diff --git "a/checkpoint-1000/trainer_state.json" "b/checkpoint-1000/trainer_state.json"
--- "a/checkpoint-1000/trainer_state.json"
+++ "b/checkpoint-1000/trainer_state.json"
@@ -10,7002 +10,7002 @@
   "log_history": [
     {
       "epoch": 0.001000281329123816,
-      "grad_norm": 7.503077030181885,
+      "grad_norm": 5.902005195617676,
       "learning_rate": 0.0001999999450590425,
-      "loss": 3.6513,
+      "loss": 3.1875,
       "step": 1
     },
     {
       "epoch": 0.002000562658247632,
-      "grad_norm": 3.1191189289093018,
+      "grad_norm": 3.2577760219573975,
       "learning_rate": 0.00019999978023623033,
-      "loss": 2.8683,
+      "loss": 2.3666,
       "step": 2
     },
     {
       "epoch": 0.003000843987371448,
-      "grad_norm": 1.9282511472702026,
+      "grad_norm": 5.3700995445251465,
       "learning_rate": 0.0001999995055317446,
-      "loss": 2.7882,
+      "loss": 2.8282,
       "step": 3
     },
     {
       "epoch": 0.004001125316495264,
-      "grad_norm": 1.726026177406311,
+      "grad_norm": 2.1445534229278564,
       "learning_rate": 0.00019999912094588717,
-      "loss": 2.7716,
+      "loss": 2.2322,
       "step": 4
     },
     {
       "epoch": 0.005001406645619081,
-      "grad_norm": 1.4632996320724487,
+      "grad_norm": 1.5143821239471436,
       "learning_rate": 0.00019999862647908064,
-      "loss": 2.6869,
+      "loss": 2.1709,
       "step": 5
     },
     {
       "epoch": 0.006001687974742896,
-      "grad_norm": 1.5544543266296387,
+      "grad_norm": 2.0491714477539062,
       "learning_rate": 0.00019999802213186834,
-      "loss": 2.6952,
+      "loss": 2.2863,
       "step": 6
     },
     {
       "epoch": 0.007001969303866712,
-      "grad_norm": 1.5888980627059937,
+      "grad_norm": 1.2016857862472534,
       "learning_rate": 0.0001999973079049143,
-      "loss": 2.1237,
+      "loss": 1.5595,
       "step": 7
     },
     {
       "epoch": 0.008002250632990529,
-      "grad_norm": 1.8750641345977783,
+      "grad_norm": 1.3860406875610352,
       "learning_rate": 0.00019999648379900338,
-      "loss": 2.3376,
+      "loss": 1.7264,
       "step": 8
     },
     {
       "epoch": 0.009002531962114344,
-      "grad_norm": 1.0540648698806763,
+      "grad_norm": 1.0861930847167969,
       "learning_rate": 0.0001999955498150411,
-      "loss": 2.4896,
+      "loss": 2.0533,
       "step": 9
     },
     {
       "epoch": 0.010002813291238161,
-      "grad_norm": 1.0269274711608887,
+      "grad_norm": 2.233243703842163,
       "learning_rate": 0.00019999450595405374,
-      "loss": 2.1365,
+      "loss": 1.9378,
       "step": 10
     },
     {
       "epoch": 0.011003094620361977,
-      "grad_norm": 1.0851730108261108,
+      "grad_norm": 1.302808165550232,
       "learning_rate": 0.0001999933522171883,
-      "loss": 2.235,
+      "loss": 1.9182,
       "step": 11
     },
     {
       "epoch": 0.012003375949485792,
-      "grad_norm": 0.927042543888092,
+      "grad_norm": 0.8285257816314697,
       "learning_rate": 0.00019999208860571255,
-      "loss": 2.2438,
+      "loss": 1.9146,
       "step": 12
     },
     {
       "epoch": 0.01300365727860961,
-      "grad_norm": 1.3729208707809448,
+      "grad_norm": 1.2248319387435913,
       "learning_rate": 0.00019999071512101496,
-      "loss": 2.0845,
+      "loss": 1.7467,
       "step": 13
     },
     {
       "epoch": 0.014003938607733425,
-      "grad_norm": 1.1325910091400146,
+      "grad_norm": 0.8307135105133057,
       "learning_rate": 0.00019998923176460474,
-      "loss": 2.0668,
+      "loss": 1.6896,
       "step": 14
     },
     {
       "epoch": 0.01500421993685724,
-      "grad_norm": 0.9290457367897034,
+      "grad_norm": 1.1531301736831665,
       "learning_rate": 0.00019998763853811184,
-      "loss": 2.0227,
+      "loss": 1.7549,
       "step": 15
     },
     {
       "epoch": 0.016004501265981057,
-      "grad_norm": 0.942140519618988,
+      "grad_norm": 1.0071958303451538,
       "learning_rate": 0.00019998593544328692,
-      "loss": 2.1598,
+      "loss": 1.903,
       "step": 16
     },
     {
       "epoch": 0.017004782595104875,
-      "grad_norm": 1.096635103225708,
+      "grad_norm": 0.9111937284469604,
       "learning_rate": 0.00019998412248200138,
-      "loss": 2.1897,
+      "loss": 1.8372,
       "step": 17
     },
     {
       "epoch": 0.01800506392422869,
-      "grad_norm": 1.1107186079025269,
+      "grad_norm": 0.9943836331367493,
       "learning_rate": 0.00019998219965624734,
-      "loss": 2.0546,
+      "loss": 1.7304,
       "step": 18
     },
     {
       "epoch": 0.019005345253352506,
-      "grad_norm": 0.9696593880653381,
+      "grad_norm": 0.8139007687568665,
       "learning_rate": 0.0001999801669681376,
-      "loss": 2.0317,
+      "loss": 1.6932,
       "step": 19
     },
     {
       "epoch": 0.020005626582476323,
-      "grad_norm": 0.9394300580024719,
+      "grad_norm": 0.7991273999214172,
       "learning_rate": 0.00019997802441990573,
-      "loss": 2.2883,
+      "loss": 1.9596,
       "step": 20
     },
     {
       "epoch": 0.021005907911600136,
-      "grad_norm": 1.08865225315094,
+      "grad_norm": 0.832266628742218,
       "learning_rate": 0.00019997577201390606,
-      "loss": 1.9838,
+      "loss": 1.7116,
       "step": 21
     },
     {
       "epoch": 0.022006189240723954,
-      "grad_norm": 1.0712405443191528,
+      "grad_norm": 0.8465655446052551,
       "learning_rate": 0.00019997340975261353,
-      "loss": 2.1177,
+      "loss": 1.7711,
       "step": 22
     },
     {
       "epoch": 0.02300647056984777,
-      "grad_norm": 1.3190314769744873,
+      "grad_norm": 1.032426118850708,
       "learning_rate": 0.00019997093763862383,
-      "loss": 1.9755,
+      "loss": 1.6746,
       "step": 23
     },
     {
       "epoch": 0.024006751898971584,
-      "grad_norm": 1.0659812688827515,
+      "grad_norm": 1.0036743879318237,
       "learning_rate": 0.0001999683556746534,
-      "loss": 1.9829,
+      "loss": 1.7274,
       "step": 24
     },
     {
       "epoch": 0.0250070332280954,
-      "grad_norm": 1.1824345588684082,
+      "grad_norm": 0.9491412043571472,
       "learning_rate": 0.0001999656638635393,
-      "loss": 2.4219,
+      "loss": 2.0302,
       "step": 25
     },
     {
       "epoch": 0.02600731455721922,
-      "grad_norm": 1.3446214199066162,
+      "grad_norm": 0.9477822184562683,
       "learning_rate": 0.0001999628622082394,
-      "loss": 1.9644,
+      "loss": 1.6107,
       "step": 26
     },
     {
       "epoch": 0.027007595886343033,
-      "grad_norm": 1.2527475357055664,
+      "grad_norm": 1.0687041282653809,
       "learning_rate": 0.0001999599507118322,
-      "loss": 2.1889,
+      "loss": 1.8225,
       "step": 27
     },
     {
       "epoch": 0.02800787721546685,
-      "grad_norm": 1.4738999605178833,
+      "grad_norm": 1.6572712659835815,
       "learning_rate": 0.00019995692937751683,
-      "loss": 2.1949,
+      "loss": 1.896,
       "step": 28
     },
     {
       "epoch": 0.029008158544590667,
-      "grad_norm": 1.0533576011657715,
+      "grad_norm": 1.013258695602417,
       "learning_rate": 0.0001999537982086133,
-      "loss": 2.1034,
+      "loss": 1.7847,
       "step": 29
     },
     {
       "epoch": 0.03000843987371448,
-      "grad_norm": 1.0343223810195923,
+      "grad_norm": 0.7584932446479797,
       "learning_rate": 0.00019995055720856218,
-      "loss": 1.9561,
+      "loss": 1.5841,
       "step": 30
     },
     {
       "epoch": 0.031008721202838298,
-      "grad_norm": 1.1149976253509521,
+      "grad_norm": 1.1543537378311157,
       "learning_rate": 0.00019994720638092468,
-      "loss": 2.0981,
+      "loss": 1.8362,
       "step": 31
     },
     {
       "epoch": 0.032009002531962115,
-      "grad_norm": 1.197178840637207,
+      "grad_norm": 0.8389608860015869,
       "learning_rate": 0.00019994374572938277,
-      "loss": 2.1587,
+      "loss": 1.7913,
       "step": 32
     },
     {
       "epoch": 0.03300928386108593,
-      "grad_norm": 0.9382303953170776,
+      "grad_norm": 0.7582125663757324,
       "learning_rate": 0.00019994017525773913,
-      "loss": 1.869,
+      "loss": 1.5406,
       "step": 33
     },
     {
       "epoch": 0.03400956519020975,
-      "grad_norm": 1.0526461601257324,
+      "grad_norm": 0.7866935133934021,
       "learning_rate": 0.00019993649496991705,
-      "loss": 1.9045,
+      "loss": 1.5363,
       "step": 34
     },
     {
       "epoch": 0.03500984651933356,
-      "grad_norm": 0.8510498404502869,
+      "grad_norm": 0.8007768988609314,
       "learning_rate": 0.00019993270486996046,
-      "loss": 2.1005,
+      "loss": 1.7597,
       "step": 35
     },
     {
       "epoch": 0.03601012784845738,
-      "grad_norm": 0.9990401268005371,
+      "grad_norm": 0.8109031319618225,
       "learning_rate": 0.000199928804962034,
-      "loss": 1.8569,
+      "loss": 1.5554,
       "step": 36
     },
     {
       "epoch": 0.037010409177581194,
-      "grad_norm": 0.9243854284286499,
+      "grad_norm": 0.7722628116607666,
       "learning_rate": 0.00019992479525042303,
-      "loss": 1.9666,
+      "loss": 1.6437,
       "step": 37
     },
     {
       "epoch": 0.03801069050670501,
-      "grad_norm": 0.7774227261543274,
+      "grad_norm": 0.7336480021476746,
       "learning_rate": 0.00019992067573953342,
-      "loss": 2.0376,
+      "loss": 1.7276,
       "step": 38
     },
     {
       "epoch": 0.03901097183582883,
-      "grad_norm": 0.8114833235740662,
+      "grad_norm": 0.6940280795097351,
       "learning_rate": 0.0001999164464338918,
-      "loss": 2.1608,
+      "loss": 1.846,
       "step": 39
     },
     {
       "epoch": 0.040011253164952645,
-      "grad_norm": 0.8716320395469666,
+      "grad_norm": 0.7079702615737915,
       "learning_rate": 0.0001999121073381454,
-      "loss": 2.0743,
+      "loss": 1.7017,
       "step": 40
     },
     {
       "epoch": 0.041011534494076456,
-      "grad_norm": 0.9571239948272705,
+      "grad_norm": 0.7438498139381409,
       "learning_rate": 0.0001999076584570621,
-      "loss": 2.0128,
+      "loss": 1.665,
       "step": 41
     },
     {
       "epoch": 0.04201181582320027,
-      "grad_norm": 1.038691520690918,
+      "grad_norm": 0.6951525211334229,
       "learning_rate": 0.00019990309979553045,
-      "loss": 1.976,
+      "loss": 1.588,
       "step": 42
     },
     {
       "epoch": 0.04301209715232409,
-      "grad_norm": 1.0576292276382446,
+      "grad_norm": 0.9398604035377502,
       "learning_rate": 0.00019989843135855958,
-      "loss": 1.94,
+      "loss": 1.6513,
       "step": 43
     },
     {
       "epoch": 0.04401237848144791,
-      "grad_norm": 1.0991204977035522,
+      "grad_norm": 0.7384347319602966,
       "learning_rate": 0.00019989365315127922,
-      "loss": 1.9397,
+      "loss": 1.5975,
       "step": 44
     },
     {
       "epoch": 0.045012659810571724,
-      "grad_norm": 0.9268686175346375,
+      "grad_norm": 0.9856846332550049,
       "learning_rate": 0.0001998887651789398,
-      "loss": 1.9305,
+      "loss": 1.644,
       "step": 45
     },
     {
       "epoch": 0.04601294113969554,
-      "grad_norm": 0.8459104299545288,
+      "grad_norm": 0.7322820425033569,
       "learning_rate": 0.0001998837674469123,
-      "loss": 1.7941,
+      "loss": 1.5207,
       "step": 46
     },
     {
       "epoch": 0.04701322246881936,
-      "grad_norm": 0.9260527491569519,
+      "grad_norm": 0.8695257902145386,
       "learning_rate": 0.00019987865996068833,
-      "loss": 1.8843,
+      "loss": 1.5572,
       "step": 47
     },
     {
       "epoch": 0.04801350379794317,
-      "grad_norm": 0.8370497226715088,
+      "grad_norm": 0.7231017351150513,
       "learning_rate": 0.00019987344272588006,
-      "loss": 1.8779,
+      "loss": 1.5841,
       "step": 48
     },
     {
       "epoch": 0.049013785127066986,
-      "grad_norm": 0.9228008389472961,
+      "grad_norm": 0.7147384285926819,
       "learning_rate": 0.00019986811574822033,
-      "loss": 2.1713,
+      "loss": 1.8628,
       "step": 49
     },
     {
       "epoch": 0.0500140664561908,
-      "grad_norm": 1.013746738433838,
+      "grad_norm": 0.8631477355957031,
       "learning_rate": 0.00019986267903356254,
-      "loss": 2.1443,
+      "loss": 1.8487,
       "step": 50
     },
     {
       "epoch": 0.05101434778531462,
-      "grad_norm": 1.0155737400054932,
+      "grad_norm": 0.7995486855506897,
       "learning_rate": 0.0001998571325878806,
-      "loss": 1.9679,
+      "loss": 1.6491,
       "step": 51
     },
     {
       "epoch": 0.05201462911443844,
-      "grad_norm": 0.9591345191001892,
+      "grad_norm": 0.7828657031059265,
       "learning_rate": 0.0001998514764172691,
-      "loss": 2.0611,
+      "loss": 1.7496,
       "step": 52
     },
     {
       "epoch": 0.053014910443562255,
-      "grad_norm": 0.9030050039291382,
+      "grad_norm": 0.7789833545684814,
       "learning_rate": 0.00019984571052794313,
-      "loss": 1.9698,
+      "loss": 1.6628,
       "step": 53
     },
     {
       "epoch": 0.054015191772686065,
-      "grad_norm": 0.7697799205780029,
+      "grad_norm": 0.7077661752700806,
       "learning_rate": 0.00019983983492623833,
-      "loss": 2.0609,
+      "loss": 1.771,
       "step": 54
     },
     {
       "epoch": 0.05501547310180988,
-      "grad_norm": 0.8806005716323853,
+      "grad_norm": 0.7939582467079163,
       "learning_rate": 0.00019983384961861096,
-      "loss": 1.9756,
+      "loss": 1.707,
       "step": 55
     },
     {
       "epoch": 0.0560157544309337,
-      "grad_norm": 0.9424449801445007,
+      "grad_norm": 0.9438828229904175,
       "learning_rate": 0.0001998277546116378,
-      "loss": 2.0913,
+      "loss": 1.8334,
       "step": 56
     },
     {
       "epoch": 0.05701603576005752,
-      "grad_norm": 1.139495849609375,
+      "grad_norm": 0.8028286695480347,
       "learning_rate": 0.00019982154991201608,
-      "loss": 2.2524,
+      "loss": 1.9117,
       "step": 57
     },
     {
       "epoch": 0.058016317089181334,
-      "grad_norm": 1.094347357749939,
+      "grad_norm": 0.6563037037849426,
       "learning_rate": 0.00019981523552656377,
-      "loss": 1.8501,
+      "loss": 1.4767,
       "step": 58
     },
     {
       "epoch": 0.05901659841830515,
-      "grad_norm": 1.1519278287887573,
+      "grad_norm": 0.6600964665412903,
       "learning_rate": 0.00019980881146221914,
-      "loss": 1.9866,
+      "loss": 1.6656,
       "step": 59
     },
     {
       "epoch": 0.06001687974742896,
-      "grad_norm": 1.2018250226974487,
+      "grad_norm": 0.7966578602790833,
       "learning_rate": 0.00019980227772604112,
-      "loss": 1.8226,
+      "loss": 1.4844,
       "step": 60
     },
     {
       "epoch": 0.06101716107655278,
-      "grad_norm": 0.9565753936767578,
+      "grad_norm": 0.8019976615905762,
       "learning_rate": 0.0001997956343252091,
-      "loss": 1.8434,
+      "loss": 1.5682,
       "step": 61
     },
     {
       "epoch": 0.062017442405676595,
-      "grad_norm": 1.0832768678665161,
+      "grad_norm": 0.8935349583625793,
       "learning_rate": 0.00019978888126702296,
-      "loss": 2.1271,
+      "loss": 1.8131,
       "step": 62
     },
     {
       "epoch": 0.06301772373480041,
-      "grad_norm": 0.8973837494850159,
+      "grad_norm": 0.8085179924964905,
       "learning_rate": 0.00019978201855890308,
-      "loss": 1.8331,
+      "loss": 1.5602,
       "step": 63
     },
     {
       "epoch": 0.06401800506392423,
-      "grad_norm": 0.8754604458808899,
+      "grad_norm": 0.7631951570510864,
       "learning_rate": 0.00019977504620839035,
-      "loss": 2.1379,
+      "loss": 1.8008,
       "step": 64
     },
     {
       "epoch": 0.06501828639304805,
-      "grad_norm": 0.8244839310646057,
+      "grad_norm": 0.7315165996551514,
       "learning_rate": 0.00019976796422314615,
