "loss": 0.4122436046600342,
      "step": 2729
    },
    {
      "epoch": 2.5803402646502835,
      "grad_norm": 0.458776307686498,
      "learning_rate": 5.279465116402438e-06,
      "loss": 0.45329684019088745,
      "step": 2730
    },
    {
      "epoch": 2.5812854442344046,
      "grad_norm": 0.46619317408321814,
      "learning_rate": 5.256188128161116e-06,
      "loss": 0.47479647397994995,
      "step": 2731
    },
    {
      "epoch": 2.5822306238185257,
      "grad_norm": 0.40544269609042377,
      "learning_rate": 5.2329597199805124e-06,
      "loss": 0.36445462703704834,
      "step": 2732
    },
    {
      "epoch": 2.5831758034026464,
      "grad_norm": 0.37922790073971274,
      "learning_rate": 5.209779917080632e-06,
      "loss": 0.34290075302124023,
      "step": 2733
    },
    {
      "epoch": 2.5841209829867675,
      "grad_norm": 0.42192838705033237,
      "learning_rate": 5.186648744628636e-06,
      "loss": 0.4211428761482239,
      "step": 2734
    },
    {
      "epoch": 2.5850661625708886,
      "grad_norm": 0.4811827757870293,
      "learning_rate": 5.163566227738936e-06,
      "loss": 0.4754933714866638,
      "step": 2735
    },
    {
      "epoch": 2.5860113421550093,
      "grad_norm": 0.41591946608825975,
      "learning_rate": 5.140532391473107e-06,
      "loss": 0.4169895350933075,
      "step": 2736
    },
    {
      "epoch": 2.5869565217391304,
      "grad_norm": 0.4460409608476691,
      "learning_rate": 5.117547260839845e-06,
      "loss": 0.45776432752609253,
      "step": 2737
    },
    {
      "epoch": 2.5879017013232515,
      "grad_norm": 0.4661492141782366,
      "learning_rate": 5.094610860795002e-06,
      "loss": 0.4346708655357361,
      "step": 2738
    },
    {
      "epoch": 2.588846880907372,
      "grad_norm": 0.46401550687728765,
      "learning_rate": 5.071723216241475e-06,
      "loss": 0.43988537788391113,
      "step": 2739
    },
    {
      "epoch": 2.5897920604914932,
      "grad_norm": 0.4313689584235239,
      "learning_rate": 5.048884352029271e-06,
      "loss": 0.3782297968864441,
      "step": 2740
    },
    {
      "epoch": 2.5907372400756143,
      "grad_norm": 0.4393154544939062,
      "learning_rate": 5.0260942929554185e-06,
      "loss": 0.462138295173645,
      "step": 2741
    },
    {
      "epoch": 2.5916824196597354,
      "grad_norm": 0.4788859696970755,
      "learning_rate": 5.003353063763932e-06,
      "loss": 0.45805612206459045,
      "step": 2742
    },
    {
      "epoch": 2.5926275992438566,
      "grad_norm": 0.42292894236948714,
      "learning_rate": 4.980660689145855e-06,
      "loss": 0.43052878975868225,
      "step": 2743
    },
    {
      "epoch": 2.593572778827977,
      "grad_norm": 0.4742941193663325,
      "learning_rate": 4.958017193739139e-06,
      "loss": 0.43938717246055603,
      "step": 2744
    },
    {
      "epoch": 2.5945179584120983,
      "grad_norm": 0.39937555693435395,
      "learning_rate": 4.935422602128697e-06,
      "loss": 0.4415702223777771,
      "step": 2745
    },
    {
      "epoch": 2.5954631379962194,
      "grad_norm": 0.4519518864893613,
      "learning_rate": 4.912876938846344e-06,
      "loss": 0.4390557110309601,
      "step": 2746
    },
    {
      "epoch": 2.59640831758034,
      "grad_norm": 0.4319365013289578,
      "learning_rate": 4.890380228370778e-06,
      "loss": 0.39091792702674866,
      "step": 2747
    },
    {
      "epoch": 2.597353497164461,
      "grad_norm": 0.3948547728077508,
      "learning_rate": 4.867932495127508e-06,
      "loss": 0.37681078910827637,
      "step": 2748
    },
    {
      "epoch": 2.5982986767485823,
      "grad_norm": 0.40169761577704255,
      "learning_rate": 4.845533763488902e-06,
      "loss": 0.38256704807281494,
      "step": 2749
    },
    {
      "epoch": 2.599243856332703,
      "grad_norm": 0.42726930368670735,
      "learning_rate": 4.823184057774116e-06,
      "loss": 0.38020384311676025,
      "step": 2750
    },
    {
      "epoch": 2.600189035916824,
      "grad_norm": 0.41896275202975336,
      "learning_rate": 4.800883402249079e-06,
      "loss": 0.4404976963996887,
      "step": 2751
    },
    {
      "epoch": 2.601134215500945,
      "grad_norm": 0.43909787135639655,
      "learning_rate": 4.778631821126473e-06,
      "loss": 0.4429529011249542,
      "step": 2752
    },
    {
      "epoch": 2.6020793950850662,
      "grad_norm": 0.4293556467175474,
      "learning_rate": 4.756429338565671e-06,
      "loss": 0.39113688468933105,
      "step": 2753
    },
    {
      "epoch": 2.6030245746691874,
      "grad_norm": 0.42754083170899987,
      "learning_rate": 4.734275978672753e-06,
      "loss": 0.4213194251060486,
      "step": 2754
    },
    {
      "epoch": 2.603969754253308,
      "grad_norm": 0.4453332501370138,
      "learning_rate": 4.712171765500484e-06,
      "loss": 0.4348399341106415,
      "step": 2755
    },
    {
      "epoch": 2.604914933837429,
      "grad_norm": 0.4134779528555368,
      "learning_rate": 4.6901167230482216e-06,
      "loss": 0.39428192377090454,
      "step": 2756
    },
    {
      "epoch": 2.60586011342155,
      "grad_norm": 0.43153926705310985,
      "learning_rate": 4.668110875261983e-06,
      "loss": 0.40666037797927856,
      "step": 2757
    },
    {
      "epoch": 2.606805293005671,
      "grad_norm": 0.448249820403538,
      "learning_rate": 4.6461542460343564e-06,
      "loss": 0.453396201133728,
      "step": 2758
    },
    {
      "epoch": 2.607750472589792,
      "grad_norm": 0.4761466921684805,
      "learning_rate": 4.624246859204495e-06,
      "loss": 0.42173081636428833,
      "step": 2759
    },
    {
      "epoch": 2.608695652173913,
      "grad_norm": 0.45294493407643727,
      "learning_rate": 4.602388738558078e-06,
      "loss": 0.45928263664245605,
      "step": 2760
    },
    {
      "epoch": 2.609640831758034,
      "grad_norm": 0.488925845657439,
      "learning_rate": 4.580579907827287e-06,
      "loss": 0.4566939175128937,
      "step": 2761
    },
    {
      "epoch": 2.610586011342155,
      "grad_norm": 0.432976573596524,
      "learning_rate": 4.55882039069081e-06,
      "loss": 0.42092475295066833,
      "step": 2762
    },
    {
      "epoch": 2.611531190926276,
      "grad_norm": 0.479903418256763,
      "learning_rate": 4.537110210773782e-06,
      "loss": 0.4561586380004883,
      "step": 2763
    },
    {
      "epoch": 2.612476370510397,
      "grad_norm": 0.5534987956568482,
      "learning_rate": 4.5154493916477855e-06,
      "loss": 0.48704490065574646,
      "step": 2764
    },
    {
      "epoch": 2.613421550094518,
      "grad_norm": 0.4707788973638857,
      "learning_rate": 4.493837956830788e-06,
      "loss": 0.4318850636482239,
      "step": 2765
    },
    {
      "epoch": 2.614366729678639,
      "grad_norm": 0.44218924288986516,
      "learning_rate": 4.472275929787134e-06,
      "loss": 0.41488927602767944,
      "step": 2766
    },
    {
      "epoch": 2.61531190926276,
      "grad_norm": 0.4929719459192536,
      "learning_rate": 4.450763333927549e-06,
      "loss": 0.49288368225097656,
      "step": 2767
    },
    {
      "epoch": 2.616257088846881,
      "grad_norm": 0.4170183040125721,
      "learning_rate": 4.429300192609076e-06,
      "loss": 0.44385826587677,
      "step": 2768
    },
    {
      "epoch": 2.6172022684310017,
      "grad_norm": 0.4443273354575549,
      "learning_rate": 4.407886529135075e-06,
      "loss": 0.4122145175933838,
      "step": 2769
    },
    {
      "epoch": 2.618147448015123,
      "grad_norm": 0.488261594004584,
      "learning_rate": 4.386522366755169e-06,
      "loss": 0.4807531237602234,
      "step": 2770
    },
    {
      "epoch": 2.619092627599244,
      "grad_norm": 0.4287982121015052,
      "learning_rate": 4.365207728665233e-06,
      "loss": 0.447624146938324,
      "step": 2771
    },
    {
      "epoch": 2.620037807183365,
      "grad_norm": 0.47316512045321263,
      "learning_rate": 4.343942638007392e-06,
      "loss": 0.47162097692489624,
      "step": 2772
    },
    {
      "epoch": 2.620982986767486,
      "grad_norm": 0.424209129787756,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.376830130815506,
      "step": 2773
    },
    {
      "epoch": 2.6219281663516067,
      "grad_norm": 0.5026259609938232,
      "learning_rate": 4.301561191287435e-06,
      "loss": 0.46253857016563416,
      "step": 2774
    },
    {
      "epoch": 2.622873345935728,
      "grad_norm": 0.4856906810947079,
      "learning_rate": 4.280444881240475e-06,
      "loss": 0.427216112613678,
      "step": 2775
    },
    {
      "epoch": 2.623818525519849,
      "grad_norm": 0.4752160377566553,
      "learning_rate": 4.2593782106558675e-06,
      "loss": 0.472520112991333,
      "step": 2776
    },
    {
      "epoch": 2.6247637051039696,
      "grad_norm": 0.4405331716291504,
      "learning_rate": 4.238361202406488e-06,
      "loss": 0.42081713676452637,
      "step": 2777
    },
    {
      "epoch": 2.6257088846880907,
      "grad_norm": 0.39033291590727215,
      "learning_rate": 4.217393879311299e-06,
      "loss": 0.3783055543899536,
      "step": 2778
    },
    {
      "epoch": 2.626654064272212,
      "grad_norm": 0.43325564833202895,
      "learning_rate": 4.1964762641353294e-06,
      "loss": 0.4369395971298218,
      "step": 2779
    },
    {
      "epoch": 2.6275992438563325,
      "grad_norm": 0.4840910256834369,
      "learning_rate": 4.175608379589624e-06,
      "loss": 0.41766059398651123,
      "step": 2780
    },
    {
      "epoch": 2.6285444234404536,
      "grad_norm": 0.493462449418432,
      "learning_rate": 4.154790248331264e-06,
      "loss": 0.4512481391429901,
      "step": 2781
    },
    {
      "epoch": 2.6294896030245747,
      "grad_norm": 0.4316013243695353,
      "learning_rate": 4.134021892963264e-06,
      "loss": 0.37747722864151,
      "step": 2782
    },
    {
      "epoch": 2.630434782608696,
      "grad_norm": 0.4498541195196379,
      "learning_rate": 4.113303336034613e-06,
      "loss": 0.41995301842689514,
      "step": 2783
    },
    {
      "epoch": 2.631379962192817,
      "grad_norm": 0.44645989506158595,
      "learning_rate": 4.092634600040252e-06,
      "loss": 0.44984424114227295,
      "step": 2784
    },
    {
      "epoch": 2.6323251417769375,
      "grad_norm": 0.47853772347404583,
      "learning_rate": 4.072015707421006e-06,
      "loss": 0.45966339111328125,