-      "loss": 1.8431,
+      "loss": 1.5735,
       "step": 65
     },
     {
       "epoch": 0.06601856772217186,
-      "grad_norm": 0.8213551044464111,
+      "grad_norm": 0.745726466178894,
       "learning_rate": 0.00019976077261095226,
-      "loss": 1.9155,
+      "loss": 1.5775,
       "step": 66
     },
     {
       "epoch": 0.06701884905129568,
-      "grad_norm": 0.9140985608100891,
+      "grad_norm": 0.9082249999046326,
       "learning_rate": 0.00019975347137971098,
-      "loss": 2.0651,
+      "loss": 1.7427,
       "step": 67
     },
     {
       "epoch": 0.0680191303804195,
-      "grad_norm": 0.8518921732902527,
+      "grad_norm": 0.6575669050216675,
       "learning_rate": 0.00019974606053744503,
-      "loss": 1.8197,
+      "loss": 1.5231,
       "step": 68
     },
     {
       "epoch": 0.06901941170954332,
-      "grad_norm": 0.8397145867347717,
+      "grad_norm": 0.7749233245849609,
       "learning_rate": 0.00019973854009229763,
-      "loss": 1.8621,
+      "loss": 1.5703,
       "step": 69
     },
     {
       "epoch": 0.07001969303866712,
-      "grad_norm": 0.8727964162826538,
+      "grad_norm": 0.7240824699401855,
       "learning_rate": 0.00019973091005253232,
-      "loss": 1.762,
+      "loss": 1.5197,
       "step": 70
     },
     {
       "epoch": 0.07101997436779094,
-      "grad_norm": 0.9284623265266418,
+      "grad_norm": 0.8683856725692749,
       "learning_rate": 0.0001997231704265332,
-      "loss": 1.8675,
+      "loss": 1.6183,
       "step": 71
     },
     {
       "epoch": 0.07202025569691475,
-      "grad_norm": 0.8280015587806702,
+      "grad_norm": 0.6885640621185303,
       "learning_rate": 0.00019971532122280464,
-      "loss": 1.931,
+      "loss": 1.6565,
       "step": 72
     },
     {
       "epoch": 0.07302053702603857,
-      "grad_norm": 0.7591394186019897,
+      "grad_norm": 0.6648329496383667,
       "learning_rate": 0.0001997073624499716,
-      "loss": 1.8485,
+      "loss": 1.5943,
       "step": 73
     },
     {
       "epoch": 0.07402081835516239,
-      "grad_norm": 0.975128710269928,
+      "grad_norm": 0.8867416977882385,
       "learning_rate": 0.0001996992941167792,
-      "loss": 2.0784,
+      "loss": 1.7855,
       "step": 74
     },
     {
       "epoch": 0.0750210996842862,
-      "grad_norm": 0.8034948110580444,
+      "grad_norm": 0.7790491580963135,
       "learning_rate": 0.00019969111623209323,
-      "loss": 1.9849,
+      "loss": 1.6723,
       "step": 75
     },
     {
       "epoch": 0.07602138101341002,
-      "grad_norm": 0.8540483713150024,
+      "grad_norm": 0.7999201416969299,
       "learning_rate": 0.00019968282880489957,
-      "loss": 1.7832,
+      "loss": 1.5619,
       "step": 76
     },
     {
       "epoch": 0.07702166234253384,
-      "grad_norm": 0.8181695342063904,
+      "grad_norm": 0.6316407322883606,
       "learning_rate": 0.00019967443184430467,
-      "loss": 1.944,
+      "loss": 1.6377,
       "step": 77
     },
     {
       "epoch": 0.07802194367165766,
-      "grad_norm": 0.8446747064590454,
+      "grad_norm": 0.7680445313453674,
       "learning_rate": 0.0001996659253595353,
-      "loss": 1.8508,
+      "loss": 1.5433,
       "step": 78
     },
     {
       "epoch": 0.07902222500078147,
-      "grad_norm": 0.8280364871025085,
+      "grad_norm": 0.7158446907997131,
       "learning_rate": 0.0001996573093599385,
-      "loss": 1.843,
+      "loss": 1.5436,
       "step": 79
     },
     {
       "epoch": 0.08002250632990529,
-      "grad_norm": 0.8016006350517273,
+      "grad_norm": 0.7354825139045715,
       "learning_rate": 0.00019964858385498172,
-      "loss": 1.9368,
+      "loss": 1.6512,
       "step": 80
     },
     {
       "epoch": 0.08102278765902911,
-      "grad_norm": 0.8450536131858826,
+      "grad_norm": 0.7031210660934448,
       "learning_rate": 0.00019963974885425266,
-      "loss": 1.9736,
+      "loss": 1.6411,
       "step": 81
     },
     {
       "epoch": 0.08202306898815291,
-      "grad_norm": 0.9172171950340271,
+      "grad_norm": 0.8451045751571655,
       "learning_rate": 0.00019963080436745945,
-      "loss": 1.9382,
+      "loss": 1.6622,
       "step": 82
     },
     {
       "epoch": 0.08302335031727673,
-      "grad_norm": 0.8581916689872742,
+      "grad_norm": 0.8329521417617798,
       "learning_rate": 0.00019962175040443044,
-      "loss": 2.224,
+      "loss": 1.9269,
       "step": 83
     },
     {
       "epoch": 0.08402363164640055,
-      "grad_norm": 0.9350367188453674,
+      "grad_norm": 0.6967645883560181,
       "learning_rate": 0.0001996125869751143,
-      "loss": 2.0519,
+      "loss": 1.7243,
       "step": 84
     },
     {
       "epoch": 0.08502391297552436,
-      "grad_norm": 0.9276247620582581,
+      "grad_norm": 0.8699042797088623,
       "learning_rate": 0.00019960331408957997,
-      "loss": 1.9657,
+      "loss": 1.7211,
       "step": 85
     },
     {
       "epoch": 0.08602419430464818,
-      "grad_norm": 0.871574342250824,
+      "grad_norm": 0.6780512928962708,
       "learning_rate": 0.00019959393175801671,
-      "loss": 1.9399,
+      "loss": 1.6376,
       "step": 86
     },
     {
       "epoch": 0.087024475633772,
-      "grad_norm": 1.0662888288497925,
+      "grad_norm": 0.7213720679283142,
       "learning_rate": 0.00019958443999073397,
-      "loss": 1.9089,
+      "loss": 1.6048,
       "step": 87
     },
     {
       "epoch": 0.08802475696289581,
-      "grad_norm": 0.8258713483810425,
+      "grad_norm": 0.6077585816383362,
       "learning_rate": 0.00019957483879816151,
-      "loss": 1.839,
+      "loss": 1.5231,
       "step": 88
     },
     {
       "epoch": 0.08902503829201963,
-      "grad_norm": 0.8154664039611816,
+      "grad_norm": 0.6854611039161682,
       "learning_rate": 0.00019956512819084928,
-      "loss": 1.8409,
+      "loss": 1.4726,
       "step": 89
     },
     {
       "epoch": 0.09002531962114345,
-      "grad_norm": 0.8584638833999634,
+      "grad_norm": 0.6969390511512756,
       "learning_rate": 0.00019955530817946748,
-      "loss": 1.9521,
+      "loss": 1.6435,
       "step": 90
     },
     {
       "epoch": 0.09102560095026727,
-      "grad_norm": 0.7917523384094238,
+      "grad_norm": 0.7178792953491211,
       "learning_rate": 0.00019954537877480655,
-      "loss": 1.9495,
+      "loss": 1.6967,
       "step": 91
     },
     {
       "epoch": 0.09202588227939108,
-      "grad_norm": 1.0129039287567139,
+      "grad_norm": 0.8248458504676819,
       "learning_rate": 0.00019953533998777706,
-      "loss": 1.949,
+      "loss": 1.5884,
       "step": 92
     },
     {
       "epoch": 0.0930261636085149,
-      "grad_norm": 0.8677986264228821,
+      "grad_norm": 0.6472075581550598,
       "learning_rate": 0.00019952519182940993,
-      "loss": 1.7875,
+      "loss": 1.5074,
       "step": 93
     },
     {
       "epoch": 0.09402644493763872,
-      "grad_norm": 0.8848614692687988,
+      "grad_norm": 0.7548672556877136,
       "learning_rate": 0.00019951493431085603,
-      "loss": 2.0675,
+      "loss": 1.6934,
       "step": 94
     },
     {
       "epoch": 0.09502672626676252,
-      "grad_norm": 0.9936463832855225,
+      "grad_norm": 0.6680666208267212,
       "learning_rate": 0.00019950456744338658,
-      "loss": 1.6761,
+      "loss": 1.4014,
       "step": 95
     },
     {
       "epoch": 0.09602700759588634,
-      "grad_norm": 1.0520148277282715,
+      "grad_norm": 0.7270862460136414,
       "learning_rate": 0.00019949409123839288,
-      "loss": 2.081,
+      "loss": 1.6991,
       "step": 96
     },
     {
       "epoch": 0.09702728892501016,
-      "grad_norm": 0.8061773180961609,
+      "grad_norm": 0.682833731174469,
       "learning_rate": 0.00019948350570738642,
-      "loss": 1.7281,
+      "loss": 1.4926,
       "step": 97
     },
     {
       "epoch": 0.09802757025413397,
-      "grad_norm": 0.7642756104469299,
+      "grad_norm": 0.6598315238952637,
       "learning_rate": 0.0001994728108619987,
-      "loss": 2.0032,
+      "loss": 1.6329,
       "step": 98
     },
     {
       "epoch": 0.09902785158325779,
-      "grad_norm": 0.8541550040245056,
+      "grad_norm": 0.830845832824707,
       "learning_rate": 0.0001994620067139815,
-      "loss": 2.1136,
+      "loss": 1.8517,
       "step": 99
     },
     {
       "epoch": 0.1000281329123816,
-      "grad_norm": 0.7868679761886597,
+      "grad_norm": 0.6964694857597351,
       "learning_rate": 0.00019945109327520658,
-      "loss": 1.8695,
+      "loss": 1.5459,
       "step": 100
     },
     {
       "epoch": 0.10102841424150542,
-      "grad_norm": 0.8776901364326477,
+      "grad_norm": 0.8640177249908447,
       "learning_rate": 0.00019944007055766586,
-      "loss": 1.9786,
+      "loss": 1.6638,
       "step": 101
     },
     {
       "epoch": 0.10202869557062924,
-      "grad_norm": 0.9013833999633789,
+      "grad_norm": 0.6485210657119751,
       "learning_rate": 0.00019942893857347128,
-      "loss": 2.1466,
+      "loss": 1.8025,
       "step": 102
     },
     {
       "epoch": 0.10302897689975306,
-      "grad_norm": 0.957558274269104,
+      "grad_norm": 0.6746248006820679,
       "learning_rate": 0.00019941769733485494,
-      "loss": 2.0473,
+      "loss": 1.6954,
       "step": 103
     },
     {
       "epoch": 0.10402925822887688,
-      "grad_norm": 0.8921108841896057,
+      "grad_norm": 0.7386549115180969,
       "learning_rate": 0.00019940634685416888,
-      "loss": 1.7882,
+      "loss": 1.4547,
       "step": 104
     },
     {
       "epoch": 0.10502953955800069,
-      "grad_norm": 0.896019697189331,
+      "grad_norm": 0.7518633008003235,
       "learning_rate": 0.00019939488714388524,
-      "loss": 1.8811,
+      "loss": 1.5098,
       "step": 105
     },
     {
       "epoch": 0.10602982088712451,
-      "grad_norm": 0.8792067766189575,
+      "grad_norm": 0.7350422739982605,
       "learning_rate": 0.00019938331821659614,
-      "loss": 1.8624,
+      "loss": 1.5452,
       "step": 106
     },
     {
       "epoch": 0.10703010221624833,
-      "grad_norm": 0.8739930391311646,
+      "grad_norm": 0.6544668674468994,
       "learning_rate": 0.0001993716400850138,
-      "loss": 1.8105,
+      "loss": 1.5106,
       "step": 107
     },
     {
       "epoch": 0.10803038354537213,
-      "grad_norm": 0.7678424715995789,
+      "grad_norm": 0.6108564138412476,
       "learning_rate": 0.0001993598527619703,
-      "loss": 1.8772,
+      "loss": 1.5818,
       "step": 108
     },
     {
       "epoch": 0.10903066487449595,
-      "grad_norm": 0.8718745112419128,
+      "grad_norm": 0.731071949005127,
       "learning_rate": 0.00019934795626041783,
-      "loss": 1.8236,
+      "loss": 1.4819,
       "step": 109
     },
     {
       "epoch": 0.11003094620361976,
-      "grad_norm": 0.8467247486114502,
+      "grad_norm": 0.5978986620903015,
       "learning_rate": 0.0001993359505934285,
-      "loss": 1.8188,
+      "loss": 1.5469,
       "step": 110
     },
     {
       "epoch": 0.11103122753274358,
-      "grad_norm": 0.8685783743858337,
+      "grad_norm": 0.7249881029129028,
       "learning_rate": 0.00019932383577419432,
-      "loss": 2.0775,
+      "loss": 1.7466,
       "step": 111
     },
     {
       "epoch": 0.1120315088618674,
-      "grad_norm": 0.7799698710441589,
+      "grad_norm": 0.6161806583404541,
       "learning_rate": 0.0001993116118160273,
-      "loss": 1.6489,
+      "loss": 1.3411,
       "step": 112
     },
     {
       "epoch": 0.11303179019099122,
-      "grad_norm": 0.7900094389915466,
+      "grad_norm": 0.6745229363441467,
       "learning_rate": 0.00019929927873235938,
-      "loss": 1.8332,
+      "loss": 1.5615,
       "step": 113
     },
     {
       "epoch": 0.11403207152011503,
-      "grad_norm": 0.9433258771896362,
+      "grad_norm": 0.6489872336387634,
       "learning_rate": 0.00019928683653674237,
-      "loss": 1.9331,
+      "loss": 1.6279,
       "step": 114
     },
     {
       "epoch": 0.11503235284923885,
-      "grad_norm": 0.8861056566238403,
+      "grad_norm": 0.7769975662231445,
       "learning_rate": 0.00019927428524284805,
-      "loss": 1.9135,
+      "loss": 1.6155,
       "step": 115
     },
     {
       "epoch": 0.11603263417836267,
-      "grad_norm": 0.8566756844520569,
+      "grad_norm": 0.734336793422699,
       "learning_rate": 0.00019926162486446792,
-      "loss": 1.9874,
+      "loss": 1.6837,
       "step": 116
     },
     {
       "epoch": 0.11703291550748648,
-      "grad_norm": 0.6897929310798645,
+      "grad_norm": 0.6966903805732727,
       "learning_rate": 0.0001992488554155135,
-      "loss": 1.946,
+      "loss": 1.6659,
       "step": 117
     },
     {
       "epoch": 0.1180331968366103,
-      "grad_norm": 0.7807729244232178,
+      "grad_norm": 0.6714586615562439,
       "learning_rate": 0.00019923597691001615,
-      "loss": 1.8127,
+      "loss": 1.5161,
       "step": 118
     },
     {
       "epoch": 0.11903347816573412,
-      "grad_norm": 0.7572523355484009,
+      "grad_norm": 0.6390894651412964,
       "learning_rate": 0.0001992229893621269,
-      "loss": 1.7768,
+      "loss": 1.4561,
       "step": 119
     },
     {
       "epoch": 0.12003375949485792,
-      "grad_norm": 0.7393172979354858,
+      "grad_norm": 0.6481143832206726,
       "learning_rate": 0.00019920989278611687,
-      "loss": 1.894,
+      "loss": 1.6331,
       "step": 120
     },
     {
       "epoch": 0.12103404082398174,
-      "grad_norm": 0.866576611995697,
+      "grad_norm": 0.6819384694099426,
       "learning_rate": 0.0001991966871963767,
-      "loss": 1.9285,
+      "loss": 1.6508,
       "step": 121
     },
     {
       "epoch": 0.12203432215310556,
-      "grad_norm": 0.7326533794403076,
+      "grad_norm": 0.6839059591293335,
       "learning_rate": 0.000199183372607417,
-      "loss": 1.9309,
+      "loss": 1.6514,
       "step": 122
     },
     {
       "epoch": 0.12303460348222937,
-      "grad_norm": 0.7655537724494934,
+      "grad_norm": 0.6401050090789795,
       "learning_rate": 0.0001991699490338681,
-      "loss": 2.1145,
+      "loss": 1.8065,
       "step": 123
     },
     {
       "epoch": 0.12403488481135319,
-      "grad_norm": 0.9714633226394653,
+      "grad_norm": 0.6860588788986206,
       "learning_rate": 0.00019915641649048005,
-      "loss": 2.0341,
+      "loss": 1.7658,
       "step": 124
     },
     {
       "epoch": 0.12503516614047702,
-      "grad_norm": 0.8542420864105225,
+      "grad_norm": 0.6286434531211853,
       "learning_rate": 0.0001991427749921227,
-      "loss": 2.1426,
+      "loss": 1.7678,
       "step": 125
     },
     {
       "epoch": 0.12603544746960083,
-      "grad_norm": 0.8286274671554565,
+      "grad_norm": 0.6609922647476196,
       "learning_rate": 0.00019912902455378556,
-      "loss": 1.8452,
+      "loss": 1.4934,
       "step": 126
     },
     {
       "epoch": 0.12703572879872463,
-      "grad_norm": 0.8823768496513367,
+      "grad_norm": 0.7058399319648743,
       "learning_rate": 0.00019911516519057788,
-      "loss": 1.8651,
+      "loss": 1.6058,
       "step": 127
     },
     {
       "epoch": 0.12803601012784846,
-      "grad_norm": 0.7301567196846008,
+      "grad_norm": 0.6362051963806152,
       "learning_rate": 0.00019910119691772863,
-      "loss": 1.7776,
+      "loss": 1.502,
       "step": 128
     },
     {
       "epoch": 0.12903629145697226,
-      "grad_norm": 0.8402552604675293,
+      "grad_norm": 0.7493100762367249,
       "learning_rate": 0.00019908711975058637,
-      "loss": 1.8617,
+      "loss": 1.5287,
       "step": 129
     },
     {
       "epoch": 0.1300365727860961,
-      "grad_norm": 0.814500093460083,
+      "grad_norm": 0.6492393612861633,
       "learning_rate": 0.0001990729337046194,
-      "loss": 1.9156,
+      "loss": 1.5716,
       "step": 130
     },
     {
       "epoch": 0.1310368541152199,
-      "grad_norm": 0.8262699246406555,
+      "grad_norm": 0.70331871509552,
       "learning_rate": 0.0001990586387954156,
-      "loss": 1.8659,
+      "loss": 1.5882,
       "step": 131
     },
     {
       "epoch": 0.13203713544434373,
-      "grad_norm": 0.8846324682235718,
+      "grad_norm": 0.7581572532653809,
       "learning_rate": 0.00019904423503868247,
-      "loss": 2.043,
+      "loss": 1.7627,
       "step": 132
     },
     {
       "epoch": 0.13303741677346753,
-      "grad_norm": 0.8757227659225464,
+      "grad_norm": 0.7087228894233704,
       "learning_rate": 0.00019902972245024715,
-      "loss": 1.9217,
+      "loss": 1.6257,
       "step": 133
     },
     {
       "epoch": 0.13403769810259136,
-      "grad_norm": 0.8476879596710205,
+      "grad_norm": 0.7920627593994141,
       "learning_rate": 0.00019901510104605637,
-      "loss": 1.8892,
+      "loss": 1.572,
       "step": 134
     },
     {
       "epoch": 0.13503797943171517,
-      "grad_norm": 0.7707583904266357,
+      "grad_norm": 0.6869202256202698,
       "learning_rate": 0.00019900037084217637,
-      "loss": 1.787,
+      "loss": 1.5478,
       "step": 135
     },
     {
       "epoch": 0.136038260760839,
-      "grad_norm": 0.7389562129974365,
+      "grad_norm": 0.6879409551620483,
       "learning_rate": 0.00019898553185479303,
-      "loss": 1.5854,
+      "loss": 1.3104,
       "step": 136
     },
     {
       "epoch": 0.1370385420899628,
-      "grad_norm": 0.7331375479698181,
+      "grad_norm": 0.6574143767356873,
       "learning_rate": 0.00019897058410021167,
-      "loss": 1.997,
+      "loss": 1.7041,
       "step": 137
     },
     {
       "epoch": 0.13803882341908663,
-      "grad_norm": 0.7219388484954834,
+      "grad_norm": 0.7793259620666504,
       "learning_rate": 0.00019895552759485722,
-      "loss": 1.8337,
+      "loss": 1.5417,
       "step": 138
     },
     {
       "epoch": 0.13903910474821043,
-      "grad_norm": 0.8535702228546143,
+      "grad_norm": 0.6310438513755798,
       "learning_rate": 0.00019894036235527395,
-      "loss": 1.7818,
+      "loss": 1.4978,
       "step": 139
     },
     {
       "epoch": 0.14003938607733424,
-      "grad_norm": 0.7627841830253601,
+      "grad_norm": 0.6298012137413025,
       "learning_rate": 0.00019892508839812584,
-      "loss": 1.8173,
+      "loss": 1.5367,
       "step": 140
     },
     {
       "epoch": 0.14103966740645807,
-      "grad_norm": 0.8397619724273682,
+      "grad_norm": 0.5647856593132019,
       "learning_rate": 0.00019890970574019617,
-      "loss": 1.8735,
+      "loss": 1.537,
       "step": 141
     },
     {
       "epoch": 0.14203994873558187,
-      "grad_norm": 0.8093482851982117,
+      "grad_norm": 0.6491876244544983,
       "learning_rate": 0.00019889421439838763,
-      "loss": 1.9918,
+      "loss": 1.6992,
       "step": 142
     },
     {
       "epoch": 0.1430402300647057,
-      "grad_norm": 0.8853684067726135,
+      "grad_norm": 0.6574720144271851,
       "learning_rate": 0.00019887861438972246,
-      "loss": 1.5825,
+      "loss": 1.3837,
       "step": 143
     },
     {
       "epoch": 0.1440405113938295,
-      "grad_norm": 0.7413788437843323,
+      "grad_norm": 0.6267092227935791,
       "learning_rate": 0.00019886290573134228,
-      "loss": 1.9068,
+      "loss": 1.6307,
       "step": 144
     },
     {
       "epoch": 0.14504079272295334,
-      "grad_norm": 0.7924477458000183,
+      "grad_norm": 0.6785029172897339,
       "learning_rate": 0.000198847088440508,
-      "loss": 1.8582,
+      "loss": 1.574,
       "step": 145
     },
     {
       "epoch": 0.14604107405207714,
-      "grad_norm": 0.8679131865501404,
+      "grad_norm": 0.6218644380569458,
       "learning_rate": 0.0001988311625346,
-      "loss": 1.7104,
+      "loss": 1.4676,
       "step": 146
     },
     {
       "epoch": 0.14704135538120097,
-      "grad_norm": 0.7480150461196899,
+      "grad_norm": 0.6047986745834351,
       "learning_rate": 0.00019881512803111796,
-      "loss": 1.7288,
+      "loss": 1.4316,
       "step": 147
     },
     {
       "epoch": 0.14804163671032478,
-      "grad_norm": 0.8382390737533569,
+      "grad_norm": 0.7340937256813049,
       "learning_rate": 0.00019879898494768093,
-      "loss": 1.8004,
+      "loss": 1.5185,
       "step": 148
     },
     {
       "epoch": 0.1490419180394486,
-      "grad_norm": 0.7360037565231323,
+      "grad_norm": 0.5874620676040649,
       "learning_rate": 0.00019878273330202717,
-      "loss": 1.85,
+      "loss": 1.5031,
       "step": 149
     },
     {
       "epoch": 0.1500421993685724,
-      "grad_norm": 0.9644019603729248,
+      "grad_norm": 0.6943556666374207,
       "learning_rate": 0.00019876637311201433,
-      "loss": 2.065,
+      "loss": 1.7323,
       "step": 150
     },
     {
       "epoch": 0.15104248069769624,
-      "grad_norm": 0.8116248250007629,
+      "grad_norm": 0.6345832347869873,
       "learning_rate": 0.00019874990439561934,
-      "loss": 1.702,
+      "loss": 1.4691,
       "step": 151
     },
     {
       "epoch": 0.15204276202682004,
-      "grad_norm": 0.9301722645759583,
+      "grad_norm": 0.7047753930091858,
       "learning_rate": 0.0001987333271709383,
-      "loss": 1.8089,
+      "loss": 1.5198,
       "step": 152
     },
     {
       "epoch": 0.15304304335594385,
-      "grad_norm": 0.7991555333137512,
+      "grad_norm": 0.6043322086334229,
       "learning_rate": 0.00019871664145618657,
-      "loss": 1.8227,
+      "loss": 1.5488,
       "step": 153
     },
     {
       "epoch": 0.15404332468506768,
-      "grad_norm": 0.8676092028617859,
+      "grad_norm": 0.5978446006774902,
       "learning_rate": 0.00019869984726969878,
-      "loss": 1.7253,
+      "loss": 1.4278,
       "step": 154
     },
     {
       "epoch": 0.15504360601419148,
-      "grad_norm": 0.8022972941398621,
+      "grad_norm": 0.6796436905860901,
       "learning_rate": 0.00019868294462992866,
-      "loss": 1.8766,
+      "loss": 1.5845,
       "step": 155
     },
     {
       "epoch": 0.1560438873433153,
-      "grad_norm": 1.128886103630066,
+      "grad_norm": 0.7113372087478638,
       "learning_rate": 0.00019866593355544922,
-      "loss": 2.0197,
+      "loss": 1.7509,
       "step": 156
     },
     {
       "epoch": 0.15704416867243912,
-      "grad_norm": 0.7420483827590942,
+      "grad_norm": 0.5908107757568359,
       "learning_rate": 0.00019864881406495246,
-      "loss": 1.8825,
+      "loss": 1.5693,
       "step": 157
     },
     {
       "epoch": 0.15804445000156295,
-      "grad_norm": 0.7797536849975586,
+      "grad_norm": 0.7135252952575684,
       "learning_rate": 0.00019863158617724967,
-      "loss": 1.8892,
+      "loss": 1.6109,
       "step": 158
     },
     {
       "epoch": 0.15904473133068675,
-      "grad_norm": 0.6859965324401855,
+      "grad_norm": 0.5621710419654846,
       "learning_rate": 0.00019861424991127115,
-      "loss": 1.8424,
+      "loss": 1.5368,
       "step": 159
     },
     {
       "epoch": 0.16004501265981058,
-      "grad_norm": 0.8115108609199524,
+      "grad_norm": 0.6205443143844604,
       "learning_rate": 0.00019859680528606637,
-      "loss": 1.8394,
+      "loss": 1.5181,
       "step": 160
     },
     {
       "epoch": 0.16104529398893438,
-      "grad_norm": 0.9756322503089905,
+      "grad_norm": 0.6933260560035706,
       "learning_rate": 0.00019857925232080373,
-      "loss": 1.726,
+      "loss": 1.4508,
       "step": 161
     },
     {
       "epoch": 0.16204557531805822,
-      "grad_norm": 0.8894350528717041,
+      "grad_norm": 0.6911661028862,
       "learning_rate": 0.00019856159103477086,
-      "loss": 1.8893,
+      "loss": 1.5423,
       "step": 162
     },
     {
       "epoch": 0.16304585664718202,
-      "grad_norm": 0.8075819611549377,
+      "grad_norm": 0.7684744000434875,
       "learning_rate": 0.00019854382144737426,
-      "loss": 1.6596,
+      "loss": 1.4097,
       "step": 163
     },
     {
       "epoch": 0.16404613797630582,
-      "grad_norm": 0.8861923813819885,
+      "grad_norm": 0.6657288074493408,
       "learning_rate": 0.00019852594357813952,
-      "loss": 1.9352,
+      "loss": 1.6145,
       "step": 164
     },
     {
       "epoch": 0.16504641930542965,
-      "grad_norm": 0.8511936068534851,
+      "grad_norm": 0.7030160427093506,
       "learning_rate": 0.00019850795744671116,
-      "loss": 1.9416,
+      "loss": 1.6551,
       "step": 165
     },
     {
       "epoch": 0.16604670063455346,
-      "grad_norm": 0.9425658583641052,
+      "grad_norm": 0.87894207239151,
       "learning_rate": 0.0001984898630728527,
-      "loss": 1.9081,
+      "loss": 1.6316,
       "step": 166
     },
     {
       "epoch": 0.1670469819636773,
-      "grad_norm": 0.7502055168151855,
+      "grad_norm": 0.6282681226730347,
       "learning_rate": 0.0001984716604764466,
-      "loss": 1.703,
+      "loss": 1.451,
       "step": 167
     },
     {
       "epoch": 0.1680472632928011,
-      "grad_norm": 0.9135978817939758,
+      "grad_norm": 0.6729792952537537,
       "learning_rate": 0.0001984533496774942,
-      "loss": 1.7641,
+      "loss": 1.4381,
       "step": 168
     },
     {
       "epoch": 0.16904754462192492,
-      "grad_norm": 0.7768126726150513,
+      "grad_norm": 0.7300116419792175,
       "learning_rate": 0.0001984349306961158,
-      "loss": 1.7053,
+      "loss": 1.4244,
       "step": 169
     },
     {
       "epoch": 0.17004782595104873,
-      "grad_norm": 0.8106538653373718,
+      "grad_norm": 0.6853480935096741,
       "learning_rate": 0.00019841640355255043,
-      "loss": 1.8646,
+      "loss": 1.6174,
       "step": 170
     },
     {
       "epoch": 0.17104810728017256,
-      "grad_norm": 0.7872330546379089,
+      "grad_norm": 0.735612690448761,
       "learning_rate": 0.00019839776826715614,
-      "loss": 1.7814,
+      "loss": 1.5085,
       "step": 171
     },
     {
       "epoch": 0.17204838860929636,
-      "grad_norm": 0.869532585144043,
+      "grad_norm": 0.6735563278198242,
       "learning_rate": 0.00019837902486040978,
-      "loss": 1.7812,
+      "loss": 1.507,
       "step": 172
     },
     {
       "epoch": 0.1730486699384202,
-      "grad_norm": 1.015028715133667,
+      "grad_norm": 0.6617917418479919,
       "learning_rate": 0.0001983601733529069,
-      "loss": 1.9432,
+      "loss": 1.6774,
       "step": 173
     },
     {
       "epoch": 0.174048951267544,
-      "grad_norm": 0.800183117389679,
+      "grad_norm": 0.7137823700904846,
       "learning_rate": 0.00019834121376536187,
-      "loss": 1.758,
+      "loss": 1.4665,
       "step": 174
     },
     {
       "epoch": 0.17504923259666783,
-      "grad_norm": 0.7427104711532593,
+      "grad_norm": 0.6372626423835754,
       "learning_rate": 0.00019832214611860793,
-      "loss": 1.6476,
+      "loss": 1.3597,
       "step": 175
     },
     {
       "epoch": 0.17604951392579163,
-      "grad_norm": 0.8289130926132202,
+      "grad_norm": 0.7131632566452026,
       "learning_rate": 0.00019830297043359692,
-      "loss": 1.7702,
+      "loss": 1.4833,
       "step": 176
     },
     {
       "epoch": 0.17704979525491543,
-      "grad_norm": 0.8298771977424622,
+      "grad_norm": 0.7538559436798096,
       "learning_rate": 0.00019828368673139947,
-      "loss": 1.7515,
+      "loss": 1.4714,
       "step": 177
     },
     {
       "epoch": 0.17805007658403926,
-      "grad_norm": 0.7602815628051758,
+      "grad_norm": 0.5684806108474731,
       "learning_rate": 0.0001982642950332049,
-      "loss": 1.7545,
+      "loss": 1.5012,
       "step": 178
     },
     {
       "epoch": 0.17905035791316307,
-      "grad_norm": 0.8110321164131165,
+      "grad_norm": 0.621658444404602,
       "learning_rate": 0.00019824479536032112,
-      "loss": 2.2604,
+      "loss": 1.9119,
       "step": 179
     },
     {
       "epoch": 0.1800506392422869,
-      "grad_norm": 0.882273256778717,
+      "grad_norm": 0.6564679741859436,
       "learning_rate": 0.0001982251877341748,
-      "loss": 1.8133,
+      "loss": 1.5131,
       "step": 180
     },
     {
       "epoch": 0.1810509205714107,
-      "grad_norm": 0.9015639424324036,
+      "grad_norm": 0.6546526551246643,
       "learning_rate": 0.00019820547217631117,
-      "loss": 1.7282,
+      "loss": 1.4493,
       "step": 181
     },
     {
       "epoch": 0.18205120190053453,
-      "grad_norm": 0.9231659173965454,
+      "grad_norm": 0.6504479050636292,
       "learning_rate": 0.00019818564870839405,
-      "loss": 1.9094,
+      "loss": 1.6131,
       "step": 182
     },
     {
       "epoch": 0.18305148322965833,
-      "grad_norm": 0.8110967874526978,
+      "grad_norm": 0.6269803047180176,
       "learning_rate": 0.00019816571735220583,
-      "loss": 1.886,
+      "loss": 1.5936,
       "step": 183
     },
     {
       "epoch": 0.18405176455878217,
-      "grad_norm": 0.7670036554336548,
+      "grad_norm": 0.6303942799568176,
       "learning_rate": 0.00019814567812964748,
-      "loss": 1.9895,
+      "loss": 1.6948,
       "step": 184
     },
     {
       "epoch": 0.18505204588790597,
-      "grad_norm": 0.7955975532531738,
+      "grad_norm": 0.6562885046005249,
       "learning_rate": 0.00019812553106273847,
-      "loss": 1.8127,
+      "loss": 1.5542,
       "step": 185
     },
     {
       "epoch": 0.1860523272170298,
-      "grad_norm": 0.8790062665939331,
+      "grad_norm": 0.5844212174415588,
       "learning_rate": 0.00019810527617361681,
-      "loss": 1.8899,
+      "loss": 1.539,
       "step": 186
     },
     {
       "epoch": 0.1870526085461536,
-      "grad_norm": 0.8818586468696594,
+      "grad_norm": 0.6402295231819153,
       "learning_rate": 0.00019808491348453894,
-      "loss": 1.7707,
+      "loss": 1.4748,
       "step": 187
     },
     {
       "epoch": 0.18805288987527743,
-      "grad_norm": 0.746442437171936,
+      "grad_norm": 0.6579477190971375,
       "learning_rate": 0.00019806444301787978,
-      "loss": 1.7281,
+      "loss": 1.5114,
       "step": 188
     },
     {
       "epoch": 0.18905317120440124,
-      "grad_norm": 0.7786905169487,
+      "grad_norm": 0.6511597037315369,
       "learning_rate": 0.0001980438647961327,
-      "loss": 1.7317,
+      "loss": 1.4678,
       "step": 189
     },
     {
       "epoch": 0.19005345253352504,
-      "grad_norm": 0.9338862299919128,
+      "grad_norm": 0.6911427974700928,
       "learning_rate": 0.00019802317884190935,
-      "loss": 1.9548,
+      "loss": 1.6876,
       "step": 190
     },
     {
       "epoch": 0.19105373386264887,
-      "grad_norm": 0.7416581511497498,
+      "grad_norm": 0.6146433353424072,
       "learning_rate": 0.00019800238517793996,
-      "loss": 1.8601,
+      "loss": 1.5986,
       "step": 191
     },
     {
       "epoch": 0.19205401519177268,
-      "grad_norm": 0.6782898902893066,
+      "grad_norm": 0.6126302480697632,
       "learning_rate": 0.00019798148382707296,
-      "loss": 1.8477,
+      "loss": 1.571,
       "step": 192
     },
     {
       "epoch": 0.1930542965208965,
-      "grad_norm": 0.7389237880706787,
+      "grad_norm": 0.5751072764396667,
       "learning_rate": 0.00019796047481227515,
-      "loss": 1.7749,
+      "loss": 1.4921,
       "step": 193
     },
     {
       "epoch": 0.1940545778500203,
-      "grad_norm": 0.9711095094680786,
+      "grad_norm": 0.6484839916229248,
       "learning_rate": 0.00019793935815663163,
-      "loss": 2.0899,
+      "loss": 1.7495,
       "step": 194
     },
     {
       "epoch": 0.19505485917914414,
-      "grad_norm": 0.7949391007423401,
+      "grad_norm": 0.6875973343849182,
       "learning_rate": 0.00019791813388334581,
-      "loss": 1.8778,
+      "loss": 1.5782,
       "step": 195
     },
     {
       "epoch": 0.19605514050826794,
-      "grad_norm": 0.8871057033538818,
+      "grad_norm": 0.8130943179130554,