      "step": 2785
    },
    {
      "epoch": 2.6332703213610587,
      "grad_norm": 0.4849090384369661,
      "learning_rate": 4.051446680563609e-06,
      "loss": 0.44520825147628784,
      "step": 2786
    },
    {
      "epoch": 2.6342155009451798,
      "grad_norm": 0.47070015650136865,
      "learning_rate": 4.030927541800617e-06,
      "loss": 0.4543454051017761,
      "step": 2787
    },
    {
      "epoch": 2.6351606805293004,
      "grad_norm": 0.4774025895055584,
      "learning_rate": 4.010458313410459e-06,
      "loss": 0.48938536643981934,
      "step": 2788
    },
    {
      "epoch": 2.6361058601134215,
      "grad_norm": 0.4545877719283823,
      "learning_rate": 3.990039017617331e-06,
      "loss": 0.4187595248222351,
      "step": 2789
    },
    {
      "epoch": 2.6370510396975426,
      "grad_norm": 0.4644660103788192,
      "learning_rate": 3.969669676591259e-06,
      "loss": 0.4439191520214081,
      "step": 2790
    },
    {
      "epoch": 2.6379962192816633,
      "grad_norm": 0.42298269575051445,
      "learning_rate": 3.9493503124480135e-06,
      "loss": 0.4623205065727234,
      "step": 2791
    },
    {
      "epoch": 2.6389413988657844,
      "grad_norm": 0.48083957803698574,
      "learning_rate": 3.929080947249087e-06,
      "loss": 0.4486094117164612,
      "step": 2792
    },
    {
      "epoch": 2.6398865784499055,
      "grad_norm": 0.4515477053132505,
      "learning_rate": 3.908861603001723e-06,
      "loss": 0.4304249882698059,
      "step": 2793
    },
    {
      "epoch": 2.6408317580340266,
      "grad_norm": 0.45472714415219895,
      "learning_rate": 3.888692301658819e-06,
      "loss": 0.4266114830970764,
      "step": 2794
    },
    {
      "epoch": 2.6417769376181477,
      "grad_norm": 0.3879895433338081,
      "learning_rate": 3.868573065118936e-06,
      "loss": 0.4477335810661316,
      "step": 2795
    },
    {
      "epoch": 2.6427221172022684,
      "grad_norm": 0.3627530303395456,
      "learning_rate": 3.848503915226337e-06,
      "loss": 0.31800973415374756,
      "step": 2796
    },
    {
      "epoch": 2.6436672967863895,
      "grad_norm": 0.4702762821083236,
      "learning_rate": 3.8284848737708325e-06,
      "loss": 0.3975920081138611,
      "step": 2797
    },
    {
      "epoch": 2.6446124763705106,
      "grad_norm": 0.43259167034314444,
      "learning_rate": 3.8085159624878687e-06,
      "loss": 0.3971887230873108,
      "step": 2798
    },
    {
      "epoch": 2.645557655954631,
      "grad_norm": 0.4313046340747982,
      "learning_rate": 3.788597203058458e-06,
      "loss": 0.3890862762928009,
      "step": 2799
    },
    {
      "epoch": 2.6465028355387523,
      "grad_norm": 0.4957426485909259,
      "learning_rate": 3.768728617109135e-06,
      "loss": 0.4713684916496277,
      "step": 2800
    },
    {
      "epoch": 2.6474480151228734,
      "grad_norm": 0.40641024235639994,
      "learning_rate": 3.748910226211999e-06,
      "loss": 0.38663333654403687,
      "step": 2801
    },
    {
      "epoch": 2.648393194706994,
      "grad_norm": 0.47030837134681885,
      "learning_rate": 3.7291420518846266e-06,
      "loss": 0.453487753868103,
      "step": 2802
    },
    {
      "epoch": 2.649338374291115,
      "grad_norm": 0.5076075267241802,
      "learning_rate": 3.7094241155900875e-06,
      "loss": 0.4682665467262268,
      "step": 2803
    },
    {
      "epoch": 2.6502835538752363,
      "grad_norm": 0.5163006908773995,
      "learning_rate": 3.6897564387368746e-06,
      "loss": 0.4792552590370178,
      "step": 2804
    },
    {
      "epoch": 2.6512287334593574,
      "grad_norm": 0.43217438534152663,
      "learning_rate": 3.670139042678955e-06,
      "loss": 0.3899480104446411,
      "step": 2805
    },
    {
      "epoch": 2.6521739130434785,
      "grad_norm": 0.4477293217871917,
      "learning_rate": 3.6505719487156697e-06,
      "loss": 0.4390020966529846,
      "step": 2806
    },
    {
      "epoch": 2.653119092627599,
      "grad_norm": 0.4380458335366412,
      "learning_rate": 3.631055178091758e-06,
      "loss": 0.4460015594959259,
      "step": 2807
    },
    {
      "epoch": 2.6540642722117203,
      "grad_norm": 0.38471631452392385,
      "learning_rate": 3.611588751997336e-06,
      "loss": 0.3785001039505005,
      "step": 2808
    },
    {
      "epoch": 2.6550094517958414,
      "grad_norm": 0.4458282040238457,
      "learning_rate": 3.5921726915678244e-06,
      "loss": 0.38587838411331177,
      "step": 2809
    },
    {
      "epoch": 2.655954631379962,
      "grad_norm": 0.4188993254112753,
      "learning_rate": 3.5728070178839943e-06,
      "loss": 0.38591229915618896,
      "step": 2810
    },
    {
      "epoch": 2.656899810964083,
      "grad_norm": 0.43122362289829047,
      "learning_rate": 3.5534917519718824e-06,
      "loss": 0.4119555950164795,
      "step": 2811
    },
    {
      "epoch": 2.6578449905482042,
      "grad_norm": 0.46283332425154955,
      "learning_rate": 3.5342269148028127e-06,
      "loss": 0.45798349380493164,
      "step": 2812
    },
    {
      "epoch": 2.658790170132325,
      "grad_norm": 0.41854695522299884,
      "learning_rate": 3.5150125272933652e-06,
      "loss": 0.3970344066619873,
      "step": 2813
    },
    {
      "epoch": 2.659735349716446,
      "grad_norm": 0.5004574172437488,
      "learning_rate": 3.495848610305319e-06,
      "loss": 0.4616374969482422,
      "step": 2814
    },
    {
      "epoch": 2.660680529300567,
      "grad_norm": 0.4839831425168395,
      "learning_rate": 3.476735184645674e-06,
      "loss": 0.40665626525878906,
      "step": 2815
    },
    {
      "epoch": 2.661625708884688,
      "grad_norm": 0.446188710060192,
      "learning_rate": 3.457672271066598e-06,
      "loss": 0.46110808849334717,
      "step": 2816
    },
    {
      "epoch": 2.6625708884688093,
      "grad_norm": 0.44077429574135163,
      "learning_rate": 3.438659890265422e-06,
      "loss": 0.4193815588951111,
      "step": 2817
    },
    {
      "epoch": 2.66351606805293,
      "grad_norm": 0.4535459935768628,
      "learning_rate": 3.41969806288463e-06,
      "loss": 0.42894214391708374,
      "step": 2818
    },
    {
      "epoch": 2.664461247637051,
      "grad_norm": 0.4764396472244361,
      "learning_rate": 3.4007868095117744e-06,
      "loss": 0.4779616594314575,
      "step": 2819
    },
    {
      "epoch": 2.665406427221172,
      "grad_norm": 0.43290738300718296,
      "learning_rate": 3.381926150679543e-06,
      "loss": 0.4659026265144348,
      "step": 2820
    },
    {
      "epoch": 2.666351606805293,
      "grad_norm": 0.44339333309457957,
      "learning_rate": 3.3631161068656603e-06,
      "loss": 0.3896799683570862,
      "step": 2821
    },
    {
      "epoch": 2.667296786389414,
      "grad_norm": 0.4367848922643944,
      "learning_rate": 3.3443566984929176e-06,
      "loss": 0.3890998065471649,
      "step": 2822
    },
    {
      "epoch": 2.668241965973535,
      "grad_norm": 0.48950342799502616,
      "learning_rate": 3.3256479459291055e-06,
      "loss": 0.45650815963745117,
      "step": 2823
    },
    {
      "epoch": 2.6691871455576557,
      "grad_norm": 0.49497288556230284,
      "learning_rate": 3.306989869487037e-06,
      "loss": 0.47000956535339355,
      "step": 2824
    },
    {
      "epoch": 2.670132325141777,
      "grad_norm": 0.48166725819054634,
      "learning_rate": 3.288382489424502e-06,
      "loss": 0.4370376169681549,
      "step": 2825
    },
    {
      "epoch": 2.671077504725898,
      "grad_norm": 0.437095459497125,
      "learning_rate": 3.2698258259442283e-06,
      "loss": 0.4270906448364258,
      "step": 2826
    },
    {
      "epoch": 2.672022684310019,
      "grad_norm": 0.44603835146410875,
      "learning_rate": 3.2513198991939053e-06,
      "loss": 0.4122464060783386,
      "step": 2827
    },
    {
      "epoch": 2.67296786389414,
      "grad_norm": 0.5365359874665366,
      "learning_rate": 3.232864729266111e-06,
      "loss": 0.4901719391345978,
      "step": 2828
    },
    {
      "epoch": 2.6739130434782608,
      "grad_norm": 0.4502826378246867,
      "learning_rate": 3.2144603361983328e-06,
      "loss": 0.41424089670181274,
      "step": 2829
    },
    {
      "epoch": 2.674858223062382,
      "grad_norm": 0.46720898561711593,
      "learning_rate": 3.196106739972926e-06,
      "loss": 0.3911948800086975,
      "step": 2830
    },
    {
      "epoch": 2.675803402646503,
      "grad_norm": 0.41578016636936393,
      "learning_rate": 3.177803960517073e-06,
      "loss": 0.377789169549942,
      "step": 2831
    },
    {
      "epoch": 2.6767485822306236,
      "grad_norm": 0.43614534420095097,
      "learning_rate": 3.1595520177028215e-06,
      "loss": 0.4432322382926941,
      "step": 2832
    },
    {
      "epoch": 2.6776937618147447,
      "grad_norm": 0.4234195770360025,
      "learning_rate": 3.1413509313469812e-06,
      "loss": 0.38369303941726685,
      "step": 2833
    },
    {
      "epoch": 2.678638941398866,
      "grad_norm": 0.41538923567692354,
      "learning_rate": 3.1232007212111725e-06,
      "loss": 0.39200299978256226,
      "step": 2834
    },
    {
      "epoch": 2.679584120982987,
      "grad_norm": 0.3926340617254343,
      "learning_rate": 3.10510140700177e-06,
      "loss": 0.3733380436897278,
      "step": 2835
    },
    {
      "epoch": 2.6805293005671076,
      "grad_norm": 0.5209301447482235,
      "learning_rate": 3.087053008369889e-06,
      "loss": 0.4739784598350525,
      "step": 2836
    },
    {
      "epoch": 2.6814744801512287,
      "grad_norm": 0.4693747945093216,
      "learning_rate": 3.0690555449113643e-06,
      "loss": 0.4697369933128357,
      "step": 2837
    },
    {
      "epoch": 2.68241965973535,
      "grad_norm": 0.47328553434500154,
      "learning_rate": 3.0511090361667216e-06,
      "loss": 0.45481106638908386,
      "step": 2838
    },
    {
      "epoch": 2.683364839319471,
      "grad_norm": 0.4540068551399434,
      "learning_rate": 3.0332135016211794e-06,
      "loss": 0.4560334086418152,
      "step": 2839
    },
    {
      "epoch": 2.6843100189035916,
      "grad_norm": 0.46828084092244304,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 0.4010257124900818,
      "step": 2840
    },
    {
      "epoch": 2.6852551984877127,
      "grad_norm": 0.4883813066838279,
      "learning_rate": 2.997575432791444e-06,
      "loss": 0.4869145154953003,