       "learning_rate": 0.00019789680201573933,
-      "loss": 1.7511,
+      "loss": 1.4964,
       "step": 196
     },
     {
       "epoch": 0.19705542183739178,
-      "grad_norm": 0.8664624094963074,
+      "grad_norm": 0.6734403371810913,
       "learning_rate": 0.00019787536257725202,
-      "loss": 1.7232,
+      "loss": 1.4787,
       "step": 197
     },
     {
       "epoch": 0.19805570316651558,
-      "grad_norm": 0.871658980846405,
+      "grad_norm": 0.6480582356452942,
       "learning_rate": 0.00019785381559144196,
-      "loss": 1.7987,
+      "loss": 1.5629,
       "step": 198
     },
     {
       "epoch": 0.1990559844956394,
-      "grad_norm": 0.7748361229896545,
+      "grad_norm": 0.6554624438285828,
       "learning_rate": 0.00019783216108198542,
-      "loss": 1.9239,
+      "loss": 1.5806,
       "step": 199
     },
     {
       "epoch": 0.2000562658247632,
-      "grad_norm": 0.9393408298492432,
+      "grad_norm": 0.705443263053894,
       "learning_rate": 0.00019781039907267677,
-      "loss": 2.0936,
+      "loss": 1.8372,
       "step": 200
     },
     {
       "epoch": 0.20105654715388704,
-      "grad_norm": 0.8519601225852966,
+      "grad_norm": 0.706923246383667,
       "learning_rate": 0.00019778852958742853,
-      "loss": 1.9108,
+      "loss": 1.6405,
       "step": 201
     },
     {
       "epoch": 0.20205682848301085,
-      "grad_norm": 0.8464863300323486,
+      "grad_norm": 0.7062544822692871,
       "learning_rate": 0.00019776655265027127,
-      "loss": 1.897,
+      "loss": 1.6,
       "step": 202
     },
     {
       "epoch": 0.20305710981213465,
-      "grad_norm": 0.8933351635932922,
+      "grad_norm": 0.7227569222450256,
       "learning_rate": 0.00019774446828535371,
-      "loss": 1.8204,
+      "loss": 1.5172,
       "step": 203
     },
     {
       "epoch": 0.20405739114125848,
-      "grad_norm": 0.8305785059928894,
+      "grad_norm": 0.6762563586235046,
       "learning_rate": 0.00019772227651694256,
-      "loss": 1.9135,
+      "loss": 1.6753,
       "step": 204
     },
     {
       "epoch": 0.20505767247038229,
-      "grad_norm": 0.8117037415504456,
+      "grad_norm": 0.6048421859741211,
       "learning_rate": 0.00019769997736942258,
-      "loss": 1.7585,
+      "loss": 1.4827,
       "step": 205
     },
     {
       "epoch": 0.20605795379950612,
-      "grad_norm": 0.7570348381996155,
+      "grad_norm": 0.6002956032752991,
       "learning_rate": 0.00019767757086729647,
-      "loss": 1.8373,
+      "loss": 1.5438,
       "step": 206
     },
     {
       "epoch": 0.20705823512862992,
-      "grad_norm": 0.9291234016418457,
+      "grad_norm": 0.7948954701423645,
       "learning_rate": 0.00019765505703518496,
-      "loss": 1.7774,
+      "loss": 1.4988,
       "step": 207
     },
     {
       "epoch": 0.20805851645775375,
-      "grad_norm": 0.8211004137992859,
+      "grad_norm": 0.6495680809020996,
       "learning_rate": 0.00019763243589782662,
-      "loss": 1.8766,
+      "loss": 1.5738,
       "step": 208
     },
     {
       "epoch": 0.20905879778687755,
-      "grad_norm": 0.6625431180000305,
+      "grad_norm": 0.6413107514381409,
       "learning_rate": 0.00019760970748007803,
-      "loss": 1.628,
+      "loss": 1.3794,
       "step": 209
     },
     {
       "epoch": 0.21005907911600138,
-      "grad_norm": 0.7974782586097717,
+      "grad_norm": 0.5999665260314941,
       "learning_rate": 0.0001975868718069136,
-      "loss": 1.6896,
+      "loss": 1.4313,
       "step": 210
     },
     {
       "epoch": 0.2110593604451252,
-      "grad_norm": 0.8364912867546082,
+      "grad_norm": 0.6355773210525513,
       "learning_rate": 0.00019756392890342563,
-      "loss": 1.7492,
+      "loss": 1.5107,
       "step": 211
     },
     {
       "epoch": 0.21205964177424902,
-      "grad_norm": 0.8730652332305908,
+      "grad_norm": 0.6068251729011536,
       "learning_rate": 0.00019754087879482422,
-      "loss": 1.8295,
+      "loss": 1.536,
       "step": 212
     },
     {
       "epoch": 0.21305992310337282,
-      "grad_norm": 0.7532863020896912,
+      "grad_norm": 0.5568909049034119,
       "learning_rate": 0.00019751772150643722,
-      "loss": 1.8309,
+      "loss": 1.5372,
       "step": 213
     },
     {
       "epoch": 0.21406020443249665,
-      "grad_norm": 0.7375178933143616,
+      "grad_norm": 0.5771281719207764,
       "learning_rate": 0.00019749445706371038,
-      "loss": 1.7854,
+      "loss": 1.487,
       "step": 214
     },
     {
       "epoch": 0.21506048576162046,
-      "grad_norm": 0.7524377703666687,
+      "grad_norm": 0.6146671772003174,
       "learning_rate": 0.00019747108549220702,
-      "loss": 1.7683,
+      "loss": 1.4585,
       "step": 215
     },
     {
       "epoch": 0.21606076709074426,
-      "grad_norm": 0.7331809997558594,
+      "grad_norm": 0.5595754981040955,
       "learning_rate": 0.00019744760681760832,
-      "loss": 1.7103,
+      "loss": 1.4224,
       "step": 216
     },
     {
       "epoch": 0.2170610484198681,
-      "grad_norm": 0.8083691596984863,
+      "grad_norm": 0.5873929858207703,
       "learning_rate": 0.00019742402106571314,
-      "loss": 1.674,
+      "loss": 1.4581,
       "step": 217
     },
     {
       "epoch": 0.2180613297489919,
-      "grad_norm": 0.8524570465087891,
+      "grad_norm": 0.5725668668746948,
       "learning_rate": 0.00019740032826243788,
-      "loss": 1.7227,
+      "loss": 1.4393,
       "step": 218
     },
     {
       "epoch": 0.21906161107811573,
-      "grad_norm": 0.7676658630371094,
+      "grad_norm": 0.6452648043632507,
       "learning_rate": 0.0001973765284338167,
-      "loss": 1.8561,
+      "loss": 1.6048,
       "step": 219
     },
     {
       "epoch": 0.22006189240723953,
-      "grad_norm": 0.7858710289001465,
+      "grad_norm": 0.6166092753410339,
       "learning_rate": 0.00019735262160600127,
-      "loss": 1.7796,
+      "loss": 1.4976,
       "step": 220
     },
     {
       "epoch": 0.22106217373636336,
-      "grad_norm": 0.7587497234344482,
+      "grad_norm": 0.7053269147872925,
       "learning_rate": 0.00019732860780526088,
-      "loss": 1.9271,
+      "loss": 1.6882,
       "step": 221
     },
     {
       "epoch": 0.22206245506548716,
-      "grad_norm": 0.8084688186645508,
+      "grad_norm": 0.7072796821594238,
       "learning_rate": 0.00019730448705798239,
-      "loss": 1.8176,
+      "loss": 1.5441,
       "step": 222
     },
     {
       "epoch": 0.223062736394611,
-      "grad_norm": 0.6736906170845032,
+      "grad_norm": 0.6704496145248413,
       "learning_rate": 0.00019728025939067008,
-      "loss": 1.6288,
+      "loss": 1.3791,
       "step": 223
     },
     {
       "epoch": 0.2240630177237348,
-      "grad_norm": 0.7483925819396973,
+      "grad_norm": 0.6141743659973145,
       "learning_rate": 0.00019725592482994583,
-      "loss": 1.8363,
+      "loss": 1.5831,
       "step": 224
     },
     {
       "epoch": 0.22506329905285863,
-      "grad_norm": 1.7995796203613281,
+      "grad_norm": 0.6235673427581787,
       "learning_rate": 0.00019723148340254892,
-      "loss": 1.9072,
+      "loss": 1.6103,
       "step": 225
     },
     {
       "epoch": 0.22606358038198243,
-      "grad_norm": 0.8028881549835205,
+      "grad_norm": 0.6383673548698425,
       "learning_rate": 0.00019720693513533598,
-      "loss": 1.9021,
+      "loss": 1.6284,
       "step": 226
     },
     {
       "epoch": 0.22706386171110624,
-      "grad_norm": 0.9853909015655518,
+      "grad_norm": 0.7666104435920715,
       "learning_rate": 0.00019718228005528122,
-      "loss": 2.0159,
+      "loss": 1.702,
       "step": 227
     },
     {
       "epoch": 0.22806414304023007,
-      "grad_norm": 0.7784947156906128,
+      "grad_norm": 0.6431383490562439,
       "learning_rate": 0.00019715751818947603,
-      "loss": 1.7816,
+      "loss": 1.4571,
       "step": 228
     },
     {
       "epoch": 0.22906442436935387,
-      "grad_norm": 0.7447614669799805,
+      "grad_norm": 0.6177626252174377,
       "learning_rate": 0.0001971326495651293,
-      "loss": 1.654,
+      "loss": 1.4326,
       "step": 229
     },
     {
       "epoch": 0.2300647056984777,
-      "grad_norm": 0.8673064112663269,
+      "grad_norm": 0.7352898120880127,
       "learning_rate": 0.00019710767420956705,
-      "loss": 2.0049,
+      "loss": 1.7427,
       "step": 230
     },
     {
       "epoch": 0.2310649870276015,
-      "grad_norm": 0.8207747936248779,
+      "grad_norm": 0.6259469389915466,
       "learning_rate": 0.0001970825921502328,
-      "loss": 1.9388,
+      "loss": 1.634,
       "step": 231
     },
     {
       "epoch": 0.23206526835672533,
-      "grad_norm": 0.742266058921814,
+      "grad_norm": 0.6699635982513428,
       "learning_rate": 0.0001970574034146871,
-      "loss": 1.7658,
+      "loss": 1.4705,
       "step": 232
     },
     {
       "epoch": 0.23306554968584914,
-      "grad_norm": 0.9097973704338074,
+      "grad_norm": 0.5577033162117004,
       "learning_rate": 0.00019703210803060782,
-      "loss": 1.8023,
+      "loss": 1.5438,
       "step": 233
     },
     {
       "epoch": 0.23406583101497297,
-      "grad_norm": 0.7512438297271729,
+      "grad_norm": 0.6063429117202759,
       "learning_rate": 0.00019700670602579008,
-      "loss": 1.8551,
+      "loss": 1.555,
       "step": 234
     },
     {
       "epoch": 0.23506611234409677,
-      "grad_norm": 0.8303943872451782,
+      "grad_norm": 0.6069104671478271,
       "learning_rate": 0.00019698119742814606,
-      "loss": 1.7723,
+      "loss": 1.5036,
       "step": 235
     },
     {
       "epoch": 0.2360663936732206,
-      "grad_norm": 0.9195139408111572,
+      "grad_norm": 0.6158379316329956,
       "learning_rate": 0.00019695558226570507,
-      "loss": 1.6426,
+      "loss": 1.3741,
       "step": 236
     },
     {
       "epoch": 0.2370666750023444,
-      "grad_norm": 0.7734714150428772,
+      "grad_norm": 0.6366294622421265,
       "learning_rate": 0.00019692986056661356,
-      "loss": 1.7798,
+      "loss": 1.4467,
       "step": 237
     },
     {
       "epoch": 0.23806695633146824,
-      "grad_norm": 0.8759648203849792,
+      "grad_norm": 0.6726595163345337,
       "learning_rate": 0.00019690403235913504,
-      "loss": 1.6465,
+      "loss": 1.3861,
       "step": 238
     },
     {
       "epoch": 0.23906723766059204,
-      "grad_norm": 0.7688003778457642,
+      "grad_norm": 0.6546512842178345,
       "learning_rate": 0.00019687809767165,
-      "loss": 2.0092,
+      "loss": 1.6886,
       "step": 239
     },
     {
       "epoch": 0.24006751898971584,
-      "grad_norm": 0.7398790121078491,
+      "grad_norm": 0.6623121500015259,
       "learning_rate": 0.000196852056532656,
-      "loss": 1.8176,
+      "loss": 1.5925,
       "step": 240
     },
     {
       "epoch": 0.24106780031883968,
-      "grad_norm": 0.8921257853507996,
+      "grad_norm": 0.6577529311180115,
       "learning_rate": 0.00019682590897076752,
-      "loss": 1.7387,
+      "loss": 1.4509,
       "step": 241
     },
     {
       "epoch": 0.24206808164796348,
-      "grad_norm": 0.7939002513885498,
+      "grad_norm": 0.5586327314376831,
       "learning_rate": 0.00019679965501471608,
-      "loss": 1.9417,
+      "loss": 1.6346,
       "step": 242
     },
     {
       "epoch": 0.2430683629770873,
-      "grad_norm": 0.7798025608062744,
+      "grad_norm": 0.6459937691688538,
       "learning_rate": 0.0001967732946933499,
-      "loss": 1.7134,
+      "loss": 1.4129,
       "step": 243
     },
     {
       "epoch": 0.2440686443062111,
-      "grad_norm": 0.8007254600524902,
+      "grad_norm": 0.778732180595398,
       "learning_rate": 0.00019674682803563428,
-      "loss": 1.7387,
+      "loss": 1.5129,
       "step": 244
     },
     {
       "epoch": 0.24506892563533494,
-      "grad_norm": 0.6257696151733398,
+      "grad_norm": 0.7264451384544373,
       "learning_rate": 0.00019672025507065131,
-      "loss": 1.767,
+      "loss": 1.4483,
       "step": 245
     },
     {
       "epoch": 0.24606920696445875,
-      "grad_norm": 0.7942785620689392,
+      "grad_norm": 0.616084635257721,
       "learning_rate": 0.00019669357582759983,
-      "loss": 1.8801,
+      "loss": 1.5947,
       "step": 246
     },
     {
       "epoch": 0.24706948829358258,
-      "grad_norm": 0.7933829426765442,
+      "grad_norm": 0.5911642909049988,
       "learning_rate": 0.00019666679033579552,
-      "loss": 1.9711,
+      "loss": 1.6407,
       "step": 247
     },
     {
       "epoch": 0.24806976962270638,
-      "grad_norm": 0.7489326596260071,
+      "grad_norm": 0.6102796792984009,
       "learning_rate": 0.00019663989862467082,
-      "loss": 1.8038,
+      "loss": 1.5251,
       "step": 248
     },
     {
       "epoch": 0.2490700509518302,
-      "grad_norm": 0.7279101014137268,
+      "grad_norm": 0.5973434448242188,
       "learning_rate": 0.00019661290072377482,
-      "loss": 1.66,
+      "loss": 1.3969,
       "step": 249
     },
     {
       "epoch": 0.25007033228095404,
-      "grad_norm": 0.6823874115943909,
+      "grad_norm": 0.8515523076057434,
       "learning_rate": 0.00019658579666277334,
-      "loss": 1.8064,
+      "loss": 1.5687,
       "step": 250
     },
     {
       "epoch": 0.2510706136100778,
-      "grad_norm": 0.6561273336410522,
+      "grad_norm": 0.5003417134284973,
       "learning_rate": 0.0001965585864714488,
-      "loss": 1.6874,
+      "loss": 1.4102,
       "step": 251
     },
     {
       "epoch": 0.25207089493920165,
-      "grad_norm": 0.6457573175430298,
+      "grad_norm": 0.5215190052986145,
       "learning_rate": 0.00019653127017970034,
-      "loss": 1.4587,
+      "loss": 1.2471,
       "step": 252
     },
     {
       "epoch": 0.2530711762683255,
-      "grad_norm": 0.7649476528167725,
+      "grad_norm": 0.6491619348526001,
       "learning_rate": 0.0001965038478175436,
-      "loss": 1.9811,
+      "loss": 1.6969,
       "step": 253
     },
     {
       "epoch": 0.25407145759744926,
-      "grad_norm": 0.8786829710006714,
+      "grad_norm": 0.6176133155822754,
       "learning_rate": 0.00019647631941511082,
-      "loss": 1.8629,
+      "loss": 1.5351,
       "step": 254
     },
     {
       "epoch": 0.2550717389265731,
-      "grad_norm": 0.7038159966468811,
+      "grad_norm": 0.6913408041000366,
       "learning_rate": 0.0001964486850026507,
-      "loss": 1.6885,
+      "loss": 1.4309,
       "step": 255
     },
     {
       "epoch": 0.2560720202556969,
-      "grad_norm": 0.7255909442901611,
+      "grad_norm": 0.5875718593597412,
       "learning_rate": 0.00019642094461052852,
-      "loss": 1.7335,
+      "loss": 1.4679,
       "step": 256
     },
     {
       "epoch": 0.25707230158482075,
-      "grad_norm": 0.7780727744102478,
+      "grad_norm": 0.6682264804840088,
       "learning_rate": 0.00019639309826922585,
-      "loss": 1.899,
+      "loss": 1.5393,
       "step": 257
     },
     {
       "epoch": 0.2580725829139445,
-      "grad_norm": 0.8533650040626526,
+      "grad_norm": 0.7241432666778564,
       "learning_rate": 0.0001963651460093409,
-      "loss": 1.7711,
+      "loss": 1.4998,
       "step": 258
     },
     {
       "epoch": 0.25907286424306836,
-      "grad_norm": 0.6440068483352661,
+      "grad_norm": 0.5210353136062622,
       "learning_rate": 0.00019633708786158806,
-      "loss": 1.6685,
+      "loss": 1.3837,
       "step": 259
     },
     {
       "epoch": 0.2600731455721922,
-      "grad_norm": 0.6873877048492432,
+      "grad_norm": 0.584020733833313,
       "learning_rate": 0.00019630892385679818,
-      "loss": 1.7502,
+      "loss": 1.4961,
       "step": 260
     },
     {
       "epoch": 0.261073426901316,
-      "grad_norm": 0.7100672721862793,
+      "grad_norm": 0.6708115935325623,
       "learning_rate": 0.00019628065402591845,
-      "loss": 1.7789,
+      "loss": 1.5277,
       "step": 261
     },
     {
       "epoch": 0.2620737082304398,
-      "grad_norm": 0.8447420001029968,
+      "grad_norm": 0.5480003952980042,
       "learning_rate": 0.00019625227840001225,
-      "loss": 1.8577,
+      "loss": 1.556,
       "step": 262
     },
     {
       "epoch": 0.2630739895595636,
-      "grad_norm": 0.767888605594635,
+      "grad_norm": 0.595191478729248,
       "learning_rate": 0.0001962237970102593,
-      "loss": 1.5936,
+      "loss": 1.3514,
       "step": 263
     },
     {
       "epoch": 0.26407427088868746,
-      "grad_norm": 0.6955805420875549,
+      "grad_norm": 0.7332099080085754,
       "learning_rate": 0.0001961952098879555,
-      "loss": 1.7733,
+      "loss": 1.5394,
       "step": 264
     },
     {
       "epoch": 0.26507455221781123,
-      "grad_norm": 0.777740478515625,
+      "grad_norm": 0.596319317817688,
       "learning_rate": 0.00019616651706451287,
-      "loss": 1.6027,
+      "loss": 1.3828,
       "step": 265
     },
     {
       "epoch": 0.26607483354693506,
-      "grad_norm": 0.7691099047660828,
+      "grad_norm": 0.5998026132583618,
       "learning_rate": 0.0001961377185714597,
-      "loss": 1.7457,
+      "loss": 1.4479,
       "step": 266
     },
     {
       "epoch": 0.2670751148760589,
-      "grad_norm": 0.6778420805931091,
+      "grad_norm": 0.6220220923423767,
       "learning_rate": 0.0001961088144404403,
-      "loss": 1.7704,
+      "loss": 1.5121,
       "step": 267
     },
     {
       "epoch": 0.2680753962051827,
-      "grad_norm": 0.7943267226219177,
+      "grad_norm": 0.5865943431854248,
       "learning_rate": 0.00019607980470321505,
-      "loss": 1.9775,
+      "loss": 1.6747,
       "step": 268
     },
     {
       "epoch": 0.2690756775343065,
-      "grad_norm": 0.6660135388374329,
+      "grad_norm": 0.5790852904319763,
       "learning_rate": 0.00019605068939166045,
-      "loss": 1.6556,
+      "loss": 1.3798,
       "step": 269
     },
     {
       "epoch": 0.27007595886343033,
-      "grad_norm": 0.8664935827255249,
+      "grad_norm": 0.6157498955726624,
       "learning_rate": 0.00019602146853776894,
-      "loss": 2.03,
+      "loss": 1.6799,
       "step": 270
     },
     {
       "epoch": 0.27107624019255416,
-      "grad_norm": 0.7783074975013733,
+      "grad_norm": 0.6214422583580017,
       "learning_rate": 0.000195992142173649,
-      "loss": 1.7426,
+      "loss": 1.4782,
       "step": 271
     },
     {
       "epoch": 0.272076521521678,
-      "grad_norm": 0.7470223903656006,
+      "grad_norm": 0.6460129618644714,
       "learning_rate": 0.0001959627103315249,
-      "loss": 1.7284,
+      "loss": 1.4874,
       "step": 272
     },
     {
       "epoch": 0.27307680285080177,
-      "grad_norm": 0.7284931540489197,
+      "grad_norm": 0.5928930640220642,
       "learning_rate": 0.00019593317304373705,
-      "loss": 1.6977,
+      "loss": 1.4557,
       "step": 273
     },
     {
       "epoch": 0.2740770841799256,
-      "grad_norm": 0.7201762795448303,
+      "grad_norm": 0.5123687982559204,
       "learning_rate": 0.00019590353034274144,
-      "loss": 1.7184,
+      "loss": 1.445,
       "step": 274
     },
     {
       "epoch": 0.27507736550904943,
-      "grad_norm": 0.6756151914596558,
+      "grad_norm": 0.607455313205719,
       "learning_rate": 0.00019587378226111014,
-      "loss": 1.7276,
+      "loss": 1.4468,
       "step": 275
     },
     {
       "epoch": 0.27607764683817326,
-      "grad_norm": 0.6784201860427856,
+      "grad_norm": 0.6108120083808899,
       "learning_rate": 0.00019584392883153088,
-      "loss": 1.642,
+      "loss": 1.3834,
       "step": 276
     },
     {
       "epoch": 0.27707792816729704,
-      "grad_norm": 0.7387176752090454,
+      "grad_norm": 0.680404543876648,
       "learning_rate": 0.00019581397008680717,
-      "loss": 1.7911,
+      "loss": 1.5094,
       "step": 277
     },
     {
       "epoch": 0.27807820949642087,
-      "grad_norm": 0.9367021918296814,
+      "grad_norm": 0.6419563889503479,
       "learning_rate": 0.00019578390605985826,
-      "loss": 2.0034,
+      "loss": 1.6933,
       "step": 278
     },
     {
       "epoch": 0.2790784908255447,
-      "grad_norm": 0.803698718547821,
+      "grad_norm": 0.5788853764533997,
       "learning_rate": 0.00019575373678371909,
-      "loss": 1.7907,
+      "loss": 1.4754,
       "step": 279
     },
     {
       "epoch": 0.2800787721546685,
-      "grad_norm": 0.7324479818344116,
+      "grad_norm": 0.5943770408630371,
       "learning_rate": 0.00019572346229154025,
-      "loss": 1.5539,
+      "loss": 1.2949,
       "step": 280
     },
     {
       "epoch": 0.2810790534837923,
-      "grad_norm": 0.7107382416725159,
+      "grad_norm": 0.5997135043144226,
       "learning_rate": 0.00019569308261658787,
-      "loss": 1.838,
+      "loss": 1.5365,
       "step": 281
     },
     {
       "epoch": 0.28207933481291614,
-      "grad_norm": 0.8698626756668091,
+      "grad_norm": 0.692401647567749,
       "learning_rate": 0.00019566259779224378,
-      "loss": 1.7433,
+      "loss": 1.4946,
       "step": 282
     },
     {
       "epoch": 0.28307961614203997,
-      "grad_norm": 0.7804028391838074,
+      "grad_norm": 0.5856708884239197,
       "learning_rate": 0.00019563200785200526,
-      "loss": 1.7161,
+      "loss": 1.426,
       "step": 283
     },
     {
       "epoch": 0.28407989747116374,
-      "grad_norm": 0.8762909173965454,
+      "grad_norm": 1.2516822814941406,
       "learning_rate": 0.00019560131282948516,
-      "loss": 1.8031,
+      "loss": 1.5119,
       "step": 284
     },
     {
       "epoch": 0.2850801788002876,
-      "grad_norm": 0.8252436518669128,
+      "grad_norm": 0.6360501050949097,
       "learning_rate": 0.0001955705127584117,
-      "loss": 1.6434,
+      "loss": 1.3916,
       "step": 285
     },
     {
       "epoch": 0.2860804601294114,
-      "grad_norm": 0.8220797181129456,
+      "grad_norm": 0.6822036504745483,
       "learning_rate": 0.00019553960767262863,
-      "loss": 1.8522,
+      "loss": 1.5565,
       "step": 286
     },
     {
       "epoch": 0.28708074145853524,
-      "grad_norm": 0.7883003950119019,
+      "grad_norm": 0.6973714828491211,
       "learning_rate": 0.00019550859760609503,
-      "loss": 1.8245,
+      "loss": 1.5559,
       "step": 287
     },
     {
       "epoch": 0.288081022787659,
-      "grad_norm": 0.9208703637123108,
+      "grad_norm": 0.6595618724822998,
       "learning_rate": 0.00019547748259288536,
-      "loss": 1.8877,
+      "loss": 1.5824,
       "step": 288
     },
     {
       "epoch": 0.28908130411678284,
-      "grad_norm": 0.8452202677726746,
+      "grad_norm": 0.5625808238983154,
       "learning_rate": 0.0001954462626671894,
-      "loss": 1.554,
+      "loss": 1.2669,
       "step": 289
     },
     {
       "epoch": 0.2900815854459067,
-      "grad_norm": 0.82865971326828,
+      "grad_norm": 0.6318663358688354,
       "learning_rate": 0.0001954149378633122,
-      "loss": 1.655,
+      "loss": 1.3896,
       "step": 290
     },
     {
       "epoch": 0.29108186677503045,
-      "grad_norm": 0.7871205806732178,
+      "grad_norm": 0.6655906438827515,
       "learning_rate": 0.00019538350821567404,
-      "loss": 1.621,
+      "loss": 1.3889,
       "step": 291
     },
     {
       "epoch": 0.2920821481041543,
-      "grad_norm": 0.8288848996162415,
+      "grad_norm": 0.5947337746620178,
       "learning_rate": 0.00019535197375881045,
-      "loss": 1.9277,
+      "loss": 1.6112,
       "step": 292
     },
     {
       "epoch": 0.2930824294332781,
-      "grad_norm": 0.7275516986846924,
+      "grad_norm": 0.6139295101165771,
       "learning_rate": 0.00019532033452737205,
-      "loss": 1.7949,
+      "loss": 1.5185,
       "step": 293
     },
     {
       "epoch": 0.29408271076240194,
-      "grad_norm": 0.7424570322036743,
+      "grad_norm": 0.579953670501709,
       "learning_rate": 0.00019528859055612468,
-      "loss": 1.6407,
+      "loss": 1.3874,
       "step": 294
     },
     {
       "epoch": 0.2950829920915257,
-      "grad_norm": 0.7031363248825073,
+      "grad_norm": 0.6101506352424622,
       "learning_rate": 0.0001952567418799492,
-      "loss": 1.8793,
+      "loss": 1.5965,
       "step": 295
     },
     {
       "epoch": 0.29608327342064955,
-      "grad_norm": 0.7190185189247131,
+      "grad_norm": 0.6393965482711792,
       "learning_rate": 0.00019522478853384155,
-      "loss": 1.6759,
+      "loss": 1.4124,
       "step": 296
     },
     {
       "epoch": 0.2970835547497734,
-      "grad_norm": 0.7270736694335938,
+      "grad_norm": 0.6147856712341309,
       "learning_rate": 0.00019519273055291266,
-      "loss": 1.6351,
+      "loss": 1.3776,
       "step": 297
     },
     {
       "epoch": 0.2980838360788972,
-      "grad_norm": 0.8894152641296387,
+      "grad_norm": 0.6056416630744934,
       "learning_rate": 0.00019516056797238846,
-      "loss": 1.7908,
+      "loss": 1.4453,
       "step": 298
     },
     {
       "epoch": 0.299084117408021,
-      "grad_norm": 0.9089106321334839,
+      "grad_norm": 0.6705831289291382,
       "learning_rate": 0.00019512830082760987,
-      "loss": 1.6018,
+      "loss": 1.3248,
       "step": 299
     },
     {
       "epoch": 0.3000843987371448,
-      "grad_norm": 0.8772429823875427,
+      "grad_norm": 0.6664314866065979,
       "learning_rate": 0.00019509592915403255,
-      "loss": 1.8474,
+      "loss": 1.5865,
       "step": 300
     },
     {
       "epoch": 0.30108468006626865,
-      "grad_norm": 0.8244933485984802,
+      "grad_norm": 0.5325604677200317,
       "learning_rate": 0.00019506345298722717,
-      "loss": 1.4324,
+      "loss": 1.0646,
       "step": 301
     },
     {
       "epoch": 0.3020849613953925,
-      "grad_norm": 0.7283012866973877,
+      "grad_norm": 0.589242160320282,
       "learning_rate": 0.00019503087236287913,
-      "loss": 1.5115,
+      "loss": 1.2297,
       "step": 302
     },
     {
       "epoch": 0.30308524272451626,
-      "grad_norm": 0.7721333503723145,
+      "grad_norm": 0.5677699446678162,
       "learning_rate": 0.00019499818731678873,
-      "loss": 1.6728,
+      "loss": 1.3961,
       "step": 303
     },
     {
       "epoch": 0.3040855240536401,
-      "grad_norm": 0.7579306960105896,
+      "grad_norm": 0.5676394701004028,
       "learning_rate": 0.00019496539788487082,
-      "loss": 1.5927,
+      "loss": 1.3276,
       "step": 304
     },
     {
       "epoch": 0.3050858053827639,
-      "grad_norm": 0.9054704308509827,
+      "grad_norm": 0.7280861139297485,
       "learning_rate": 0.0001949325041031551,
-      "loss": 1.9027,
+      "loss": 1.6731,
       "step": 305
     },
     {
       "epoch": 0.3060860867118877,
-      "grad_norm": 0.7023262977600098,
+      "grad_norm": 0.690636396408081,
       "learning_rate": 0.0001948995060077859,
-      "loss": 1.7705,
+      "loss": 1.5443,
       "step": 306
     },
     {
       "epoch": 0.3070863680410115,
-      "grad_norm": 0.7942065000534058,
+      "grad_norm": 0.611426055431366,
       "learning_rate": 0.0001948664036350221,
-      "loss": 1.8269,
+      "loss": 1.5827,
       "step": 307
     },
     {
       "epoch": 0.30808664937013536,
-      "grad_norm": 0.9305068850517273,
+      "grad_norm": 0.7112497091293335,
       "learning_rate": 0.00019483319702123732,
-      "loss": 1.8247,
+      "loss": 1.5401,
       "step": 308
     },
     {
       "epoch": 0.3090869306992592,
-      "grad_norm": 0.814664900302887,
+      "grad_norm": 0.6598275303840637,
       "learning_rate": 0.00019479988620291956,
-      "loss": 1.9179,
+      "loss": 1.6432,
       "step": 309
     },
     {
       "epoch": 0.31008721202838296,
-      "grad_norm": 0.6418014764785767,
+      "grad_norm": 0.5019932985305786,
       "learning_rate": 0.00019476647121667137,
-      "loss": 1.5011,
+      "loss": 1.2561,
       "step": 310
     },
     {
       "epoch": 0.3110874933575068,
-      "grad_norm": 0.7911447882652283,
+      "grad_norm": 0.7777897715568542,
       "learning_rate": 0.00019473295209920983,
-      "loss": 1.857,
+      "loss": 1.6118,
       "step": 311
     },
     {
       "epoch": 0.3120877746866306,
-      "grad_norm": 0.7792949676513672,
+      "grad_norm": 0.6028640866279602,
       "learning_rate": 0.00019469932888736632,
-      "loss": 1.7279,
+      "loss": 1.4682,
       "step": 312
     },
     {
       "epoch": 0.31308805601575446,
-      "grad_norm": 0.7579171657562256,
+      "grad_norm": 0.554381251335144,
       "learning_rate": 0.00019466560161808674,
-      "loss": 1.6902,
+      "loss": 1.4179,
       "step": 313
     },
     {
       "epoch": 0.31408833734487823,
-      "grad_norm": 0.7052372694015503,
+      "grad_norm": 0.6212736368179321,
       "learning_rate": 0.00019463177032843124,
-      "loss": 1.7302,
+      "loss": 1.4327,
       "step": 314
     },
     {
       "epoch": 0.31508861867400206,
-      "grad_norm": 0.7188624143600464,
+      "grad_norm": 0.6829814910888672,
       "learning_rate": 0.00019459783505557424,
-      "loss": 1.7338,
+      "loss": 1.4455,
       "step": 315
     },
     {
       "epoch": 0.3160889000031259,
-      "grad_norm": 0.6057978272438049,
+      "grad_norm": 0.5808065533638,
       "learning_rate": 0.00019456379583680452,
-      "loss": 1.6123,
+      "loss": 1.3583,
       "step": 316
     },
     {
       "epoch": 0.31708918133224967,
-      "grad_norm": 0.8339365720748901,
+      "grad_norm": 0.6354159712791443,
       "learning_rate": 0.000194529652709525,
-      "loss": 1.9765,
+      "loss": 1.6916,
       "step": 317
     },
     {
       "epoch": 0.3180894626613735,
-      "grad_norm": 0.8524260520935059,
+      "grad_norm": 0.6299159526824951,
       "learning_rate": 0.00019449540571125286,
-      "loss": 1.6803,
+      "loss": 1.47,
       "step": 318
     },
     {
       "epoch": 0.31908974399049733,
-      "grad_norm": 0.7035975456237793,
+      "grad_norm": 0.6222877502441406,
       "learning_rate": 0.00019446105487961926,
-      "loss": 1.5792,
+      "loss": 1.4137,
       "step": 319
     },
     {
       "epoch": 0.32009002531962116,
-      "grad_norm": 0.7894249558448792,
+      "grad_norm": 0.5995916724205017,
       "learning_rate": 0.0001944266002523696,
-      "loss": 1.6326,
+      "loss": 1.3679,
       "step": 320
     },
     {
       "epoch": 0.32109030664874494,
-      "grad_norm": 0.7716989517211914,
+      "grad_norm": 0.599814236164093,
       "learning_rate": 0.0001943920418673633,
-      "loss": 1.6871,
+      "loss": 1.4075,
       "step": 321
     },
     {
       "epoch": 0.32209058797786877,
-      "grad_norm": 0.7914933562278748,
+      "grad_norm": 0.5409269332885742,
       "learning_rate": 0.00019435737976257377,
-      "loss": 1.7148,
+      "loss": 1.4289,
       "step": 322
     },
     {
       "epoch": 0.3230908693069926,
-      "grad_norm": 0.7113205790519714,
+      "grad_norm": 0.5298951864242554,
       "learning_rate": 0.00019432261397608834,
-      "loss": 1.5236,
+      "loss": 1.2834,
       "step": 323
     },
     {
       "epoch": 0.32409115063611643,
-      "grad_norm": 0.8609917163848877,
+      "grad_norm": 0.7196112871170044,
       "learning_rate": 0.00019428774454610843,
-      "loss": 1.8101,
+      "loss": 1.4845,
       "step": 324
     },
     {
       "epoch": 0.3250914319652402,
-      "grad_norm": 0.7319685220718384,
+      "grad_norm": 0.5605450868606567,
       "learning_rate": 0.00019425277151094913,
-      "loss": 1.7712,
+      "loss": 1.4575,
       "step": 325
     },
     {
       "epoch": 0.32609171329436404,
-      "grad_norm": 0.6478747725486755,
+      "grad_norm": 0.573080837726593,
       "learning_rate": 0.00019421769490903957,
-      "loss": 1.8535,
+      "loss": 1.5757,
       "step": 326
     },
     {
       "epoch": 0.32709199462348787,
-      "grad_norm": 0.7025763392448425,
+      "grad_norm": 0.5017902851104736,