      "step": 2841
    },
    {
      "epoch": 2.6862003780718338,
      "grad_norm": 0.46479805130140456,
      "learning_rate": 2.9798329372008825e-06,
      "loss": 0.4238889813423157,
      "step": 2842
    },
    {
      "epoch": 2.6871455576559544,
      "grad_norm": 0.4621862229153414,
      "learning_rate": 2.9621414931965794e-06,
      "loss": 0.4640699028968811,
      "step": 2843
    },
    {
      "epoch": 2.6880907372400755,
      "grad_norm": 0.45740096601354197,
      "learning_rate": 2.9445011199868254e-06,
      "loss": 0.4442843496799469,
      "step": 2844
    },
    {
      "epoch": 2.6890359168241966,
      "grad_norm": 0.4493392086458401,
      "learning_rate": 2.9269118367244385e-06,
      "loss": 0.4179978370666504,
      "step": 2845
    },
    {
      "epoch": 2.6899810964083177,
      "grad_norm": 0.5022342202510234,
      "learning_rate": 2.90937366250677e-06,
      "loss": 0.4576505124568939,
      "step": 2846
    },
    {
      "epoch": 2.690926275992439,
      "grad_norm": 0.4359455796039392,
      "learning_rate": 2.8918866163756997e-06,
      "loss": 0.3959827125072479,
      "step": 2847
    },
    {
      "epoch": 2.6918714555765595,
      "grad_norm": 0.4460281867653098,
      "learning_rate": 2.874450717317556e-06,
      "loss": 0.414617121219635,
      "step": 2848
    },
    {
      "epoch": 2.6928166351606806,
      "grad_norm": 0.45654820763009996,
      "learning_rate": 2.8570659842631854e-06,
      "loss": 0.4314059615135193,
      "step": 2849
    },
    {
      "epoch": 2.6937618147448017,
      "grad_norm": 0.5236242306605328,
      "learning_rate": 2.839732436087833e-06,
      "loss": 0.4675754904747009,
      "step": 2850
    },
    {
      "epoch": 2.6947069943289224,
      "grad_norm": 0.5140296363440973,
      "learning_rate": 2.8224500916111952e-06,
      "loss": 0.5156667232513428,
      "step": 2851
    },
    {
      "epoch": 2.6956521739130435,
      "grad_norm": 0.4182749784308828,
      "learning_rate": 2.8052189695973894e-06,
      "loss": 0.37500423192977905,
      "step": 2852
    },
    {
      "epoch": 2.6965973534971646,
      "grad_norm": 0.505744231541097,
      "learning_rate": 2.788039088754879e-06,
      "loss": 0.46950486302375793,
      "step": 2853
    },
    {
      "epoch": 2.697542533081285,
      "grad_norm": 0.45889505647910445,
      "learning_rate": 2.7709104677365315e-06,
      "loss": 0.45169293880462646,
      "step": 2854
    },
    {
      "epoch": 2.6984877126654063,
      "grad_norm": 0.4633879584521977,
      "learning_rate": 2.7538331251395266e-06,
      "loss": 0.48943984508514404,
      "step": 2855
    },
    {
      "epoch": 2.6994328922495274,
      "grad_norm": 0.4687135388805424,
      "learning_rate": 2.7368070795053934e-06,
      "loss": 0.395643413066864,
      "step": 2856
    },
    {
      "epoch": 2.7003780718336485,
      "grad_norm": 0.4661926484256157,
      "learning_rate": 2.7198323493199728e-06,
      "loss": 0.45326530933380127,
      "step": 2857
    },
    {
      "epoch": 2.7013232514177696,
      "grad_norm": 0.4331908091595055,
      "learning_rate": 2.70290895301335e-06,
      "loss": 0.43027716875076294,
      "step": 2858
    },
    {
      "epoch": 2.7022684310018903,
      "grad_norm": 0.42436730593205935,
      "learning_rate": 2.6860369089599215e-06,
      "loss": 0.43217557668685913,
      "step": 2859
    },
    {
      "epoch": 2.7032136105860114,
      "grad_norm": 0.45175809984430887,
      "learning_rate": 2.6692162354782944e-06,
      "loss": 0.4347071051597595,
      "step": 2860
    },
    {
      "epoch": 2.7041587901701325,
      "grad_norm": 0.3698993973543292,
      "learning_rate": 2.6524469508313278e-06,
      "loss": 0.33577749133110046,
      "step": 2861
    },
    {
      "epoch": 2.705103969754253,
      "grad_norm": 0.4528974593426443,
      "learning_rate": 2.635729073226051e-06,
      "loss": 0.3851119875907898,
      "step": 2862
    },
    {
      "epoch": 2.7060491493383743,
      "grad_norm": 0.4109051473449968,
      "learning_rate": 2.619062620813728e-06,
      "loss": 0.40276890993118286,
      "step": 2863
    },
    {
      "epoch": 2.7069943289224954,
      "grad_norm": 0.4381213765163317,
      "learning_rate": 2.6024476116897456e-06,
      "loss": 0.4218601584434509,
      "step": 2864
    },
    {
      "epoch": 2.707939508506616,
      "grad_norm": 0.4288548823658911,
      "learning_rate": 2.585884063893651e-06,
      "loss": 0.4004361629486084,
      "step": 2865
    },
    {
      "epoch": 2.708884688090737,
      "grad_norm": 0.4642305395285001,
      "learning_rate": 2.5693719954091255e-06,
      "loss": 0.423369824886322,
      "step": 2866
    },
    {
      "epoch": 2.7098298676748582,
      "grad_norm": 0.48404302494774104,
      "learning_rate": 2.5529114241639395e-06,
      "loss": 0.4782102406024933,
      "step": 2867
    },
    {
      "epoch": 2.7107750472589793,
      "grad_norm": 0.46426313450775636,
      "learning_rate": 2.5365023680299747e-06,
      "loss": 0.4819539189338684,
      "step": 2868
    },
    {
      "epoch": 2.7117202268431004,
      "grad_norm": 0.4329010340309318,
      "learning_rate": 2.520144844823169e-06,
      "loss": 0.39002472162246704,
      "step": 2869
    },
    {
      "epoch": 2.712665406427221,
      "grad_norm": 0.49888028385550376,
      "learning_rate": 2.5038388723034932e-06,
      "loss": 0.5279086828231812,
      "step": 2870
    },
    {
      "epoch": 2.713610586011342,
      "grad_norm": 0.4621450090648458,
      "learning_rate": 2.487584468174986e-06,
      "loss": 0.45123398303985596,
      "step": 2871
    },
    {
      "epoch": 2.7145557655954633,
      "grad_norm": 0.47243310036629865,
      "learning_rate": 2.4713816500856468e-06,
      "loss": 0.4882585108280182,
      "step": 2872
    },
    {
      "epoch": 2.715500945179584,
      "grad_norm": 0.4780558531305996,
      "learning_rate": 2.4552304356275034e-06,
      "loss": 0.39195841550827026,
      "step": 2873
    },
    {
      "epoch": 2.716446124763705,
      "grad_norm": 0.5127393341339171,
      "learning_rate": 2.439130842336551e-06,
      "loss": 0.44555893540382385,
      "step": 2874
    },
    {
      "epoch": 2.717391304347826,
      "grad_norm": 0.42207798344070563,
      "learning_rate": 2.4230828876927294e-06,
      "loss": 0.39709731936454773,
      "step": 2875
    },
    {
      "epoch": 2.718336483931947,
      "grad_norm": 0.4401664118958827,
      "learning_rate": 2.4070865891199167e-06,
      "loss": 0.41347622871398926,
      "step": 2876
    },
    {
      "epoch": 2.719281663516068,
      "grad_norm": 0.4286983486293977,
      "learning_rate": 2.391141963985888e-06,
      "loss": 0.3952424228191376,
      "step": 2877
    },
    {
      "epoch": 2.720226843100189,
      "grad_norm": 0.4323625640299157,
      "learning_rate": 2.3752490296023387e-06,
      "loss": 0.42811721563339233,
      "step": 2878
    },
    {
      "epoch": 2.72117202268431,
      "grad_norm": 0.4731407402233115,
      "learning_rate": 2.3594078032248333e-06,
      "loss": 0.4910525679588318,
      "step": 2879
    },
    {
      "epoch": 2.7221172022684312,
      "grad_norm": 0.45777363466847754,
      "learning_rate": 2.343618302052808e-06,
      "loss": 0.4475446939468384,
      "step": 2880
    },
    {
      "epoch": 2.723062381852552,
      "grad_norm": 0.40019467114355584,
      "learning_rate": 2.327880543229505e-06,
      "loss": 0.3782731592655182,
      "step": 2881
    },
    {
      "epoch": 2.724007561436673,
      "grad_norm": 0.461262048680245,
      "learning_rate": 2.3121945438420167e-06,
      "loss": 0.4235340356826782,
      "step": 2882
    },
    {
      "epoch": 2.724952741020794,
      "grad_norm": 0.41396716524671456,
      "learning_rate": 2.296560320921237e-06,
      "loss": 0.4088025987148285,
      "step": 2883
    },
    {
      "epoch": 2.7258979206049148,
      "grad_norm": 0.4081384646443075,
      "learning_rate": 2.2809778914418156e-06,
      "loss": 0.3787614107131958,
      "step": 2884
    },
    {
      "epoch": 2.726843100189036,
      "grad_norm": 0.48242965205283744,
      "learning_rate": 2.265447272322213e-06,
      "loss": 0.4488810896873474,
      "step": 2885
    },
    {
      "epoch": 2.727788279773157,
      "grad_norm": 0.41456222554670275,
      "learning_rate": 2.2499684804246134e-06,
      "loss": 0.39070436358451843,
      "step": 2886
    },
    {
      "epoch": 2.7287334593572776,
      "grad_norm": 0.4900698187511008,
      "learning_rate": 2.2345415325549123e-06,
      "loss": 0.4820168912410736,
      "step": 2887
    },
    {
      "epoch": 2.7296786389413987,
      "grad_norm": 0.4485339999105653,
      "learning_rate": 2.21916644546275e-06,
      "loss": 0.41927406191825867,
      "step": 2888
    },
    {
      "epoch": 2.73062381852552,
      "grad_norm": 0.44035666222202346,
      "learning_rate": 2.203843235841435e-06,
      "loss": 0.43869659304618835,
      "step": 2889
    },
    {
      "epoch": 2.731568998109641,
      "grad_norm": 0.44687419780367316,
      "learning_rate": 2.1885719203279588e-06,
      "loss": 0.41197746992111206,
      "step": 2890
    },
    {
      "epoch": 2.732514177693762,
      "grad_norm": 0.46353478148050625,
      "learning_rate": 2.17335251550298e-06,
      "loss": 0.4402436316013336,
      "step": 2891
    },
    {
      "epoch": 2.7334593572778827,
      "grad_norm": 0.5003651092629762,
      "learning_rate": 2.1581850378907866e-06,
      "loss": 0.4257342517375946,
      "step": 2892
    },
    {
      "epoch": 2.734404536862004,
      "grad_norm": 0.48382751680745456,
      "learning_rate": 2.143069503959283e-06,
      "loss": 0.4028918743133545,
      "step": 2893
    },
    {
      "epoch": 2.735349716446125,
      "grad_norm": 0.4537586724217299,
      "learning_rate": 2.1280059301199694e-06,
      "loss": 0.41585269570350647,
      "step": 2894
    },
    {
      "epoch": 2.7362948960302456,
      "grad_norm": 0.415939329599401,
      "learning_rate": 2.112994332727952e-06,
      "loss": 0.38194388151168823,
      "step": 2895
    },
    {
      "epoch": 2.7372400756143667,
      "grad_norm": 0.42036206608440696,
      "learning_rate": 2.0980347280818935e-06,
      "loss": 0.38771429657936096,
      "step": 2896
    },
    {
      "epoch": 2.7381852551984878,
      "grad_norm": 0.4803937821244692,
      "learning_rate": 2.083127132424012e-06,