       "learning_rate": 0.0001941825147789225,
-      "loss": 1.9213,
+      "loss": 1.5794,
       "step": 327
     },
     {
       "epoch": 0.32809227595261165,
-      "grad_norm": 0.7595239877700806,
+      "grad_norm": 0.643267810344696,
       "learning_rate": 0.00019414723115925456,
-      "loss": 1.7449,
+      "loss": 1.4903,
       "step": 328
     },
     {
       "epoch": 0.3290925572817355,
-      "grad_norm": 0.7728105783462524,
+      "grad_norm": 0.6522070169448853,
       "learning_rate": 0.0001941118440888061,
-      "loss": 1.8821,
+      "loss": 1.5907,
       "step": 329
     },
     {
       "epoch": 0.3300928386108593,
-      "grad_norm": 0.7430977821350098,
+      "grad_norm": 0.6496105790138245,
       "learning_rate": 0.0001940763536064611,
-      "loss": 1.6904,
+      "loss": 1.4225,
       "step": 330
     },
     {
       "epoch": 0.33109311993998314,
-      "grad_norm": 0.7909367680549622,
+      "grad_norm": 0.6011468768119812,
       "learning_rate": 0.00019404075975121716,
-      "loss": 1.7899,
+      "loss": 1.5022,
       "step": 331
     },
     {
       "epoch": 0.3320934012691069,
-      "grad_norm": 0.7561226487159729,
+      "grad_norm": 0.6327878832817078,
       "learning_rate": 0.0001940050625621855,
-      "loss": 1.7746,
+      "loss": 1.468,
       "step": 332
     },
     {
       "epoch": 0.33309368259823074,
-      "grad_norm": 0.7602452635765076,
+      "grad_norm": 0.6187490820884705,
       "learning_rate": 0.00019396926207859084,
-      "loss": 1.7909,
+      "loss": 1.5183,
       "step": 333
     },
     {
       "epoch": 0.3340939639273546,
-      "grad_norm": 0.8194379806518555,
+      "grad_norm": 0.7625093460083008,
       "learning_rate": 0.0001939333583397715,
-      "loss": 1.7039,
+      "loss": 1.4813,
       "step": 334
     },
     {
       "epoch": 0.3350942452564784,
-      "grad_norm": 0.7036342024803162,
+      "grad_norm": 0.5286359190940857,
       "learning_rate": 0.00019389735138517915,
-      "loss": 1.6663,
+      "loss": 1.3674,
       "step": 335
     },
     {
       "epoch": 0.3360945265856022,
-      "grad_norm": 0.8429521918296814,
+      "grad_norm": 0.5798503160476685,
       "learning_rate": 0.00019386124125437895,
-      "loss": 1.589,
+      "loss": 1.3016,
       "step": 336
     },
     {
       "epoch": 0.337094807914726,
-      "grad_norm": 0.7271071076393127,
+      "grad_norm": 0.48794126510620117,
       "learning_rate": 0.00019382502798704935,
-      "loss": 1.646,
+      "loss": 1.3642,
       "step": 337
     },
     {
       "epoch": 0.33809508924384984,
-      "grad_norm": 0.7862086892127991,
+      "grad_norm": 0.7394312620162964,
       "learning_rate": 0.00019378871162298227,
-      "loss": 1.6085,
+      "loss": 1.327,
       "step": 338
     },
     {
       "epoch": 0.3390953705729737,
-      "grad_norm": 0.676815390586853,
+      "grad_norm": 0.5598319172859192,
       "learning_rate": 0.00019375229220208276,
-      "loss": 1.7335,
+      "loss": 1.4247,
       "step": 339
     },
     {
       "epoch": 0.34009565190209745,
-      "grad_norm": 0.8916042447090149,
+      "grad_norm": 0.6099628806114197,
       "learning_rate": 0.00019371576976436917,
-      "loss": 1.7914,
+      "loss": 1.4906,
       "step": 340
     },
     {
       "epoch": 0.3410959332312213,
-      "grad_norm": 0.7913751006126404,
+      "grad_norm": 0.6749781370162964,
       "learning_rate": 0.00019367914434997312,
-      "loss": 1.6031,
+      "loss": 1.367,
       "step": 341
     },
     {
       "epoch": 0.3420962145603451,
-      "grad_norm": 0.7409866452217102,
+      "grad_norm": 0.7721238136291504,
       "learning_rate": 0.00019364241599913924,
-      "loss": 1.6525,
+      "loss": 1.4464,
       "step": 342
     },
     {
       "epoch": 0.3430964958894689,
-      "grad_norm": 0.7472705841064453,
+      "grad_norm": 0.5762369632720947,
       "learning_rate": 0.0001936055847522254,
-      "loss": 1.6716,
+      "loss": 1.409,
       "step": 343
     },
     {
       "epoch": 0.3440967772185927,
-      "grad_norm": 0.7030773758888245,
+      "grad_norm": 0.6960498690605164,
       "learning_rate": 0.00019356865064970244,
-      "loss": 1.7134,
+      "loss": 1.3907,
       "step": 344
     },
     {
       "epoch": 0.34509705854771655,
-      "grad_norm": 0.6609564423561096,
+      "grad_norm": 0.5805984735488892,
       "learning_rate": 0.0001935316137321543,
-      "loss": 1.7127,
+      "loss": 1.4539,
       "step": 345
     },
     {
       "epoch": 0.3460973398768404,
-      "grad_norm": 0.7811393141746521,
+      "grad_norm": 0.5686045289039612,
       "learning_rate": 0.00019349447404027782,
-      "loss": 1.75,
+      "loss": 1.4493,
       "step": 346
     },
     {
       "epoch": 0.34709762120596416,
-      "grad_norm": 0.6980521082878113,
+      "grad_norm": 0.5448501706123352,
       "learning_rate": 0.00019345723161488283,
-      "loss": 1.82,
+      "loss": 1.5633,
       "step": 347
     },
     {
       "epoch": 0.348097902535088,
-      "grad_norm": 0.749796986579895,
+      "grad_norm": 0.6388784050941467,
       "learning_rate": 0.000193419886496892,
-      "loss": 1.9755,
+      "loss": 1.7179,
       "step": 348
     },
     {
       "epoch": 0.3490981838642118,
-      "grad_norm": 0.9486667513847351,
+      "grad_norm": 0.5240457653999329,
       "learning_rate": 0.00019338243872734086,
-      "loss": 1.7047,
+      "loss": 1.4411,
       "step": 349
     },
     {
       "epoch": 0.35009846519333565,
-      "grad_norm": 0.8086081147193909,
+      "grad_norm": 0.5460641384124756,
       "learning_rate": 0.00019334488834737775,
-      "loss": 1.661,
+      "loss": 1.361,
       "step": 350
     },
     {
       "epoch": 0.3510987465224594,
-      "grad_norm": 0.700549840927124,
+      "grad_norm": 0.5495695471763611,
       "learning_rate": 0.00019330723539826375,
-      "loss": 1.8696,
+      "loss": 1.5891,
       "step": 351
     },
     {
       "epoch": 0.35209902785158326,
-      "grad_norm": 0.7465476393699646,
+      "grad_norm": 0.5618153214454651,
       "learning_rate": 0.00019326947992137262,
-      "loss": 1.5444,
+      "loss": 1.3084,
       "step": 352
     },
     {
       "epoch": 0.3530993091807071,
-      "grad_norm": 0.7370999455451965,
+      "grad_norm": 0.5603707432746887,
       "learning_rate": 0.00019323162195819082,
-      "loss": 1.8805,
+      "loss": 1.5732,
       "step": 353
     },
     {
       "epoch": 0.35409959050983086,
-      "grad_norm": 0.719359278678894,
+      "grad_norm": 0.5732563138008118,
       "learning_rate": 0.0001931936615503174,
-      "loss": 1.8022,
+      "loss": 1.5045,
       "step": 354
     },
     {
       "epoch": 0.3550998718389547,
-      "grad_norm": 0.7301434278488159,
+      "grad_norm": 0.5997583866119385,
       "learning_rate": 0.000193155598739464,
-      "loss": 1.6984,
+      "loss": 1.4175,
       "step": 355
     },
     {
       "epoch": 0.3561001531680785,
-      "grad_norm": 0.7191399335861206,
+      "grad_norm": 0.5769765377044678,
       "learning_rate": 0.0001931174335674547,
-      "loss": 1.7229,
+      "loss": 1.4834,
       "step": 356
     },
     {
       "epoch": 0.35710043449720236,
-      "grad_norm": 0.7471932768821716,
+      "grad_norm": 0.5902683138847351,
       "learning_rate": 0.0001930791660762262,
-      "loss": 1.7408,
+      "loss": 1.4664,
       "step": 357
     },
     {
       "epoch": 0.35810071582632613,
-      "grad_norm": 0.8197934031486511,
+      "grad_norm": 0.6354758143424988,
       "learning_rate": 0.00019304079630782752,
-      "loss": 1.6938,
+      "loss": 1.3891,
       "step": 358
     },
     {
       "epoch": 0.35910099715544996,
-      "grad_norm": 0.7408166527748108,
+      "grad_norm": 0.6018317341804504,
       "learning_rate": 0.0001930023243044201,
-      "loss": 1.7798,
+      "loss": 1.4514,
       "step": 359
     },
     {
       "epoch": 0.3601012784845738,
-      "grad_norm": 0.7525373101234436,
+      "grad_norm": 0.5409123301506042,
       "learning_rate": 0.00019296375010827773,
-      "loss": 1.711,
+      "loss": 1.4708,
       "step": 360
     },
     {
       "epoch": 0.3611015598136976,
-      "grad_norm": 0.6712046265602112,
+      "grad_norm": 0.5457523465156555,
       "learning_rate": 0.00019292507376178643,
-      "loss": 1.8157,
+      "loss": 1.4988,
       "step": 361
     },
     {
       "epoch": 0.3621018411428214,
-      "grad_norm": 0.6712916493415833,
+      "grad_norm": 0.626768946647644,
       "learning_rate": 0.00019288629530744454,
-      "loss": 1.8707,
+      "loss": 1.5722,
       "step": 362
     },
     {
       "epoch": 0.36310212247194523,
-      "grad_norm": 0.6127772331237793,
+      "grad_norm": 0.566554069519043,
       "learning_rate": 0.0001928474147878626,
-      "loss": 1.4743,
+      "loss": 1.2135,
       "step": 363
     },
     {
       "epoch": 0.36410240380106906,
-      "grad_norm": 0.910310685634613,
+      "grad_norm": 0.7327786684036255,
       "learning_rate": 0.0001928084322457632,
-      "loss": 1.7956,
+      "loss": 1.5245,
       "step": 364
     },
     {
       "epoch": 0.3651026851301929,
-      "grad_norm": 0.6267688870429993,
+      "grad_norm": 0.5205698609352112,
       "learning_rate": 0.00019276934772398114,
-      "loss": 1.4664,
+      "loss": 1.2068,
       "step": 365
     },
     {
       "epoch": 0.36610296645931667,
-      "grad_norm": 0.8317943811416626,
+      "grad_norm": 1.0956753492355347,
       "learning_rate": 0.00019273016126546323,
-      "loss": 1.853,
+      "loss": 1.5044,
       "step": 366
     },
     {
       "epoch": 0.3671032477884405,
-      "grad_norm": 0.7581344842910767,
+      "grad_norm": 0.6484043598175049,
       "learning_rate": 0.00019269087291326833,
-      "loss": 1.9236,
+      "loss": 1.6369,
       "step": 367
     },
     {
       "epoch": 0.36810352911756433,
-      "grad_norm": 0.9311390519142151,
+      "grad_norm": 0.6363429427146912,
       "learning_rate": 0.00019265148271056722,
-      "loss": 1.7019,
+      "loss": 1.4338,
       "step": 368
     },
     {
       "epoch": 0.3691038104466881,
-      "grad_norm": 0.9513958096504211,
+      "grad_norm": 0.6295244693756104,
       "learning_rate": 0.0001926119907006426,
-      "loss": 1.7617,
+      "loss": 1.4701,
       "step": 369
     },
     {
       "epoch": 0.37010409177581194,
-      "grad_norm": 0.7407613396644592,
+      "grad_norm": 0.6013259887695312,
       "learning_rate": 0.00019257239692688907,
-      "loss": 2.1057,
+      "loss": 1.7629,
       "step": 370
     },
     {
       "epoch": 0.37110437310493577,
-      "grad_norm": 0.7530227899551392,
+      "grad_norm": 0.6949493885040283,
       "learning_rate": 0.00019253270143281296,
-      "loss": 1.9844,
+      "loss": 1.6713,
       "step": 371
     },
     {
       "epoch": 0.3721046544340596,
-      "grad_norm": 0.6733037233352661,
+      "grad_norm": 0.6933801174163818,
       "learning_rate": 0.00019249290426203252,
-      "loss": 1.8939,
+      "loss": 1.6131,
       "step": 372
     },
     {
       "epoch": 0.3731049357631834,
-      "grad_norm": 0.7037007808685303,
+      "grad_norm": 0.5847527384757996,
       "learning_rate": 0.0001924530054582776,
-      "loss": 1.6897,
+      "loss": 1.3968,
       "step": 373
     },
     {
       "epoch": 0.3741052170923072,
-      "grad_norm": 0.7689145803451538,
+      "grad_norm": 0.6053057312965393,
       "learning_rate": 0.0001924130050653898,
-      "loss": 1.6544,
+      "loss": 1.3311,
       "step": 374
     },
     {
       "epoch": 0.37510549842143104,
-      "grad_norm": 0.7100968360900879,
+      "grad_norm": 0.5513793230056763,
       "learning_rate": 0.00019237290312732226,
-      "loss": 1.7932,
+      "loss": 1.5063,
       "step": 375
     },
     {
       "epoch": 0.37610577975055487,
-      "grad_norm": 0.7645193338394165,
+      "grad_norm": 0.5859197378158569,
       "learning_rate": 0.00019233269968813984,
-      "loss": 1.6576,
+      "loss": 1.3556,
       "step": 376
     },
     {
       "epoch": 0.37710606107967864,
-      "grad_norm": 0.6831678152084351,
+      "grad_norm": 0.5623495578765869,
       "learning_rate": 0.00019229239479201876,
-      "loss": 1.6675,
+      "loss": 1.3859,
       "step": 377
     },
     {
       "epoch": 0.3781063424088025,
-      "grad_norm": 0.8677794337272644,
+      "grad_norm": 0.602118968963623,
       "learning_rate": 0.0001922519884832469,
-      "loss": 1.6064,
+      "loss": 1.334,
       "step": 378
     },
     {
       "epoch": 0.3791066237379263,
-      "grad_norm": 0.6727691888809204,
+      "grad_norm": 0.5212380886077881,
       "learning_rate": 0.0001922114808062234,
-      "loss": 1.6249,
+      "loss": 1.401,
       "step": 379
     },
     {
       "epoch": 0.3801069050670501,
-      "grad_norm": 0.6588670611381531,
+      "grad_norm": 0.4969455599784851,
       "learning_rate": 0.00019217087180545893,
-      "loss": 1.5541,
+      "loss": 1.2292,
       "step": 380
     },
     {
       "epoch": 0.3811071863961739,
-      "grad_norm": 0.7666369080543518,
+      "grad_norm": 0.578629732131958,
       "learning_rate": 0.0001921301615255754,
-      "loss": 1.7806,
+      "loss": 1.5015,
       "step": 381
     },
     {
       "epoch": 0.38210746772529774,
-      "grad_norm": 0.6465156078338623,
+      "grad_norm": 0.593053936958313,
       "learning_rate": 0.0001920893500113061,
-      "loss": 1.512,
+      "loss": 1.302,
       "step": 382
     },
     {
       "epoch": 0.3831077490544216,
-      "grad_norm": 0.7854346632957458,
+      "grad_norm": 0.5832563638687134,
       "learning_rate": 0.00019204843730749547,
-      "loss": 1.6857,
+      "loss": 1.3695,
       "step": 383
     },
     {
       "epoch": 0.38410803038354535,
-      "grad_norm": 0.6625111103057861,
+      "grad_norm": 0.5608510375022888,
       "learning_rate": 0.00019200742345909915,
-      "loss": 1.7033,
+      "loss": 1.3792,
       "step": 384
     },
     {
       "epoch": 0.3851083117126692,
-      "grad_norm": 0.7273709177970886,
+      "grad_norm": 0.5337334275245667,
       "learning_rate": 0.00019196630851118398,
-      "loss": 1.665,
+      "loss": 1.4163,
       "step": 385
     },
     {
       "epoch": 0.386108593041793,
-      "grad_norm": 0.6861465573310852,
+      "grad_norm": 0.5460125803947449,
       "learning_rate": 0.0001919250925089278,
-      "loss": 1.5028,
+      "loss": 1.2439,
       "step": 386
     },
     {
       "epoch": 0.38710887437091684,
-      "grad_norm": 0.845456063747406,
+      "grad_norm": 0.6217851638793945,
       "learning_rate": 0.00019188377549761963,
-      "loss": 1.967,
+      "loss": 1.6428,
       "step": 387
     },
     {
       "epoch": 0.3881091557000406,
-      "grad_norm": 0.6481165289878845,
+      "grad_norm": 0.7154502868652344,
       "learning_rate": 0.00019184235752265928,
-      "loss": 1.6053,
+      "loss": 1.3468,
       "step": 388
     },
     {
       "epoch": 0.38910943702916445,
-      "grad_norm": 0.6312947273254395,
+      "grad_norm": 0.5044635534286499,
       "learning_rate": 0.00019180083862955772,
-      "loss": 1.4427,
+      "loss": 1.1877,
       "step": 389
     },
     {
       "epoch": 0.3901097183582883,
-      "grad_norm": 0.7874154448509216,
+      "grad_norm": 0.5755971074104309,
       "learning_rate": 0.00019175921886393666,
-      "loss": 1.6099,
+      "loss": 1.3475,
       "step": 390
     },
     {
       "epoch": 0.39110999968741206,
-      "grad_norm": 0.6839481592178345,
+      "grad_norm": 0.6121137738227844,
       "learning_rate": 0.00019171749827152869,
-      "loss": 1.7004,
+      "loss": 1.4342,
       "step": 391
     },
     {
       "epoch": 0.3921102810165359,
-      "grad_norm": 0.7239277362823486,
+      "grad_norm": 0.5615536570549011,
       "learning_rate": 0.0001916756768981772,
-      "loss": 1.8813,
+      "loss": 1.5471,
       "step": 392
     },
     {