      "loss": 0.44784826040267944,
      "step": 2897
    },
    {
      "epoch": 2.7391304347826084,
      "grad_norm": 0.4533554206411885,
      "learning_rate": 2.0682715619400438e-06,
      "loss": 0.4842197895050049,
      "step": 2898
    },
    {
      "epoch": 2.7400756143667295,
      "grad_norm": 0.4376646783717641,
      "learning_rate": 2.0534680327592426e-06,
      "loss": 0.4360918700695038,
      "step": 2899
    },
    {
      "epoch": 2.7410207939508506,
      "grad_norm": 0.43955393286156846,
      "learning_rate": 2.0387165609543736e-06,
      "loss": 0.4137001633644104,
      "step": 2900
    },
    {
      "epoch": 2.7419659735349717,
      "grad_norm": 0.43671721676392894,
      "learning_rate": 2.024017162541664e-06,
      "loss": 0.37250959873199463,
      "step": 2901
    },
    {
      "epoch": 2.742911153119093,
      "grad_norm": 0.43027230997402294,
      "learning_rate": 2.0093698534808246e-06,
      "loss": 0.4412021040916443,
      "step": 2902
    },
    {
      "epoch": 2.7438563327032135,
      "grad_norm": 0.4375112441058645,
      "learning_rate": 1.9947746496749796e-06,
      "loss": 0.4216412305831909,
      "step": 2903
    },
    {
      "epoch": 2.7448015122873346,
      "grad_norm": 0.45110423640578,
      "learning_rate": 1.980231566970714e-06,
      "loss": 0.4360377788543701,
      "step": 2904
    },
    {
      "epoch": 2.7457466918714557,
      "grad_norm": 0.44823726564156485,
      "learning_rate": 1.9657406211579966e-06,
      "loss": 0.40627503395080566,
      "step": 2905
    },
    {
      "epoch": 2.7466918714555764,
      "grad_norm": 0.4466399220202717,
      "learning_rate": 1.951301827970192e-06,
      "loss": 0.41477635502815247,
      "step": 2906
    },
    {
      "epoch": 2.7476370510396975,
      "grad_norm": 0.431038198251035,
      "learning_rate": 1.9369152030840556e-06,
      "loss": 0.3584004044532776,
      "step": 2907
    },
    {
      "epoch": 2.7485822306238186,
      "grad_norm": 0.46737932001445504,
      "learning_rate": 1.9225807621196967e-06,
      "loss": 0.41146862506866455,
      "step": 2908
    },
    {
      "epoch": 2.7495274102079397,
      "grad_norm": 0.4509609661955297,
      "learning_rate": 1.908298520640572e-06,
      "loss": 0.4594671428203583,
      "step": 2909
    },
    {
      "epoch": 2.7504725897920603,
      "grad_norm": 0.48013871433300076,
      "learning_rate": 1.8940684941534392e-06,
      "loss": 0.49168309569358826,
      "step": 2910
    },
    {
      "epoch": 2.7514177693761814,
      "grad_norm": 0.42758335047859847,
      "learning_rate": 1.879890698108383e-06,
      "loss": 0.41360458731651306,
      "step": 2911
    },
    {
      "epoch": 2.7523629489603025,
      "grad_norm": 0.4441588368201936,
      "learning_rate": 1.8657651478987791e-06,
      "loss": 0.4512834846973419,
      "step": 2912
    },
    {
      "epoch": 2.7533081285444236,
      "grad_norm": 0.48283731959757514,
      "learning_rate": 1.8516918588612708e-06,
      "loss": 0.4554654359817505,
      "step": 2913
    },
    {
      "epoch": 2.7542533081285443,
      "grad_norm": 0.4550431774370647,
      "learning_rate": 1.8376708462757796e-06,
      "loss": 0.4241878092288971,
      "step": 2914
    },
    {
      "epoch": 2.7551984877126654,
      "grad_norm": 0.45416429698510413,
      "learning_rate": 1.8237021253654396e-06,
      "loss": 0.45071732997894287,
      "step": 2915
    },
    {
      "epoch": 2.7561436672967865,
      "grad_norm": 0.4537239623665792,
      "learning_rate": 1.8097857112966243e-06,
      "loss": 0.40416979789733887,
      "step": 2916
    },
    {
      "epoch": 2.757088846880907,
      "grad_norm": 0.46727615131216127,
      "learning_rate": 1.7959216191789141e-06,
      "loss": 0.40620285272598267,
      "step": 2917
    },
    {
      "epoch": 2.7580340264650283,
      "grad_norm": 0.4575446923292323,
      "learning_rate": 1.782109864065079e-06,
      "loss": 0.444033682346344,
      "step": 2918
    },
    {
      "epoch": 2.7589792060491494,
      "grad_norm": 0.4758770193866977,
      "learning_rate": 1.7683504609510848e-06,
      "loss": 0.4563744068145752,
      "step": 2919
    },
    {
      "epoch": 2.7599243856332705,
      "grad_norm": 0.4513645168273106,
      "learning_rate": 1.7546434247760146e-06,
      "loss": 0.4116581678390503,
      "step": 2920
    },
    {
      "epoch": 2.7608695652173916,
      "grad_norm": 0.5040851534694669,
      "learning_rate": 1.7409887704221418e-06,
      "loss": 0.4595235586166382,
      "step": 2921
    },
    {
      "epoch": 2.7618147448015122,
      "grad_norm": 0.4311472603807616,
      "learning_rate": 1.7273865127148293e-06,
      "loss": 0.40190935134887695,
      "step": 2922
    },
    {
      "epoch": 2.7627599243856333,
      "grad_norm": 0.4779480909708149,
      "learning_rate": 1.713836666422569e-06,
      "loss": 0.46864473819732666,
      "step": 2923
    },
    {
      "epoch": 2.7637051039697544,
      "grad_norm": 0.4701963305639444,
      "learning_rate": 1.7003392462569545e-06,
      "loss": 0.4465354084968567,
      "step": 2924
    },
    {
      "epoch": 2.764650283553875,
      "grad_norm": 0.5221529841362399,
      "learning_rate": 1.6868942668726407e-06,
      "loss": 0.49709129333496094,
      "step": 2925
    },
    {
      "epoch": 2.765595463137996,
      "grad_norm": 0.4765862068404574,
      "learning_rate": 1.6735017428673561e-06,
      "loss": 0.43945538997650146,
      "step": 2926
    },
    {
      "epoch": 2.7665406427221173,
      "grad_norm": 0.39059160869193577,
      "learning_rate": 1.6601616887818749e-06,
      "loss": 0.38754746317863464,
      "step": 2927
    },
    {
      "epoch": 2.767485822306238,
      "grad_norm": 0.47033152454192545,
      "learning_rate": 1.6468741190999949e-06,
      "loss": 0.41859176754951477,
      "step": 2928
    },
    {
      "epoch": 2.768431001890359,
      "grad_norm": 0.5433786809782114,
      "learning_rate": 1.633639048248542e-06,
      "loss": 0.4892963767051697,
      "step": 2929
    },
    {
      "epoch": 2.76937618147448,
      "grad_norm": 0.4452174836083594,
      "learning_rate": 1.6204564905973386e-06,
      "loss": 0.4126704931259155,
      "step": 2930
    },
    {
      "epoch": 2.7703213610586013,
      "grad_norm": 0.4593381295728348,
      "learning_rate": 1.6073264604591964e-06,
      "loss": 0.4702644944190979,
      "step": 2931
    },
    {
      "epoch": 2.7712665406427224,
      "grad_norm": 0.40644000535290803,
      "learning_rate": 1.594248972089879e-06,
      "loss": 0.3881847858428955,
      "step": 2932
    },
    {
      "epoch": 2.772211720226843,
      "grad_norm": 0.47193860728479897,
      "learning_rate": 1.5812240396881061e-06,
      "loss": 0.4634220600128174,
      "step": 2933
    },
    {
      "epoch": 2.773156899810964,
      "grad_norm": 0.44601244435626436,
      "learning_rate": 1.5682516773955547e-06,
      "loss": 0.42718005180358887,
      "step": 2934
    },
    {
      "epoch": 2.7741020793950852,
      "grad_norm": 0.46362351449512773,
      "learning_rate": 1.555331899296808e-06,
      "loss": 0.4327946603298187,
      "step": 2935
    },
    {
      "epoch": 2.775047258979206,
      "grad_norm": 0.45566147219007497,
      "learning_rate": 1.542464719419373e-06,
      "loss": 0.46008551120758057,
      "step": 2936
    },
    {
      "epoch": 2.775992438563327,
      "grad_norm": 0.475756929119225,
      "learning_rate": 1.5296501517336193e-06,
      "loss": 0.47109168767929077,
      "step": 2937
    },
    {
      "epoch": 2.776937618147448,
      "grad_norm": 0.4906454345922015,
      "learning_rate": 1.516888210152828e-06,
      "loss": 0.4401085376739502,
      "step": 2938
    },
    {
      "epoch": 2.7778827977315688,
      "grad_norm": 0.44248957961747826,
      "learning_rate": 1.5041789085331048e-06,
      "loss": 0.4351924657821655,
      "step": 2939
    },
    {
      "epoch": 2.77882797731569,
      "grad_norm": 0.4533681720698251,
      "learning_rate": 1.4915222606734392e-06,
      "loss": 0.44652292132377625,
      "step": 2940
    },
    {
      "epoch": 2.779773156899811,
      "grad_norm": 0.4357825184653124,
      "learning_rate": 1.4789182803156332e-06,
      "loss": 0.4565678536891937,
      "step": 2941
    },
    {
      "epoch": 2.780718336483932,
      "grad_norm": 0.4797563125205837,
      "learning_rate": 1.4663669811443014e-06,
      "loss": 0.4202726483345032,
      "step": 2942
    },
    {
      "epoch": 2.781663516068053,
      "grad_norm": 0.47326010682161307,
      "learning_rate": 1.4538683767868822e-06,
      "loss": 0.4824743866920471,
      "step": 2943
    },
    {
      "epoch": 2.782608695652174,
      "grad_norm": 0.4327370930493396,
      "learning_rate": 1.4414224808135702e-06,
      "loss": 0.41451922059059143,
      "step": 2944
    },
    {
      "epoch": 2.783553875236295,
      "grad_norm": 0.47633134828235724,
      "learning_rate": 1.429029306737345e-06,
      "loss": 0.45273709297180176,
      "step": 2945
    },
    {
      "epoch": 2.784499054820416,
      "grad_norm": 0.44514888121863605,
      "learning_rate": 1.4166888680139657e-06,
      "loss": 0.4363822638988495,
      "step": 2946
    },
    {
      "epoch": 2.7854442344045367,
      "grad_norm": 0.40740880974220994,
      "learning_rate": 1.404401178041903e-06,
      "loss": 0.44579190015792847,
      "step": 2947
    },
    {
      "epoch": 2.786389413988658,
      "grad_norm": 0.47754015253218873,
      "learning_rate": 1.3921662501623855e-06,
      "loss": 0.42761433124542236,
      "step": 2948
    },
    {
      "epoch": 2.787334593572779,
      "grad_norm": 0.46808256511982654,
      "learning_rate": 1.3799840976593204e-06,
      "loss": 0.4604160487651825,
      "step": 2949
    },
    {
      "epoch": 2.7882797731568996,
      "grad_norm": 0.42374654335140166,
      "learning_rate": 1.3678547337593494e-06,
      "loss": 0.3988068103790283,
      "step": 2950
    },
    {
      "epoch": 2.7892249527410207,
      "grad_norm": 0.49605949440667346,
      "learning_rate": 1.3557781716317774e-06,
      "loss": 0.45192304253578186,
      "step": 2951
    },
    {
      "epoch": 2.7901701323251418,
      "grad_norm": 0.3997798023279822,
      "learning_rate": 1.343754424388588e-06,
      "loss": 0.3615618050098419,
      "step": 2952
    },
    {
      "epoch": 2.791115311909263,
      "grad_norm": 0.4218780506098901,
      "learning_rate": 1.3317835050844275e-06,
      "loss": 0.3860602378845215,
      "step": 2953
    },
    {
      "epoch": 2.792060491493384,
      "grad_norm": 0.418225984784419,
      "learning_rate": 1.319865426716571e-06,
      "loss": 0.37652450799942017,
      "step": 2954
    },
    {
      "epoch": 2.7930056710775046,
      "grad_norm": 0.44075611964180345,
      "learning_rate": 1.3080002022249405e-06,
      "loss": 0.4157121181488037,
      "step": 2955
    },
    {
      "epoch": 2.7939508506616257,
      "grad_norm": 0.4690172149879725,
      "learning_rate": 1.2961878444920528e-06,
      "loss": 0.46241623163223267,
      "step": 2956
    },
    {
      "epoch": 2.794896030245747,
      "grad_norm": 0.4596216508427531,
      "learning_rate": 1.2844283663430324e-06,
      "loss": 0.38835999369621277,
      "step": 2957
    },
    {
      "epoch": 2.7958412098298675,
      "grad_norm": 0.4602959937849937,
      "learning_rate": 1.272721780545605e-06,
      "loss": 0.4190332293510437,
      "step": 2958
    },
    {
      "epoch": 2.7967863894139886,
      "grad_norm": 0.5180745185463717,
      "learning_rate": 1.2610680998100478e-06,
      "loss": 0.49982139468193054,
      "step": 2959
    },
    {
      "epoch": 2.7977315689981097,
      "grad_norm": 0.44063943072866213,
      "learning_rate": 1.2494673367892062e-06,
      "loss": 0.4197500944137573,
      "step": 2960
    },
    {
      "epoch": 2.7986767485822304,
      "grad_norm": 0.4640346745843304,
      "learning_rate": 1.237919504078472e-06,
      "loss": 0.36017078161239624,
      "step": 2961
    },
    {
      "epoch": 2.7996219281663515,
      "grad_norm": 0.4193246258716871,
      "learning_rate": 1.2264246142157653e-06,
      "loss": 0.38449156284332275,
      "step": 2962
    },
    {
      "epoch": 2.8005671077504726,
      "grad_norm": 0.4672794850504573,
      "learning_rate": 1.2149826796815367e-06,
      "loss": 0.4355344772338867,
      "step": 2963
    },
    {
      "epoch": 2.8015122873345937,
      "grad_norm": 0.457237371391946,
      "learning_rate": 1.2035937128987096e-06,
      "loss": 0.42571377754211426,
      "step": 2964
    },
    {
      "epoch": 2.802457466918715,
      "grad_norm": 0.46686001713566166,
      "learning_rate": 1.1922577262327373e-06,
      "loss": 0.44667574763298035,
      "step": 2965
    },
    {
      "epoch": 2.8034026465028354,
      "grad_norm": 0.425558366672215,
      "learning_rate": 1.1809747319915187e-06,
      "loss": 0.4480895698070526,
      "step": 2966
    },
    {
      "epoch": 2.8043478260869565,
      "grad_norm": 0.4365623413367783,
      "learning_rate": 1.1697447424254438e-06,
      "loss": 0.41138219833374023,
      "step": 2967
    },
    {
      "epoch": 2.8052930056710776,
      "grad_norm": 0.4831722928887866,
      "learning_rate": 1.1585677697273312e-06,
      "loss": 0.44401198625564575,
      "step": 2968
    },
    {
      "epoch": 2.8062381852551983,
      "grad_norm": 0.4716689664397205,
      "learning_rate": 1.147443826032446e-06,
      "loss": 0.4811118543148041,
      "step": 2969
    },
    {
      "epoch": 2.8071833648393194,
      "grad_norm": 0.4514670354549091,
      "learning_rate": 1.1363729234184827e-06,
      "loss": 0.44799041748046875,
      "step": 2970
    },
    {
      "epoch": 2.8081285444234405,
      "grad_norm": 0.4465717675115479,
      "learning_rate": 1.1253550739055373e-06,
      "loss": 0.4696815609931946,
      "step": 2971
    },
    {
      "epoch": 2.809073724007561,
      "grad_norm": 0.43577556246057625,
      "learning_rate": 1.1143902894561132e-06,
      "loss": 0.4255073070526123,
      "step": 2972
    },
    {
      "epoch": 2.8100189035916823,
      "grad_norm": 0.4108965153569258,
      "learning_rate": 1.103478581975087e-06,
      "loss": 0.39006859064102173,
      "step": 2973
    },
    {
      "epoch": 2.8109640831758034,
      "grad_norm": 0.49521972314244256,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.44066542387008667,
      "step": 2974
    },
    {
      "epoch": 2.8119092627599245,
      "grad_norm": 0.4379257632100853,
      "learning_rate": 1.0818144452496292e-06,
      "loss": 0.4136120080947876,
      "step": 2975
    },
    {
      "epoch": 2.8128544423440456,
      "grad_norm": 0.4687768967562004,
      "learning_rate": 1.0710620395267711e-06,
      "loss": 0.4307497441768646,
      "step": 2976
    },
    {
      "epoch": 2.8137996219281662,
      "grad_norm": 0.4483013195499789,
      "learning_rate": 1.060362757815453e-06,
      "loss": 0.44938555359840393,
      "step": 2977
    },
    {
      "epoch": 2.8147448015122873,
      "grad_norm": 0.4215165680633903,
      "learning_rate": 1.049716611732282e-06,
      "loss": 0.396475613117218,
      "step": 2978
    },
    {
      "epoch": 2.8156899810964084,
      "grad_norm": 0.4304073733209891,
      "learning_rate": 1.0391236128361903e-06,
      "loss": 0.382732093334198,
      "step": 2979
    },
    {
      "epoch": 2.816635160680529,
      "grad_norm": 0.4238910729569748,
      "learning_rate": 1.0285837726283998e-06,
      "loss": 0.404712438583374,
      "step": 2980
    },
    {
      "epoch": 2.81758034026465,
      "grad_norm": 0.43125099983296755,
      "learning_rate": 1.0180971025524121e-06,
      "loss": 0.4304637312889099,
      "step": 2981
    },
    {
      "epoch": 2.8185255198487713,
      "grad_norm": 0.4493982028294385,
      "learning_rate": 1.0076636139940088e-06,
      "loss": 0.4751417934894562,
      "step": 2982
    },
    {
      "epoch": 2.8194706994328924,
      "grad_norm": 0.4729941302153935,
      "learning_rate": 9.972833182812224e-07,
      "loss": 0.4719465970993042,
      "step": 2983
    },
    {
      "epoch": 2.820415879017013,
      "grad_norm": 0.43220120041194254,
      "learning_rate": 9.869562266843379e-07,
      "loss": 0.39151430130004883,
      "step": 2984
    },
    {
      "epoch": 2.821361058601134,
      "grad_norm": 0.4924497168106336,
      "learning_rate": 9.76682350415875e-07,
      "loss": 0.4562298059463501,
      "step": 2985
    },
    {
      "epoch": 2.8223062381852553,
      "grad_norm": 0.4565056444487534,
      "learning_rate": 9.664617006305664e-07,
      "loss": 0.4322068393230438,
      "step": 2986
    },
    {
      "epoch": 2.8232514177693764,
      "grad_norm": 0.4737741814123948,
      "learning_rate": 9.562942884253633e-07,
      "loss": 0.44013693928718567,
      "step": 2987
    },
    {
      "epoch": 2.824196597353497,
      "grad_norm": 0.4362619444997381,
      "learning_rate": 9.461801248394131e-07,
      "loss": 0.4072737693786621,
      "step": 2988
    },
    {
      "epoch": 2.825141776937618,
      "grad_norm": 0.44247519279753394,
      "learning_rate": 9.361192208540425e-07,
      "loss": 0.40677428245544434,
      "step": 2989
    },
    {
      "epoch": 2.8260869565217392,
      "grad_norm": 0.515963419195998,
      "learning_rate": 9.261115873927695e-07,
      "loss": 0.4405622184276581,
      "step": 2990
    },
    {
      "epoch": 2.82703213610586,
      "grad_norm": 0.435734915424571,
      "learning_rate": 9.161572353212578e-07,
      "loss": 0.43056613206863403,
      "step": 2991
    },
    {
      "epoch": 2.827977315689981,
      "grad_norm": 0.4621133300411559,
      "learning_rate": 9.062561754473231e-07,
      "loss": 0.407729834318161,
      "step": 2992
    },
    {
      "epoch": 2.828922495274102,
      "grad_norm": 0.46628410038315754,
      "learning_rate": 8.96408418520922e-07,
      "loss": 0.4225010275840759,
      "step": 2993
    },
    {
      "epoch": 2.829867674858223,
      "grad_norm": 0.48921511017073543,
      "learning_rate": 8.866139752341518e-07,
      "loss": 0.5262871980667114,
      "step": 2994
    },
    {
      "epoch": 2.8308128544423443,
      "grad_norm": 0.4659202618544926,
      "learning_rate": 8.768728562211947e-07,
      "loss": 0.4229423999786377,
      "step": 2995
    },
    {
      "epoch": 2.831758034026465,
      "grad_norm": 0.5168763412878847,
      "learning_rate": 8.67185072058363e-07,
      "loss": 0.4845583140850067,
      "step": 2996
    },
    {
      "epoch": 2.832703213610586,
      "grad_norm": 0.4400888161620784,
      "learning_rate": 8.575506332640593e-07,
      "loss": 0.4105900526046753,
      "step": 2997
    },
    {
      "epoch": 2.833648393194707,
      "grad_norm": 0.42412624979213354,
      "learning_rate": 8.47969550298755e-07,
      "loss": 0.4325970411300659,
      "step": 2998
    },
    {
      "epoch": 2.834593572778828,
      "grad_norm": 0.4395186702760115,
      "learning_rate": 8.384418335649958e-07,
      "loss": 0.4371253550052643,
      "step": 2999
    },
    {
      "epoch": 2.835538752362949,
      "grad_norm": 0.4494656873122901,
      "learning_rate": 8.289674934073844e-07,
      "loss": 0.4552825093269348,
      "step": 3000
    },
    {
      "epoch": 2.83648393194707,
      "grad_norm": 0.4749268860514407,
      "learning_rate": 8.195465401125813e-07,
      "loss": 0.4626786708831787,
      "step": 3001
    },
    {
      "epoch": 2.8374291115311907,
      "grad_norm": 0.44765143876571467,
      "learning_rate": 8.101789839092655e-07,
      "loss": 0.3720012903213501,
      "step": 3002
    },
    {
      "epoch": 2.838374291115312,
      "grad_norm": 0.4269751356607903,
      "learning_rate": 8.008648349681569e-07,
      "loss": 0.38515177369117737,
      "step": 3003
    },
    {
      "epoch": 2.839319470699433,
      "grad_norm": 0.41310132848787445,
      "learning_rate": 7.916041034019772e-07,
      "loss": 0.4204266667366028,
      "step": 3004
    },
    {
      "epoch": 2.840264650283554,
      "grad_norm": 0.45280027336628054,
      "learning_rate": 7.823967992654502e-07,
      "loss": 0.4035249352455139,
      "step": 3005
    },
    {
      "epoch": 2.841209829867675,
      "grad_norm": 0.4622739283048699,
      "learning_rate": 7.73242932555307e-07,
      "loss": 0.451436847448349,
      "step": 3006
    },
    {
      "epoch": 2.8421550094517958,
      "grad_norm": 0.44130357837158235,
      "learning_rate": 7.64142513210242e-07,
      "loss": 0.4266582131385803,
      "step": 3007
    },
    {
      "epoch": 2.843100189035917,
      "grad_norm": 0.5067700229002491,
      "learning_rate": 7.550955511109348e-07,
      "loss": 0.43929463624954224,
      "step": 3008
    },
    {
      "epoch": 2.844045368620038,
      "grad_norm": 0.5041733694566991,