       "epoch": 0.3931105623456597,
-      "grad_norm": 0.8241100311279297,
+      "grad_norm": 0.6527026295661926,
       "learning_rate": 0.00019163375478983632,
-      "loss": 1.9443,
+      "loss": 1.6363,
       "step": 393
     },
     {
       "epoch": 0.39411084367478355,
-      "grad_norm": 0.7401999235153198,
+      "grad_norm": 0.6465044617652893,
       "learning_rate": 0.00019159173199257085,
-      "loss": 1.6663,
+      "loss": 1.3823,
       "step": 394
     },
     {
       "epoch": 0.3951111250039073,
-      "grad_norm": 0.8297036290168762,
+      "grad_norm": 0.5620000958442688,
       "learning_rate": 0.00019154960855255628,
-      "loss": 1.8012,
+      "loss": 1.5418,
       "step": 395
     },
     {
       "epoch": 0.39611140633303116,
-      "grad_norm": 0.9661216735839844,
+      "grad_norm": 0.7090588808059692,
       "learning_rate": 0.0001915073845160786,
-      "loss": 1.7007,
+      "loss": 1.4593,
       "step": 396
     },
     {
       "epoch": 0.397111687662155,
-      "grad_norm": 1.4041926860809326,
+      "grad_norm": 0.6644489169120789,
       "learning_rate": 0.00019146505992953446,
-      "loss": 1.7232,
+      "loss": 1.4236,
       "step": 397
     },
     {
       "epoch": 0.3981119689912788,
-      "grad_norm": 0.8469036221504211,
+      "grad_norm": 0.6038135886192322,
       "learning_rate": 0.00019142263483943085,
-      "loss": 1.4479,
+      "loss": 1.1805,
       "step": 398
     },
     {
       "epoch": 0.3991122503204026,
-      "grad_norm": 0.9476561546325684,
+      "grad_norm": 0.6746726036071777,
       "learning_rate": 0.00019138010929238534,
-      "loss": 1.8572,
+      "loss": 1.5264,
       "step": 399
     },
     {
       "epoch": 0.4001125316495264,
-      "grad_norm": 0.7196705937385559,
+      "grad_norm": 0.5871374607086182,
       "learning_rate": 0.00019133748333512575,
-      "loss": 1.6184,
+      "loss": 1.3709,
       "step": 400
     },
     {
       "epoch": 0.40111281297865026,
-      "grad_norm": 0.8957480192184448,
+      "grad_norm": 0.5743412375450134,
       "learning_rate": 0.00019129475701449035,
-      "loss": 1.762,
+      "loss": 1.4677,
       "step": 401
     },
     {
       "epoch": 0.4021130943077741,
-      "grad_norm": 0.7850635647773743,
+      "grad_norm": 0.6184396743774414,
       "learning_rate": 0.0001912519303774276,
-      "loss": 1.6764,
+      "loss": 1.4228,
       "step": 402
     },
     {
       "epoch": 0.40311337563689786,
-      "grad_norm": 0.7579814791679382,
+      "grad_norm": 0.5872434973716736,
       "learning_rate": 0.0001912090034709963,
-      "loss": 1.6231,
+      "loss": 1.3495,
       "step": 403
     },
     {
       "epoch": 0.4041136569660217,
-      "grad_norm": 0.7173107266426086,
+      "grad_norm": 0.6500155925750732,
       "learning_rate": 0.00019116597634236525,
-      "loss": 1.7107,
+      "loss": 1.4315,
       "step": 404
     },
     {
       "epoch": 0.4051139382951455,
-      "grad_norm": 0.7832950353622437,
+      "grad_norm": 0.5240740180015564,
       "learning_rate": 0.0001911228490388136,
-      "loss": 1.8608,
+      "loss": 1.4954,
       "step": 405
     },
     {
       "epoch": 0.4061142196242693,
-      "grad_norm": 0.716299295425415,
+      "grad_norm": 0.5531806945800781,
       "learning_rate": 0.00019107962160773035,
-      "loss": 1.652,
+      "loss": 1.3949,
       "step": 406
     },
     {
       "epoch": 0.40711450095339313,
-      "grad_norm": 0.6675253510475159,
+      "grad_norm": 0.5266262888908386,
       "learning_rate": 0.0001910362940966147,
-      "loss": 1.5963,
+      "loss": 1.2859,
       "step": 407
     },
     {
       "epoch": 0.40811478228251696,
-      "grad_norm": 0.6555336713790894,
+      "grad_norm": 0.5734869241714478,
       "learning_rate": 0.00019099286655307568,
-      "loss": 1.4991,
+      "loss": 1.2451,
       "step": 408
     },
     {
       "epoch": 0.4091150636116408,
-      "grad_norm": 0.7307867407798767,
+      "grad_norm": 0.5922874212265015,
       "learning_rate": 0.0001909493390248324,
-      "loss": 1.8221,
+      "loss": 1.5429,
       "step": 409
     },
     {
       "epoch": 0.41011534494076457,
-      "grad_norm": 0.6557430624961853,
+      "grad_norm": 0.542540431022644,
       "learning_rate": 0.00019090571155971366,
-      "loss": 1.6484,
+      "loss": 1.4138,
       "step": 410
     },
     {
       "epoch": 0.4111156262698884,
-      "grad_norm": 0.6816605925559998,
+      "grad_norm": 0.57356196641922,
       "learning_rate": 0.00019086198420565823,
-      "loss": 1.5052,
+      "loss": 1.2592,
       "step": 411
     },
     {
       "epoch": 0.41211590759901223,
-      "grad_norm": 0.66513592004776,
+      "grad_norm": 0.6042733192443848,
       "learning_rate": 0.00019081815701071445,
-      "loss": 1.8818,
+      "loss": 1.5524,
       "step": 412
     },
     {
       "epoch": 0.41311618892813606,
-      "grad_norm": 0.6807469129562378,
+      "grad_norm": 0.46550241112709045,
       "learning_rate": 0.0001907742300230406,
-      "loss": 1.5997,
+      "loss": 1.308,
       "step": 413
     },
     {
       "epoch": 0.41411647025725984,
-      "grad_norm": 0.8060654401779175,
+      "grad_norm": 0.6283137798309326,
       "learning_rate": 0.00019073020329090444,
-      "loss": 1.8099,
+      "loss": 1.4753,
       "step": 414
     },
     {
       "epoch": 0.41511675158638367,
-      "grad_norm": 0.7440110445022583,
+      "grad_norm": 0.5254876613616943,
       "learning_rate": 0.0001906860768626834,
-      "loss": 1.4876,
+      "loss": 1.2157,
       "step": 415
     },
     {
       "epoch": 0.4161170329155075,
-      "grad_norm": 0.7675415277481079,
+      "grad_norm": 0.59089195728302,
       "learning_rate": 0.00019064185078686443,
-      "loss": 1.4722,
+      "loss": 1.2684,
       "step": 416
     },
     {
       "epoch": 0.4171173142446313,
-      "grad_norm": 0.6656553149223328,
+      "grad_norm": 0.7129126787185669,
       "learning_rate": 0.000190597525112044,
-      "loss": 1.6453,
+      "loss": 1.3974,
       "step": 417
     },
     {
       "epoch": 0.4181175955737551,
-      "grad_norm": 0.730689287185669,
+      "grad_norm": 0.607305109500885,
       "learning_rate": 0.000190553099886928,
-      "loss": 1.6584,
+      "loss": 1.4312,
       "step": 418
     },
     {
       "epoch": 0.41911787690287894,
-      "grad_norm": 0.8425858616828918,
+      "grad_norm": 0.49921515583992004,
       "learning_rate": 0.00019050857516033173,
-      "loss": 1.6249,
+      "loss": 1.3469,
       "step": 419
     },
     {
       "epoch": 0.42011815823200277,
-      "grad_norm": 0.7816892266273499,
+      "grad_norm": 0.6167325377464294,
       "learning_rate": 0.00019046395098117983,
-      "loss": 1.7532,
+      "loss": 1.4723,
       "step": 420
     },
     {
       "epoch": 0.42111843956112655,
-      "grad_norm": 0.7324026823043823,
+      "grad_norm": 0.6144593358039856,
       "learning_rate": 0.00019041922739850616,
-      "loss": 1.8523,
+      "loss": 1.5502,
       "step": 421
     },
     {
       "epoch": 0.4221187208902504,
-      "grad_norm": 0.7473389506340027,
+      "grad_norm": 0.61333167552948,
       "learning_rate": 0.00019037440446145385,
-      "loss": 1.583,
+      "loss": 1.3283,
       "step": 422
     },
     {
       "epoch": 0.4231190022193742,
-      "grad_norm": 0.8720895051956177,
+      "grad_norm": 0.5881702303886414,
       "learning_rate": 0.00019032948221927524,
-      "loss": 1.6806,
+      "loss": 1.4206,
       "step": 423
     },
     {
       "epoch": 0.42411928354849804,
-      "grad_norm": 0.728528618812561,
+      "grad_norm": 0.5334322452545166,
       "learning_rate": 0.00019028446072133175,
-      "loss": 1.7283,
+      "loss": 1.4603,
       "step": 424
     },
     {
       "epoch": 0.4251195648776218,
-      "grad_norm": 0.739930272102356,
+      "grad_norm": 0.5730605721473694,
       "learning_rate": 0.00019023934001709383,
-      "loss": 1.7244,
+      "loss": 1.4375,
       "step": 425
     },
     {
       "epoch": 0.42611984620674564,
-      "grad_norm": 0.7825399041175842,
+      "grad_norm": 0.6227820515632629,
       "learning_rate": 0.00019019412015614098,
-      "loss": 1.7871,
+      "loss": 1.4888,
       "step": 426
     },
     {
       "epoch": 0.4271201275358695,
-      "grad_norm": 0.8878734707832336,
+      "grad_norm": 0.5811313390731812,
       "learning_rate": 0.00019014880118816164,
-      "loss": 1.6018,
+      "loss": 1.3492,
       "step": 427
     },
     {
       "epoch": 0.4281204088649933,
-      "grad_norm": 0.726259708404541,
+      "grad_norm": 0.5685800313949585,
       "learning_rate": 0.0001901033831629532,
-      "loss": 1.7732,
+      "loss": 1.5052,
       "step": 428
     },
     {
       "epoch": 0.4291206901941171,
-      "grad_norm": 0.7620319724082947,
+      "grad_norm": 0.5961394309997559,
       "learning_rate": 0.00019005786613042185,
-      "loss": 1.6466,
+      "loss": 1.3324,
       "step": 429
     },
     {
       "epoch": 0.4301209715232409,
-      "grad_norm": 0.7295501828193665,
+      "grad_norm": 0.5845314860343933,
       "learning_rate": 0.00019001225014058255,
-      "loss": 1.8708,
+      "loss": 1.5733,
       "step": 430
     },
     {
       "epoch": 0.43112125285236474,
-      "grad_norm": 0.7419458031654358,
+      "grad_norm": 0.5400176048278809,
       "learning_rate": 0.00018996653524355902,
-      "loss": 1.6583,
+      "loss": 1.3973,
       "step": 431
     },
     {
       "epoch": 0.4321215341814885,
-      "grad_norm": 0.7701705098152161,
+      "grad_norm": 0.5462201833724976,
       "learning_rate": 0.00018992072148958368,
-      "loss": 1.4421,
+      "loss": 1.2167,
       "step": 432
     },
     {
       "epoch": 0.43312181551061235,
-      "grad_norm": 0.8237659931182861,
+      "grad_norm": 0.6200360059738159,
       "learning_rate": 0.00018987480892899758,
-      "loss": 1.844,
+      "loss": 1.5596,
       "step": 433
     },
     {
       "epoch": 0.4341220968397362,
-      "grad_norm": 0.6167672276496887,
+      "grad_norm": 0.5230718851089478,
       "learning_rate": 0.00018982879761225027,
-      "loss": 1.6193,
+      "loss": 1.3661,
       "step": 434
     },
     {
       "epoch": 0.43512237816886,
-      "grad_norm": 0.7565534710884094,
+      "grad_norm": 0.5868643522262573,
       "learning_rate": 0.00018978268758989991,
-      "loss": 1.7655,
+      "loss": 1.4792,
       "step": 435
     },
     {
       "epoch": 0.4361226594979838,
-      "grad_norm": 0.8333333730697632,
+      "grad_norm": 0.580892026424408,
       "learning_rate": 0.00018973647891261307,
-      "loss": 1.5764,
+      "loss": 1.3275,
       "step": 436
     },
     {
       "epoch": 0.4371229408271076,
-      "grad_norm": 0.7404434084892273,
+      "grad_norm": 0.5903263688087463,
       "learning_rate": 0.00018969017163116472,
-      "loss": 1.7922,
+      "loss": 1.4721,
       "step": 437
     },
     {
       "epoch": 0.43812322215623145,
-      "grad_norm": 0.7129400372505188,
+      "grad_norm": 0.5108968019485474,
       "learning_rate": 0.0001896437657964382,
-      "loss": 1.6925,
+      "loss": 1.3785,
       "step": 438
     },
     {
       "epoch": 0.4391235034853553,
-      "grad_norm": 0.7750307321548462,
+      "grad_norm": 0.6707500219345093,
       "learning_rate": 0.00018959726145942508,
-      "loss": 1.8133,
+      "loss": 1.5033,
       "step": 439
     },
     {
       "epoch": 0.44012378481447906,
-      "grad_norm": 0.7244920134544373,
+      "grad_norm": 0.5793184638023376,
       "learning_rate": 0.00018955065867122528,
-      "loss": 1.6425,
+      "loss": 1.3629,
       "step": 440
     },
     {
       "epoch": 0.4411240661436029,
-      "grad_norm": 0.7624644637107849,
+      "grad_norm": 0.5549041628837585,
       "learning_rate": 0.00018950395748304678,
-      "loss": 1.886,
+      "loss": 1.5557,
       "step": 441
     },
     {
       "epoch": 0.4421243474727267,
-      "grad_norm": 0.7016286849975586,
+      "grad_norm": 0.5406919121742249,
       "learning_rate": 0.0001894571579462058,
-      "loss": 1.7308,
+      "loss": 1.4441,
       "step": 442
     },
     {
       "epoch": 0.4431246288018505,
-      "grad_norm": 0.6965353488922119,
+      "grad_norm": 0.5131089091300964,
       "learning_rate": 0.00018941026011212654,
-      "loss": 1.5872,
+      "loss": 1.3051,
       "step": 443
     },
     {
       "epoch": 0.4441249101309743,
-      "grad_norm": 0.7479360699653625,
+      "grad_norm": 0.601586639881134,
       "learning_rate": 0.00018936326403234125,
-      "loss": 1.8129,
+      "loss": 1.5297,
       "step": 444
     },
     {
       "epoch": 0.44512519146009816,
-      "grad_norm": 0.7027686834335327,
+      "grad_norm": 0.5036457180976868,
       "learning_rate": 0.00018931616975849006,
-      "loss": 1.6433,
+      "loss": 1.357,
       "step": 445
     },
     {
       "epoch": 0.446125472789222,
-      "grad_norm": 0.7771592140197754,
+      "grad_norm": 0.5471266508102417,
       "learning_rate": 0.00018926897734232115,
-      "loss": 1.4645,
+      "loss": 1.2176,
       "step": 446
     },
     {
       "epoch": 0.44712575411834576,
-      "grad_norm": 0.7766458988189697,
+      "grad_norm": 0.6057867407798767,
       "learning_rate": 0.0001892216868356904,
-      "loss": 1.7873,
+      "loss": 1.4763,
       "step": 447
     },
     {
       "epoch": 0.4481260354474696,
-      "grad_norm": 0.8146182298660278,
+      "grad_norm": 0.5384593605995178,
       "learning_rate": 0.0001891742982905615,
-      "loss": 1.7935,
+      "loss": 1.513,
       "step": 448
     },
     {
       "epoch": 0.4491263167765934,
-      "grad_norm": 0.6744781136512756,
+      "grad_norm": 0.6144880056381226,
       "learning_rate": 0.00018912681175900598,
-      "loss": 1.8916,
+      "loss": 1.5782,
       "step": 449
     },
     {
       "epoch": 0.45012659810571726,
-      "grad_norm": 0.6259024143218994,
+      "grad_norm": 0.4838174879550934,
       "learning_rate": 0.00018907922729320285,
-      "loss": 1.6616,
+      "loss": 1.4085,
       "step": 450
     },
     {
       "epoch": 0.45112687943484103,
-      "grad_norm": 0.7717494368553162,
+      "grad_norm": 0.6852928400039673,
       "learning_rate": 0.00018903154494543889,
-      "loss": 1.817,
+      "loss": 1.5989,
       "step": 451
     },
     {
       "epoch": 0.45212716076396486,
-      "grad_norm": 0.648040771484375,
+      "grad_norm": 0.47527411580085754,
       "learning_rate": 0.00018898376476810834,
-      "loss": 1.6309,
+      "loss": 1.3409,
       "step": 452
     },
     {
       "epoch": 0.4531274420930887,
-      "grad_norm": 0.7560017704963684,
+      "grad_norm": 0.5665884613990784,
       "learning_rate": 0.00018893588681371303,
-      "loss": 1.8016,
+      "loss": 1.5395,
       "step": 453
     },
     {
       "epoch": 0.45412772342221247,
-      "grad_norm": 0.8778628706932068,
+      "grad_norm": 0.5792158246040344,
       "learning_rate": 0.00018888791113486213,
-      "loss": 1.7797,
+      "loss": 1.516,
       "step": 454
     },
     {
       "epoch": 0.4551280047513363,
-      "grad_norm": 0.6795655488967896,
+      "grad_norm": 0.5223523378372192,
       "learning_rate": 0.00018883983778427227,
-      "loss": 1.6343,
+      "loss": 1.3678,
       "step": 455
     },
     {
       "epoch": 0.45612828608046013,
-      "grad_norm": 0.6690213084220886,
+      "grad_norm": 0.5927590131759644,
       "learning_rate": 0.0001887916668147673,
-      "loss": 1.6224,
+      "loss": 1.3617,
       "step": 456
     },
     {
       "epoch": 0.45712856740958396,
-      "grad_norm": 0.7529327869415283,
+      "grad_norm": 0.7266496419906616,
       "learning_rate": 0.00018874339827927846,
-      "loss": 1.6396,
+      "loss": 1.3734,
       "step": 457
     },
     {
       "epoch": 0.45812884873870774,
-      "grad_norm": 0.8393098711967468,
+      "grad_norm": 0.6495805978775024,