      "learning_rate": 7.46102056080017e-07,
      "loss": 0.491469144821167,
      "step": 3009
    },
    {
      "epoch": 2.8449905482041586,
      "grad_norm": 0.4372041544175998,
      "learning_rate": 7.371620378820554e-07,
      "loss": 0.4228332042694092,
      "step": 3010
    },
    {
      "epoch": 2.8459357277882797,
      "grad_norm": 0.48785056309120545,
      "learning_rate": 7.282755062235746e-07,
      "loss": 0.42936480045318604,
      "step": 3011
    },
    {
      "epoch": 2.846880907372401,
      "grad_norm": 0.486449865696361,
      "learning_rate": 7.194424707530234e-07,
      "loss": 0.49455398321151733,
      "step": 3012
    },
    {
      "epoch": 2.8478260869565215,
      "grad_norm": 0.43253558248701496,
      "learning_rate": 7.106629410607691e-07,
      "loss": 0.4191705584526062,
      "step": 3013
    },
    {
      "epoch": 2.8487712665406426,
      "grad_norm": 0.44484078808258454,
      "learning_rate": 7.019369266790809e-07,
      "loss": 0.41612958908081055,
      "step": 3014
    },
    {
      "epoch": 2.8497164461247637,
      "grad_norm": 0.44140246302036756,
      "learning_rate": 6.932644370821085e-07,
      "loss": 0.41005632281303406,
      "step": 3015
    },
    {
      "epoch": 2.850661625708885,
      "grad_norm": 0.42976988342730166,
      "learning_rate": 6.84645481685925e-07,
      "loss": 0.41273486614227295,
      "step": 3016
    },
    {
      "epoch": 2.851606805293006,
      "grad_norm": 0.4595959882657995,
      "learning_rate": 6.760800698484448e-07,
      "loss": 0.43199387192726135,
      "step": 3017
    },
    {
      "epoch": 2.8525519848771266,
      "grad_norm": 0.43092391723683243,
      "learning_rate": 6.675682108694625e-07,
      "loss": 0.39245736598968506,
      "step": 3018
    },
    {
      "epoch": 2.8534971644612477,
      "grad_norm": 0.4368588774153899,
      "learning_rate": 6.59109913990641e-07,
      "loss": 0.4291701316833496,
      "step": 3019
    },
    {
      "epoch": 2.854442344045369,
      "grad_norm": 0.3931603214551327,
      "learning_rate": 6.507051883954618e-07,
      "loss": 0.43118786811828613,
      "step": 3020
    },
    {
      "epoch": 2.8553875236294894,
      "grad_norm": 0.4479280144035679,
      "learning_rate": 6.42354043209259e-07,
      "loss": 0.43895918130874634,
      "step": 3021
    },
    {
      "epoch": 2.8563327032136105,
      "grad_norm": 0.47072147740224657,
      "learning_rate": 6.340564874991905e-07,
      "loss": 0.4550941288471222,
      "step": 3022
    },
    {
      "epoch": 2.8572778827977316,
      "grad_norm": 0.41455258308638465,
      "learning_rate": 6.258125302742279e-07,
      "loss": 0.42883431911468506,
      "step": 3023
    },
    {
      "epoch": 2.8582230623818523,
      "grad_norm": 0.4612545693749443,
      "learning_rate": 6.176221804851501e-07,
      "loss": 0.44481369853019714,
      "step": 3024
    },
    {
      "epoch": 2.8591682419659734,
      "grad_norm": 0.4410849920534078,
      "learning_rate": 6.094854470245326e-07,
      "loss": 0.3964496850967407,
      "step": 3025
    },
    {
      "epoch": 2.8601134215500945,
      "grad_norm": 0.49024831280071324,
      "learning_rate": 6.01402338726731e-07,
      "loss": 0.46829700469970703,
      "step": 3026
    },
    {
      "epoch": 2.8610586011342156,
      "grad_norm": 0.44006311493943634,
      "learning_rate": 5.933728643678915e-07,
      "loss": 0.4393198490142822,
      "step": 3027
    },
    {
      "epoch": 2.8620037807183367,
      "grad_norm": 0.47906943519152273,
      "learning_rate": 5.853970326659186e-07,
      "loss": 0.420960009098053,
      "step": 3028
    },
    {
      "epoch": 2.8629489603024574,
      "grad_norm": 0.4441043569103324,
      "learning_rate": 5.774748522804685e-07,
      "loss": 0.3811356723308563,
      "step": 3029
    },
    {
      "epoch": 2.8638941398865785,
      "grad_norm": 0.4713157355828422,
      "learning_rate": 5.696063318129663e-07,
      "loss": 0.45706257224082947,
      "step": 3030
    },
    {
      "epoch": 2.8648393194706996,
      "grad_norm": 0.4129080595263492,
      "learning_rate": 5.617914798065616e-07,
      "loss": 0.3878282308578491,
      "step": 3031
    },
    {
      "epoch": 2.8657844990548202,
      "grad_norm": 0.4052619580402437,
      "learning_rate": 5.540303047461282e-07,
      "loss": 0.4188312888145447,
      "step": 3032
    },
    {
      "epoch": 2.8667296786389413,
      "grad_norm": 0.4523174091265258,
      "learning_rate": 5.463228150582811e-07,
      "loss": 0.445027232170105,
      "step": 3033
    },
    {
      "epoch": 2.8676748582230625,
      "grad_norm": 0.4275029161286434,
      "learning_rate": 5.386690191113209e-07,
      "loss": 0.3741629123687744,
      "step": 3034
    },
    {
      "epoch": 2.868620037807183,
      "grad_norm": 0.3846485571487627,
      "learning_rate": 5.310689252152834e-07,
      "loss": 0.38979780673980713,
      "step": 3035
    },
    {
      "epoch": 2.869565217391304,
      "grad_norm": 0.4754244080573572,
      "learning_rate": 5.235225416218626e-07,
      "loss": 0.4063766598701477,
      "step": 3036
    },
    {
      "epoch": 2.8705103969754253,
      "grad_norm": 0.42667103485973956,
      "learning_rate": 5.160298765244709e-07,
      "loss": 0.45556774735450745,
      "step": 3037
    },
    {
      "epoch": 2.8714555765595464,
      "grad_norm": 0.4732579021487559,
      "learning_rate": 5.085909380581677e-07,
      "loss": 0.4343782663345337,
      "step": 3038
    },
    {
      "epoch": 2.8724007561436675,
      "grad_norm": 0.39327242275859564,
      "learning_rate": 5.012057342996978e-07,
      "loss": 0.37194591760635376,
      "step": 3039
    },
    {
      "epoch": 2.873345935727788,
      "grad_norm": 0.4168698059279024,
      "learning_rate": 4.938742732674529e-07,
      "loss": 0.37049397826194763,
      "step": 3040
    },
    {
      "epoch": 2.8742911153119093,
      "grad_norm": 0.4515876676104429,
      "learning_rate": 4.865965629214819e-07,
      "loss": 0.433299720287323,
      "step": 3041
    },
    {
      "epoch": 2.8752362948960304,
      "grad_norm": 0.47419387430698584,
      "learning_rate": 4.79372611163481e-07,
      "loss": 0.4486941397190094,
      "step": 3042
    },
    {
      "epoch": 2.876181474480151,
      "grad_norm": 0.42372752372528566,
      "learning_rate": 4.7220242583675967e-07,
      "loss": 0.4198722243309021,
      "step": 3043
    },
    {
      "epoch": 2.877126654064272,
      "grad_norm": 0.4589892660235846,
      "learning_rate": 4.65086014726257e-07,
      "loss": 0.44702595472335815,
      "step": 3044
    },
    {
      "epoch": 2.8780718336483933,
      "grad_norm": 0.421818021014443,
      "learning_rate": 4.5802338555854254e-07,
      "loss": 0.40658804774284363,
      "step": 3045
    },
    {
      "epoch": 2.879017013232514,
      "grad_norm": 0.4498048104893387,
      "learning_rate": 4.5101454600177674e-07,
      "loss": 0.44114601612091064,
      "step": 3046
    },
    {
      "epoch": 2.879962192816635,
      "grad_norm": 0.40157422106372354,
      "learning_rate": 4.44059503665728e-07,
      "loss": 0.41818320751190186,
      "step": 3047
    },
    {
      "epoch": 2.880907372400756,
      "grad_norm": 0.48601012477281286,
      "learning_rate": 4.3715826610175035e-07,
      "loss": 0.4970472455024719,
      "step": 3048
    },
    {
      "epoch": 2.881852551984877,
      "grad_norm": 0.5127399002571698,
      "learning_rate": 4.303108408027667e-07,
      "loss": 0.47684359550476074,
      "step": 3049
    },
    {
      "epoch": 2.8827977315689983,
      "grad_norm": 0.4515662832215411,
      "learning_rate": 4.235172352033023e-07,
      "loss": 0.39280790090560913,
      "step": 3050
    },
    {
      "epoch": 2.883742911153119,
      "grad_norm": 0.48256475998620274,
      "learning_rate": 4.1677745667942915e-07,
      "loss": 0.4364182949066162,
      "step": 3051
    },
    {
      "epoch": 2.88468809073724,
      "grad_norm": 0.4351376118864178,
      "learning_rate": 4.1009151254878256e-07,
      "loss": 0.4629242420196533,
      "step": 3052
    },
    {
      "epoch": 2.885633270321361,
      "grad_norm": 0.4384956717765726,
      "learning_rate": 4.034594100705447e-07,
      "loss": 0.40346041321754456,
      "step": 3053
    },
    {
      "epoch": 2.886578449905482,
      "grad_norm": 0.401103127228462,
      "learning_rate": 3.968811564454444e-07,
      "loss": 0.37023717164993286,
      "step": 3054
    },
    {
      "epoch": 2.887523629489603,
      "grad_norm": 0.48937560206882924,
      "learning_rate": 3.903567588157353e-07,
      "loss": 0.4392241835594177,
      "step": 3055
    },
    {
      "epoch": 2.888468809073724,
      "grad_norm": 0.4901522585662055,
      "learning_rate": 3.8388622426521194e-07,
      "loss": 0.4398655295372009,
      "step": 3056
    },
    {
      "epoch": 2.889413988657845,
      "grad_norm": 0.4264910150535971,
      "learning_rate": 3.774695598191769e-07,
      "loss": 0.41714346408843994,
      "step": 3057
    },
    {
      "epoch": 2.890359168241966,
      "grad_norm": 0.45588178839544535,
      "learning_rate": 3.7110677244445167e-07,
      "loss": 0.3962968587875366,
      "step": 3058
    },
    {
      "epoch": 2.891304347826087,
      "grad_norm": 0.472419871134126,
      "learning_rate": 3.647978690493492e-07,
      "loss": 0.4824625551700592,
      "step": 3059
    },
    {
      "epoch": 2.892249527410208,
      "grad_norm": 0.4678477272265469,
      "learning_rate": 3.585428564836957e-07,
      "loss": 0.40426814556121826,
      "step": 3060
    },
    {
      "epoch": 2.893194706994329,
      "grad_norm": 0.4447673176262239,
      "learning_rate": 3.523417415387864e-07,
      "loss": 0.4565306305885315,
      "step": 3061
    },
    {
      "epoch": 2.89413988657845,
      "grad_norm": 0.44516582121096104,
      "learning_rate": 3.4619453094740797e-07,
      "loss": 0.43773311376571655,
      "step": 3062
    },
    {
      "epoch": 2.895085066162571,
      "grad_norm": 0.41063035301801487,
      "learning_rate": 3.4010123138383256e-07,
      "loss": 0.3699529767036438,
      "step": 3063
    },
    {
      "epoch": 2.896030245746692,
      "grad_norm": 0.4584847665419475,
      "learning_rate": 3.340618494637793e-07,
      "loss": 0.44485020637512207,
      "step": 3064
    },
    {
      "epoch": 2.8969754253308126,
      "grad_norm": 0.42067015887101766,
      "learning_rate": 3.280763917444363e-07,
      "loss": 0.4254212975502014,
      "step": 3065
    },
    {
      "epoch": 2.8979206049149338,
      "grad_norm": 0.4461674181661373,
      "learning_rate": 3.2214486472444405e-07,
      "loss": 0.43201199173927307,
      "step": 3066
    },
    {
      "epoch": 2.898865784499055,
      "grad_norm": 0.4016486542516848,
      "learning_rate": 3.162672748438844e-07,
      "loss": 0.4052746891975403,
      "step": 3067
    },
    {
      "epoch": 2.899810964083176,
      "grad_norm": 0.3994553478221251,
      "learning_rate": 3.1044362848428044e-07,
      "loss": 0.4170307517051697,
      "step": 3068
    },
    {
      "epoch": 2.900756143667297,
      "grad_norm": 0.4200676707355208,
      "learning_rate": 3.0467393196859653e-07,
      "loss": 0.3982323110103607,
      "step": 3069
    },
    {
      "epoch": 2.9017013232514177,
      "grad_norm": 0.42588859609373886,
      "learning_rate": 2.9895819156119943e-07,
      "loss": 0.41938671469688416,
      "step": 3070
    },
    {
      "epoch": 2.902646502835539,
      "grad_norm": 0.4748650267733835,
      "learning_rate": 2.932964134678917e-07,
      "loss": 0.45696982741355896,
      "step": 3071
    },
    {
      "epoch": 2.90359168241966,
      "grad_norm": 0.4438307245945971,
      "learning_rate": 2.876886038358839e-07,
      "loss": 0.43934303522109985,
      "step": 3072
    },
    {
      "epoch": 2.9045368620037806,
      "grad_norm": 0.44344706460530214,
      "learning_rate": 2.8213476875378344e-07,
      "loss": 0.4322962760925293,
      "step": 3073
    },
    {
      "epoch": 2.9054820415879017,
      "grad_norm": 0.4725251001831651,
      "learning_rate": 2.7663491425160007e-07,
      "loss": 0.3998124599456787,
      "step": 3074
    },
    {
      "epoch": 2.906427221172023,
      "grad_norm": 0.4752855275308823,
      "learning_rate": 2.711890463007405e-07,
      "loss": 0.40755152702331543,
      "step": 3075
    },
    {
      "epoch": 2.9073724007561434,
      "grad_norm": 0.44168943704830077,
      "learning_rate": 2.657971708139917e-07,
      "loss": 0.3954423666000366,
      "step": 3076
    },
    {
      "epoch": 2.9083175803402646,
      "grad_norm": 0.4399693097630317,
      "learning_rate": 2.6045929364551525e-07,
      "loss": 0.40786001086235046,
      "step": 3077
    },
    {
      "epoch": 2.9092627599243857,
      "grad_norm": 0.4467496993144796,
      "learning_rate": 2.551754205908419e-07,
      "loss": 0.4639531373977661,
      "step": 3078
    },
    {
      "epoch": 2.9102079395085068,
      "grad_norm": 0.4639217609112255,
      "learning_rate": 2.499455573868825e-07,
      "loss": 0.45689886808395386,
      "step": 3079
    },
    {
      "epoch": 2.911153119092628,
      "grad_norm": 0.40104859518898195,
      "learning_rate": 2.447697097118951e-07,
      "loss": 0.3699387311935425,
      "step": 3080
    },
    {
      "epoch": 2.9120982986767485,
      "grad_norm": 0.4065506688098619,
      "learning_rate": 2.3964788318549535e-07,
      "loss": 0.40650197863578796,
      "step": 3081
    },
    {
      "epoch": 2.9130434782608696,
      "grad_norm": 0.4688514582756848,
      "learning_rate": 2.3458008336864622e-07,
      "loss": 0.4448573589324951,
      "step": 3082
    },
    {
      "epoch": 2.9139886578449907,
      "grad_norm": 0.4728528049183167,
      "learning_rate": 2.295663157636463e-07,
      "loss": 0.44780558347702026,
      "step": 3083
    },
    {
      "epoch": 2.9149338374291114,
      "grad_norm": 0.4789969678690829,
      "learning_rate": 2.2460658581414107e-07,
      "loss": 0.40694481134414673,
      "step": 3084
    },
    {
      "epoch": 2.9158790170132325,
      "grad_norm": 0.4248048449467331,
      "learning_rate": 2.1970089890509527e-07,
      "loss": 0.3625985085964203,
      "step": 3085
    },
    {
      "epoch": 2.9168241965973536,
      "grad_norm": 0.4633464320201849,
      "learning_rate": 2.1484926036279828e-07,
      "loss": 0.441043496131897,
      "step": 3086
    },
    {
      "epoch": 2.9177693761814743,
      "grad_norm": 0.4611698514591132,
      "learning_rate": 2.1005167545486426e-07,
      "loss": 0.43370240926742554,
      "step": 3087
    },
    {
      "epoch": 2.9187145557655954,
      "grad_norm": 0.4871748860782549,
      "learning_rate": 2.0530814939020427e-07,
      "loss": 0.4575813114643097,
      "step": 3088
    },
    {
      "epoch": 2.9196597353497165,
      "grad_norm": 0.495538527044584,
      "learning_rate": 2.0061868731905408e-07,
      "loss": 0.49632716178894043,
      "step": 3089
    },
    {
      "epoch": 2.9206049149338376,
      "grad_norm": 0.48668493876023666,
      "learning_rate": 1.9598329433293538e-07,
      "loss": 0.4182392656803131,
      "step": 3090
    },
    {
      "epoch": 2.9215500945179587,
      "grad_norm": 0.3999829653062112,
      "learning_rate": 1.9140197546467785e-07,
      "loss": 0.41473904252052307,
      "step": 3091
    },
    {
      "epoch": 2.9224952741020793,
      "grad_norm": 0.40594326462368857,
      "learning_rate": 1.8687473568839154e-07,
      "loss": 0.38366347551345825,
      "step": 3092
    },
    {
      "epoch": 2.9234404536862004,
      "grad_norm": 0.41935380511172393,
      "learning_rate": 1.824015799194667e-07,
      "loss": 0.3733406066894531,
      "step": 3093
    },
    {
      "epoch": 2.9243856332703215,
      "grad_norm": 0.4933256850819389,
      "learning_rate": 1.7798251301458513e-07,
      "loss": 0.43504273891448975,
      "step": 3094
    },
    {
      "epoch": 2.925330812854442,
      "grad_norm": 0.4677906584026489,
      "learning_rate": 1.7361753977169215e-07,
      "loss": 0.48186221718788147,
      "step": 3095
    },
    {
      "epoch": 2.9262759924385633,
      "grad_norm": 0.411072791201186,
      "learning_rate": 1.6930666493000792e-07,
      "loss": 0.421378493309021,
      "step": 3096
    },
    {
      "epoch": 2.9272211720226844,
      "grad_norm": 0.4315832307567673,
      "learning_rate": 1.6504989317001618e-07,
      "loss": 0.4407663643360138,
      "step": 3097
    },
    {
      "epoch": 2.928166351606805,
      "grad_norm": 0.4415577917325542,
      "learning_rate": 1.6084722911344775e-07,
      "loss": 0.42365315556526184,
      "step": 3098
    },
    {
      "epoch": 2.929111531190926,
      "grad_norm": 0.4108202987538003,
      "learning_rate": 1.5669867732330812e-07,
      "loss": 0.41086888313293457,
      "step": 3099
    },
    {
      "epoch": 2.9300567107750473,
      "grad_norm": 0.454140106572278,
      "learning_rate": 1.5260424230382763e-07,
      "loss": 0.4612917900085449,
      "step": 3100
    },
    {
      "epoch": 2.9310018903591684,
      "grad_norm": 0.3932501411292024,
      "learning_rate": 1.4856392850048917e-07,
      "loss": 0.3924476206302643,
      "step": 3101
    },
    {
      "epoch": 2.9319470699432895,
      "grad_norm": 0.47095731373250976,
      "learning_rate": 1.4457774030002259e-07,
      "loss": 0.4596758782863617,
      "step": 3102
    },
    {
      "epoch": 2.93289224952741,
      "grad_norm": 0.4925689011992791,
      "learning_rate": 1.4064568203037697e-07,
      "loss": 0.4137478470802307,
      "step": 3103
    },
    {
      "epoch": 2.9338374291115312,
      "grad_norm": 0.45704646769205354,
      "learning_rate": 1.3676775796074293e-07,
      "loss": 0.4531920552253723,
      "step": 3104
    },
    {
      "epoch": 2.9347826086956523,
      "grad_norm": 0.4650286316390841,
      "learning_rate": 1.3294397230153577e-07,
      "loss": 0.4688876271247864,
      "step": 3105
    },
    {
      "epoch": 2.935727788279773,
      "grad_norm": 0.406270967023074,
      "learning_rate": 1.2917432920437345e-07,
      "loss": 0.4078441262245178,
      "step": 3106
    },
    {
      "epoch": 2.936672967863894,
      "grad_norm": 0.4693429444490565,
      "learning_rate": 1.2545883276210425e-07,
      "loss": 0.4607255160808563,
      "step": 3107
    },
    {
      "epoch": 2.937618147448015,
      "grad_norm": 0.4687286479474011,
      "learning_rate": 1.2179748700879012e-07,
      "loss": 0.40267786383628845,
      "step": 3108
    },
    {
      "epoch": 2.938563327032136,
      "grad_norm": 0.44977162639638396,
      "learning_rate": 1.1819029591968455e-07,
      "loss": 0.41337093710899353,
      "step": 3109
    },
    {
      "epoch": 2.939508506616257,
      "grad_norm": 0.40285391245022206,
      "learning_rate": 1.1463726341126025e-07,
      "loss": 0.38492023944854736,
      "step": 3110
    },
    {
      "epoch": 2.940453686200378,
      "grad_norm": 0.40110494184617107,
      "learning_rate": 1.1113839334117581e-07,
      "loss": 0.39250093698501587,
      "step": 3111
    },
    {
      "epoch": 2.941398865784499,
      "grad_norm": 0.3988536283814516,
      "learning_rate": 1.076936895082925e-07,
      "loss": 0.37680113315582275,
      "step": 3112
    },
    {
      "epoch": 2.9423440453686203,
      "grad_norm": 0.39263962602162333,
      "learning_rate": 1.0430315565264636e-07,
      "loss": 0.38656002283096313,
      "step": 3113
    },
    {
      "epoch": 2.943289224952741,
      "grad_norm": 0.443520390924849,
      "learning_rate": 1.009667954554816e-07,
      "loss": 0.41770216822624207,
      "step": 3114
    },
    {
      "epoch": 2.944234404536862,
      "grad_norm": 0.5182867507884004,
      "learning_rate": 9.768461253920614e-08,
      "loss": 0.5051183700561523,
      "step": 3115
    },
    {
      "epoch": 2.945179584120983,
      "grad_norm": 0.43574736743269193,
      "learning_rate": 9.445661046740273e-08,
      "loss": 0.4298206567764282,
      "step": 3116
    },
    {
      "epoch": 2.946124763705104,
      "grad_norm": 0.4665558762528446,
      "learning_rate": 9.128279274485119e-08,
      "loss": 0.4080411195755005,
      "step": 3117
    },
    {
      "epoch": 2.947069943289225,
      "grad_norm": 0.45676290067645403,
      "learning_rate": 8.816316281747838e-08,
      "loss": 0.4332302212715149,
      "step": 3118
    },
    {
      "epoch": 2.948015122873346,
      "grad_norm": 0.4787557295632943,
      "learning_rate": 8.509772407238048e-08,
      "loss": 0.45118099451065063,
      "step": 3119
    },
    {
      "epoch": 2.9489603024574667,
      "grad_norm": 0.5029255063959912,
      "learning_rate": 8.208647983782847e-08,
      "loss": 0.4123859405517578,
      "step": 3120
    },
    {
      "epoch": 2.9499054820415878,
      "grad_norm": 0.43619390946134495,
      "learning_rate": 7.912943338324596e-08,
      "loss": 0.39408043026924133,
      "step": 3121
    },
    {
      "epoch": 2.950850661625709,
      "grad_norm": 0.41743181860574896,
      "learning_rate": 7.622658791920923e-08,
      "loss": 0.4379613399505615,
      "step": 3122
    },
    {
      "epoch": 2.95179584120983,
      "grad_norm": 0.4865703507802435,
      "learning_rate": 7.337794659744157e-08,
      "loss": 0.4526894688606262,
      "step": 3123
    },
    {
      "epoch": 2.952741020793951,
      "grad_norm": 0.46931809192460666,
      "learning_rate": 7.058351251083007e-08,
      "loss": 0.3996427059173584,
      "step": 3124
    },
    {
      "epoch": 2.9536862003780717,
      "grad_norm": 0.45281087804630116,
      "learning_rate": 6.784328869339218e-08,
      "loss": 0.41968482732772827,
      "step": 3125
    },
    {
      "epoch": 2.954631379962193,
      "grad_norm": 0.4337852785666559,
      "learning_rate": 6.515727812029804e-08,
      "loss": 0.39995962381362915,
      "step": 3126
    },
    {
      "epoch": 2.955576559546314,
      "grad_norm": 0.47443642285020593,
      "learning_rate": 6.252548370784817e-08,
      "loss": 0.49917441606521606,
      "step": 3127
    },
    {
      "epoch": 2.9565217391304346,
      "grad_norm": 0.4227740671366921,
      "learning_rate": 5.994790831348463e-08,
      "loss": 0.4265303611755371,
      "step": 3128
    },
    {
      "epoch": 2.9574669187145557,
      "grad_norm": 0.5308675216327835,
      "learning_rate": 5.74245547357799e-08,
      "loss": 0.4827321767807007,
      "step": 3129
    },
    {
      "epoch": 2.958412098298677,
      "grad_norm": 0.43382186223065483,
      "learning_rate": 5.4955425714431353e-08,
      "loss": 0.40003910660743713,
      "step": 3130
    },
    {
      "epoch": 2.9593572778827975,
      "grad_norm": 0.3968643397983853,
      "learning_rate": 5.254052393026676e-08,
      "loss": 0.3684311807155609,
      "step": 3131
    },
    {
      "epoch": 2.9603024574669186,
      "grad_norm": 0.4567056241489821,
      "learning_rate": 5.017985200523878e-08,
      "loss": 0.39223241806030273,
      "step": 3132
    },
    {
      "epoch": 2.9612476370510397,
      "grad_norm": 0.41986064101860354,
      "learning_rate": 4.787341250241384e-08,
      "loss": 0.4447387158870697,
      "step": 3133
    },
    {
      "epoch": 2.9621928166351608,
      "grad_norm": 0.4254605362176183,
      "learning_rate": 4.5621207925994336e-08,
      "loss": 0.3566884994506836,
      "step": 3134
    },
    {
      "epoch": 2.963137996219282,
      "grad_norm": 0.46355452826823673,
      "learning_rate": 4.3423240721268686e-08,
      "loss": 0.4191737771034241,
      "step": 3135
    },
    {
      "epoch": 2.9640831758034025,
      "grad_norm": 0.4733928125716045,
      "learning_rate": 4.1279513274666834e-08,
      "loss": 0.4649035632610321,
      "step": 3136
    },
    {
      "epoch": 2.9650283553875236,
      "grad_norm": 0.45304034741719207,
      "learning_rate": 3.919002791370474e-08,
      "loss": 0.40910306572914124,
      "step": 3137
    },
    {
      "epoch": 2.9659735349716447,
      "grad_norm": 0.5193408053610964,
      "learning_rate": 3.715478690702323e-08,
      "loss": 0.45158687233924866,
      "step": 3138
    },
    {
      "epoch": 2.9669187145557654,
      "grad_norm": 0.44554896827302826,
      "learning_rate": 3.517379246436026e-08,
      "loss": 0.41955530643463135,
      "step": 3139
    },
    {
      "epoch": 2.9678638941398865,
      "grad_norm": 0.4484963530774763,
      "learning_rate": 3.324704673655088e-08,
      "loss": 0.38138699531555176,
      "step": 3140
    },
    {
      "epoch": 2.9688090737240076,
      "grad_norm": 0.4833738015702413,
      "learning_rate": 3.137455181554394e-08,
      "loss": 0.4660564661026001,
      "step": 3141
    },
    {
      "epoch": 2.9697542533081287,
      "grad_norm": 0.4873133184674293,
      "learning_rate": 2.955630973437429e-08,
      "loss": 0.4716896116733551,
      "step": 3142
    },
    {
      "epoch": 2.97069943289225,
      "grad_norm": 0.40925677673965205,
      "learning_rate": 2.7792322467173893e-08,
      "loss": 0.4102913737297058,
      "step": 3143
    },
    {
      "epoch": 2.9716446124763705,
      "grad_norm": 0.4041433117280831,
      "learning_rate": 2.608259192917739e-08,
      "loss": 0.3886527419090271,
      "step": 3144
    },
    {
      "epoch": 2.9725897920604916,
      "grad_norm": 0.4287198600026296,
      "learning_rate": 2.442711997670544e-08,
      "loss": 0.4287281632423401,
      "step": 3145
    },
    {
      "epoch": 2.9735349716446127,
      "grad_norm": 0.40922551776418414,
      "learning_rate": 2.2825908407159148e-08,
      "loss": 0.4343321621417999,
      "step": 3146
    },
    {
      "epoch": 2.9744801512287333,
      "grad_norm": 0.4387595404000053,
      "learning_rate": 2.1278958959036754e-08,
      "loss": 0.4652763605117798,
      "step": 3147
    },
    {
      "epoch": 2.9754253308128544,
      "grad_norm": 0.4604397629435243,
      "learning_rate": 1.9786273311928062e-08,
      "loss": 0.41343867778778076,
      "step": 3148
    },
    {
      "epoch": 2.9763705103969755,
      "grad_norm": 0.44748669600052293,
      "learning_rate": 1.8347853086492228e-08,
      "loss": 0.4194350242614746,
      "step": 3149
    },
    {
      "epoch": 2.977315689981096,
      "grad_norm": 0.4070661207194153,
      "learning_rate": 1.6963699844474434e-08,
      "loss": 0.39647412300109863,
      "step": 3150
    },
    {
      "epoch": 2.9782608695652173,
      "grad_norm": 0.46712820017341344,
      "learning_rate": 1.5633815088705873e-08,
      "loss": 0.42024651169776917,
      "step": 3151
    },
    {
      "epoch": 2.9792060491493384,
      "grad_norm": 0.4792827259915058,
      "learning_rate": 1.4358200263092647e-08,
      "loss": 0.4347875118255615,
      "step": 3152
    },
    {
      "epoch": 2.9801512287334595,
      "grad_norm": 0.48926068746770274,
      "learning_rate": 1.3136856752621329e-08,
      "loss": 0.49222856760025024,
      "step": 3153
    },
    {
      "epoch": 2.9810964083175806,
      "grad_norm": 0.44245140498556773,
      "learning_rate": 1.196978588334785e-08,
      "loss": 0.46418362855911255,
      "step": 3154
    },
    {
      "epoch": 2.9820415879017013,
      "grad_norm": 0.46286471396736056,
      "learning_rate": 1.0856988922403056e-08,
      "loss": 0.42722827196121216,
      "step": 3155
    },
    {
      "epoch": 2.9829867674858224,
      "grad_norm": 0.45923206823970436,
      "learning_rate": 9.798467077998252e-09,
      "loss": 0.45172232389450073,
      "step": 3156
    },
    {
      "epoch": 2.9839319470699435,
      "grad_norm": 0.5357439773549312,
      "learning_rate": 8.794221499408561e-09,
      "loss": 0.530500590801239,
      "step": 3157
    },
    {
      "epoch": 2.984877126654064,
      "grad_norm": 0.42049055640370253,
      "learning_rate": 7.844253276978464e-09,
      "loss": 0.40834999084472656,
      "step": 3158
    },
    {
      "epoch": 2.9858223062381852,
      "grad_norm": 0.42928952909912477,
      "learning_rate": 6.948563442132905e-09,
      "loss": 0.39962679147720337,
      "step": 3159
    },
    {
      "epoch": 2.9867674858223063,
      "grad_norm": 0.42737506877946446,
      "learning_rate": 6.107152967349539e-09,
      "loss": 0.4035497307777405,
      "step": 3160
    },
    {
      "epoch": 2.987712665406427,
      "grad_norm": 0.4685598142087859,
      "learning_rate": 5.320022766180932e-09,
      "loss": 0.4788234233856201,
      "step": 3161
    },
    {
      "epoch": 2.988657844990548,
      "grad_norm": 0.3976405593779703,
      "learning_rate": 4.587173693249014e-09,
      "loss": 0.44273799657821655,
      "step": 3162
    },
    {
      "epoch": 2.989603024574669,
      "grad_norm": 0.46201720165779103,
      "learning_rate": 3.908606544228422e-09,
      "loss": 0.46272099018096924,
      "step": 3163
    },
    {
      "epoch": 2.9905482041587903,
      "grad_norm": 0.4477663415853195,
      "learning_rate": 3.284322055868705e-09,
      "loss": 0.408573716878891,
      "step": 3164
    },
    {
      "epoch": 2.9914933837429114,
      "grad_norm": 0.45133041587722345,
      "learning_rate": 2.714320905977674e-09,
      "loss": 0.44013702869415283,
      "step": 3165
    },
    {
      "epoch": 2.992438563327032,
      "grad_norm": 0.4318182678775036,
      "learning_rate": 2.198603713432501e-09,
      "loss": 0.42172056436538696,
      "step": 3166
    },
    {
      "epoch": 2.993383742911153,
      "grad_norm": 0.45961341710893594,
      "learning_rate": 1.7371710381630657e-09,
      "loss": 0.4547739624977112,
      "step": 3167
    },
    {
      "epoch": 2.9943289224952743,
      "grad_norm": 0.4522178837408639,
      "learning_rate": 1.3300233811686103e-09,
      "loss": 0.42141491174697876,
      "step": 3168
    },
    {
      "epoch": 2.995274102079395,
      "grad_norm": 0.4029001038562794,
      "learning_rate": 9.771611844955342e-10,
      "loss": 0.3889071047306061,
      "step": 3169
    },
    {
      "epoch": 2.996219281663516,
      "grad_norm": 0.5559060019874665,
      "learning_rate": 6.785848312707011e-10,
      "loss": 0.5548129081726074,
      "step": 3170
    },
    {
      "epoch": 2.997164461247637,
      "grad_norm": 0.5027642829166534,
      "learning_rate": 4.3429464566258074e-10,
      "loss": 0.4608549475669861,
      "step": 3171
    },
    {
      "epoch": 2.998109640831758,
      "grad_norm": 0.48411890275516684,
      "learning_rate": 2.4429089290345376e-10,
      "loss": 0.42092782258987427,
      "step": 3172
    },
    {
      "epoch": 2.999054820415879,
      "grad_norm": 0.4539786741862714,
      "learning_rate": 1.085737792949626e-10,
      "loss": 0.4265294671058655,
      "step": 3173
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.49428467423273154,
      "learning_rate": 2.7143452191458196e-11,
      "loss": 0.4809390902519226,
      "step": 3174
    },
    {
      "epoch": 3.0,
      "step": 3174,
      "total_flos": 3870553987219456.0,
      "train_loss": 0.3277500697330025,
      "train_runtime": 39472.6051,
      "train_samples_per_second": 1.287,
      "train_steps_per_second": 0.08
    }
  ],
  "logging_steps": 1,
  "max_steps": 3174,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3870553987219456.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}