       "learning_rate": 0.00018869503223084414,
-      "loss": 1.8374,
+      "loss": 1.5282,
       "step": 458
     },
     {
       "epoch": 0.45912913006783157,
-      "grad_norm": 0.7435776591300964,
+      "grad_norm": 0.6099816560745239,
       "learning_rate": 0.00018864656872260985,
-      "loss": 1.7363,
+      "loss": 1.4691,
       "step": 459
     },
     {
       "epoch": 0.4601294113969554,
-      "grad_norm": 0.6737055778503418,
+      "grad_norm": 0.5208227038383484,
       "learning_rate": 0.00018859800780782828,
-      "loss": 1.6661,
+      "loss": 1.3949,
       "step": 460
     },
     {
       "epoch": 0.46112969272607923,
-      "grad_norm": 0.7674340605735779,
+      "grad_norm": 0.5526600480079651,
       "learning_rate": 0.000188549349539859,
-      "loss": 1.6269,
+      "loss": 1.3557,
       "step": 461
     },
     {
       "epoch": 0.462129974055203,
-      "grad_norm": 0.7329950928688049,
+      "grad_norm": 0.5537740588188171,
       "learning_rate": 0.00018850059397216876,
-      "loss": 1.6989,
+      "loss": 1.4703,
       "step": 462
     },
     {
       "epoch": 0.46313025538432684,
-      "grad_norm": 0.7075778841972351,
+      "grad_norm": 0.5553976893424988,
       "learning_rate": 0.00018845174115833099,
-      "loss": 1.7286,
+      "loss": 1.4356,
       "step": 463
     },
     {
       "epoch": 0.46413053671345067,
-      "grad_norm": 0.7973611354827881,
+      "grad_norm": 0.6027779579162598,
       "learning_rate": 0.0001884027911520262,
-      "loss": 1.7478,
+      "loss": 1.4763,
       "step": 464
     },
     {
       "epoch": 0.4651308180425745,
-      "grad_norm": 0.7790057063102722,
+      "grad_norm": 0.5559154748916626,
       "learning_rate": 0.00018835374400704154,
-      "loss": 1.6659,
+      "loss": 1.4148,
       "step": 465
     },
     {
       "epoch": 0.4661310993716983,
-      "grad_norm": 0.8505310416221619,
+      "grad_norm": 0.6124109029769897,
       "learning_rate": 0.00018830459977727096,
-      "loss": 1.6879,
+      "loss": 1.4468,
       "step": 466
     },
     {
       "epoch": 0.4671313807008221,
-      "grad_norm": 0.7616267800331116,
+      "grad_norm": 0.4762580692768097,
       "learning_rate": 0.0001882553585167151,
-      "loss": 1.6525,
+      "loss": 1.3714,
       "step": 467
     },
     {
       "epoch": 0.46813166202994594,
-      "grad_norm": 0.8038133978843689,
+      "grad_norm": 0.5793487429618835,
       "learning_rate": 0.00018820602027948114,
-      "loss": 1.7929,
+      "loss": 1.4828,
       "step": 468
     },
     {
       "epoch": 0.4691319433590697,
-      "grad_norm": 0.6762365698814392,
+      "grad_norm": 0.55177241563797,
       "learning_rate": 0.00018815658511978298,
-      "loss": 1.6926,
+      "loss": 1.4157,
       "step": 469
     },
     {
       "epoch": 0.47013222468819355,
-      "grad_norm": 0.6515015959739685,
+      "grad_norm": 0.5065292716026306,
       "learning_rate": 0.00018810705309194083,
-      "loss": 1.7817,
+      "loss": 1.4519,
       "step": 470
     },
     {
       "epoch": 0.4711325060173174,
-      "grad_norm": 0.696675181388855,
+      "grad_norm": 0.5401413440704346,
       "learning_rate": 0.00018805742425038145,
-      "loss": 1.7195,
+      "loss": 1.4344,
       "step": 471
     },
     {
       "epoch": 0.4721327873464412,
-      "grad_norm": 0.7929533123970032,
+      "grad_norm": 0.7173880338668823,
       "learning_rate": 0.00018800769864963802,
-      "loss": 2.0165,
+      "loss": 1.7325,
       "step": 472
     },
     {
       "epoch": 0.473133068675565,
-      "grad_norm": 0.7223886251449585,
+      "grad_norm": 0.507682204246521,
       "learning_rate": 0.00018795787634434994,
-      "loss": 1.6708,
+      "loss": 1.37,
       "step": 473
     },
     {
       "epoch": 0.4741333500046888,
-      "grad_norm": 0.7982028126716614,
+      "grad_norm": 0.551888644695282,
       "learning_rate": 0.0001879079573892629,
-      "loss": 1.628,
+      "loss": 1.3695,
       "step": 474
     },
     {
       "epoch": 0.47513363133381264,
-      "grad_norm": 0.6962152123451233,
+      "grad_norm": 0.5109260082244873,
       "learning_rate": 0.00018785794183922883,
-      "loss": 1.6429,
+      "loss": 1.4001,
       "step": 475
     },
     {
       "epoch": 0.4761339126629365,
-      "grad_norm": 0.687489926815033,
+      "grad_norm": 0.4565551280975342,
       "learning_rate": 0.00018780782974920572,
-      "loss": 1.4546,
+      "loss": 1.1752,
       "step": 476
     },
     {
       "epoch": 0.47713419399206025,
-      "grad_norm": 0.7260375022888184,
+      "grad_norm": 0.5651509761810303,
       "learning_rate": 0.00018775762117425777,
-      "loss": 1.739,
+      "loss": 1.4291,
       "step": 477
     },
     {
       "epoch": 0.4781344753211841,
-      "grad_norm": 0.759400486946106,
+      "grad_norm": 0.5827792286872864,
       "learning_rate": 0.0001877073161695551,
-      "loss": 1.6465,
+      "loss": 1.3438,
       "step": 478
     },
     {
       "epoch": 0.4791347566503079,
-      "grad_norm": 0.7412364482879639,
+      "grad_norm": 0.5719752907752991,
       "learning_rate": 0.00018765691479037376,
-      "loss": 1.7333,
+      "loss": 1.4683,
       "step": 479
     },
     {
       "epoch": 0.4801350379794317,
-      "grad_norm": 0.6909674406051636,
+      "grad_norm": 0.5153111815452576,
       "learning_rate": 0.00018760641709209583,
-      "loss": 1.6936,
+      "loss": 1.4392,
       "step": 480
     },
     {
       "epoch": 0.4811353193085555,
-      "grad_norm": 0.6478050947189331,
+      "grad_norm": 0.5455904603004456,
       "learning_rate": 0.0001875558231302091,
-      "loss": 1.4435,
+      "loss": 1.1603,
       "step": 481
     },
     {
       "epoch": 0.48213560063767935,
-      "grad_norm": 0.6662638187408447,
+      "grad_norm": 0.5857074856758118,
       "learning_rate": 0.00018750513296030718,
-      "loss": 1.5567,
+      "loss": 1.3099,
       "step": 482
     },
     {
       "epoch": 0.4831358819668032,
-      "grad_norm": 0.6973413825035095,
+      "grad_norm": 0.6051676273345947,
       "learning_rate": 0.00018745434663808942,
-      "loss": 1.6434,
+      "loss": 1.3587,
       "step": 483
     },
     {
       "epoch": 0.48413616329592696,
-      "grad_norm": 0.8381956815719604,
+      "grad_norm": 0.588749885559082,
       "learning_rate": 0.0001874034642193608,
-      "loss": 1.8568,
+      "loss": 1.5277,
       "step": 484
     },
     {
       "epoch": 0.4851364446250508,
-      "grad_norm": 0.6522262096405029,
+      "grad_norm": 0.5295410752296448,
       "learning_rate": 0.0001873524857600319,
-      "loss": 1.4265,
+      "loss": 1.2084,
       "step": 485
     },
     {
       "epoch": 0.4861367259541746,
-      "grad_norm": 0.7018440961837769,
+      "grad_norm": 0.5313368439674377,
       "learning_rate": 0.00018730141131611882,
-      "loss": 1.6914,
+      "loss": 1.4002,
       "step": 486
     },
     {
       "epoch": 0.48713700728329845,
-      "grad_norm": 0.8237236142158508,
+      "grad_norm": 0.5166353583335876,
       "learning_rate": 0.00018725024094374315,
-      "loss": 1.4462,
+      "loss": 1.208,
       "step": 487
     },
     {
       "epoch": 0.4881372886124222,
-      "grad_norm": 0.6507928967475891,
+      "grad_norm": 0.5478363037109375,
       "learning_rate": 0.00018719897469913184,
-      "loss": 1.5802,
+      "loss": 1.3236,
       "step": 488
     },
     {
       "epoch": 0.48913756994154606,
-      "grad_norm": 0.8120100498199463,
+      "grad_norm": 0.5531913042068481,
       "learning_rate": 0.00018714761263861728,
-      "loss": 1.7819,
+      "loss": 1.4938,
       "step": 489
     },
     {
       "epoch": 0.4901378512706699,
-      "grad_norm": 0.759722888469696,
+      "grad_norm": 0.5334530472755432,
       "learning_rate": 0.000187096154818637,
-      "loss": 1.7481,
+      "loss": 1.4172,
       "step": 490
     },
     {
       "epoch": 0.4911381325997937,
-      "grad_norm": 0.7146822214126587,
+      "grad_norm": 0.5667001605033875,
       "learning_rate": 0.00018704460129573391,
-      "loss": 1.6217,
+      "loss": 1.3517,
       "step": 491
     },
     {
       "epoch": 0.4921384139289175,
-      "grad_norm": 0.7138429284095764,
+      "grad_norm": 0.5568780303001404,
       "learning_rate": 0.00018699295212655596,
-      "loss": 1.7242,
+      "loss": 1.4287,
       "step": 492
     },
     {
       "epoch": 0.4931386952580413,
-      "grad_norm": 0.8145613670349121,
+      "grad_norm": 0.6663610935211182,
       "learning_rate": 0.00018694120736785632,
-      "loss": 1.8759,
+      "loss": 1.5416,
       "step": 493
     },
     {
       "epoch": 0.49413897658716516,
-      "grad_norm": 0.6624435186386108,
+      "grad_norm": 0.5753045082092285,
       "learning_rate": 0.00018688936707649304,
-      "loss": 1.8632,
+      "loss": 1.5552,
       "step": 494
     },
     {
       "epoch": 0.49513925791628893,
-      "grad_norm": 0.6550843119621277,
+      "grad_norm": 0.5707410573959351,
       "learning_rate": 0.00018683743130942928,
-      "loss": 1.7645,
+      "loss": 1.5332,
       "step": 495
     },
     {
       "epoch": 0.49613953924541276,
-      "grad_norm": 0.6931488513946533,
+      "grad_norm": 0.5847951173782349,
       "learning_rate": 0.00018678540012373302,
-      "loss": 1.5899,
+      "loss": 1.3488,
       "step": 496
     },
     {
       "epoch": 0.4971398205745366,
-      "grad_norm": 0.684229850769043,
+      "grad_norm": 0.60503751039505,
       "learning_rate": 0.00018673327357657715,
-      "loss": 1.6902,
+      "loss": 1.3924,
       "step": 497
     },
     {
       "epoch": 0.4981401019036604,
-      "grad_norm": 0.7380666136741638,
+      "grad_norm": 0.635142982006073,
       "learning_rate": 0.0001866810517252393,
-      "loss": 1.7115,
+      "loss": 1.4392,
       "step": 498
     },
     {
       "epoch": 0.4991403832327842,
-      "grad_norm": 0.6782827973365784,
+      "grad_norm": 0.5536782741546631,
       "learning_rate": 0.00018662873462710184,
-      "loss": 1.495,
+      "loss": 1.286,
       "step": 499
     },
     {
       "epoch": 0.5001406645619081,
-      "grad_norm": 0.730248749256134,
+      "grad_norm": 0.5676659345626831,
       "learning_rate": 0.0001865763223396518,
-      "loss": 1.5147,
+      "loss": 1.3006,
       "step": 500
     },
     {
       "epoch": 0.5011409458910319,
-      "grad_norm": 0.7644149661064148,
+      "grad_norm": 0.5546663403511047,
       "learning_rate": 0.00018652381492048083,
-      "loss": 1.7278,
+      "loss": 1.418,
       "step": 501
     },
     {
       "epoch": 0.5021412272201556,
-      "grad_norm": 0.6977668404579163,
+      "grad_norm": 0.5137162804603577,
       "learning_rate": 0.00018647121242728506,
-      "loss": 1.5194,
+      "loss": 1.3173,
       "step": 502
     },
     {
       "epoch": 0.5031415085492795,
-      "grad_norm": 0.7714502215385437,
+      "grad_norm": 0.5474348068237305,
       "learning_rate": 0.00018641851491786512,
-      "loss": 2.0039,
+      "loss": 1.6652,
       "step": 503
     },
     {
       "epoch": 0.5041417898784033,
-      "grad_norm": 0.9013757109642029,
+      "grad_norm": 0.5563383102416992,
       "learning_rate": 0.00018636572245012606,
-      "loss": 1.8069,
+      "loss": 1.4519,
       "step": 504
     },
     {
       "epoch": 0.5051420712075271,
-      "grad_norm": 0.7173192501068115,
+      "grad_norm": 0.5621083974838257,
       "learning_rate": 0.00018631283508207725,
-      "loss": 1.841,
+      "loss": 1.5418,
       "step": 505
     },
     {
       "epoch": 0.506142352536651,
-      "grad_norm": 0.7639481425285339,
+      "grad_norm": 0.49915972352027893,
       "learning_rate": 0.00018625985287183233,
-      "loss": 1.5333,
+      "loss": 1.2969,
       "step": 506
     },
     {
       "epoch": 0.5071426338657747,
-      "grad_norm": 0.8779808282852173,
+      "grad_norm": 0.601996660232544,
       "learning_rate": 0.00018620677587760916,
-      "loss": 1.8304,
+      "loss": 1.4483,
       "step": 507
     },
     {
       "epoch": 0.5081429151948985,
-      "grad_norm": 0.8956230282783508,
+      "grad_norm": 0.5594652891159058,
       "learning_rate": 0.00018615360415772978,
-      "loss": 1.7228,
+      "loss": 1.4094,
       "step": 508
     },
     {
       "epoch": 0.5091431965240224,
-      "grad_norm": 0.8217945098876953,
+      "grad_norm": 0.557381808757782,
       "learning_rate": 0.00018610033777062025,
-      "loss": 1.4977,
+      "loss": 1.216,
       "step": 509
     },
     {
       "epoch": 0.5101434778531462,
-      "grad_norm": 0.7628902196884155,
+      "grad_norm": 0.5841740369796753,
       "learning_rate": 0.0001860469767748108,
-      "loss": 1.7634,
+      "loss": 1.4924,
       "step": 510
     },
     {
       "epoch": 0.5111437591822701,
-      "grad_norm": 0.635911226272583,
+      "grad_norm": 0.4968324899673462,
       "learning_rate": 0.00018599352122893539,
-      "loss": 1.5103,
+      "loss": 1.2474,
       "step": 511
     },
     {
       "epoch": 0.5121440405113938,
-      "grad_norm": 0.6439646482467651,
+      "grad_norm": 0.5390318632125854,
       "learning_rate": 0.00018593997119173205,
-      "loss": 1.7281,
+      "loss": 1.4484,
       "step": 512
     },
     {
       "epoch": 0.5131443218405176,
-      "grad_norm": 0.7753567099571228,
+      "grad_norm": 0.6626128554344177,
       "learning_rate": 0.00018588632672204264,
-      "loss": 1.9028,
+      "loss": 1.5664,
       "step": 513
     },
     {
       "epoch": 0.5141446031696415,
-      "grad_norm": 0.8296717405319214,
+      "grad_norm": 0.6183133721351624,
       "learning_rate": 0.0001858325878788126,
-      "loss": 1.9049,
+      "loss": 1.5603,
       "step": 514
     },
     {
       "epoch": 0.5151448844987653,
-      "grad_norm": 0.7379167079925537,
+      "grad_norm": 0.5574773550033569,
       "learning_rate": 0.00018577875472109134,
-      "loss": 1.6262,
+      "loss": 1.3668,
       "step": 515
     },
     {
       "epoch": 0.516145165827889,
-      "grad_norm": 0.634040355682373,
+      "grad_norm": 0.5127518773078918,
       "learning_rate": 0.0001857248273080317,
-      "loss": 1.5416,
+      "loss": 1.264,
       "step": 516
     },
     {
       "epoch": 0.5171454471570129,
-      "grad_norm": 0.7394217252731323,
+      "grad_norm": 0.6540619134902954,
       "learning_rate": 0.00018567080569889015,
-      "loss": 1.6035,
+      "loss": 1.3091,
       "step": 517
     },
     {
       "epoch": 0.5181457284861367,
-      "grad_norm": 0.705426037311554,
+      "grad_norm": 0.5286336541175842,
       "learning_rate": 0.00018561668995302667,
-      "loss": 1.616,
+      "loss": 1.3581,
       "step": 518
     },
     {
       "epoch": 0.5191460098152605,
-      "grad_norm": 0.778021514415741,
+      "grad_norm": 0.6609972715377808,
       "learning_rate": 0.00018556248012990468,
-      "loss": 1.5206,
+      "loss": 1.3123,
       "step": 519
     },
     {
       "epoch": 0.5201462911443844,
-      "grad_norm": 0.7203211188316345,
+      "grad_norm": 0.48230236768722534,
       "learning_rate": 0.000185508176289091,
-      "loss": 1.5369,
+      "loss": 1.2372,
       "step": 520
     },
     {
       "epoch": 0.5211465724735082,
-      "grad_norm": 0.7390999794006348,
+      "grad_norm": 0.5173765420913696,
       "learning_rate": 0.00018545377849025566,
-      "loss": 1.6438,
+      "loss": 1.327,
       "step": 521
     },
     {
       "epoch": 0.522146853802632,
-      "grad_norm": 0.6779179573059082,
+      "grad_norm": 0.5822583436965942,
       "learning_rate": 0.0001853992867931721,
-      "loss": 1.6268,
+      "loss": 1.3851,
       "step": 522
     },
     {
       "epoch": 0.5231471351317558,
-      "grad_norm": 0.6589105725288391,
+      "grad_norm": 0.6025621891021729,
       "learning_rate": 0.00018534470125771674,
-      "loss": 1.8353,
+      "loss": 1.5627,
       "step": 523
     },
     {
       "epoch": 0.5241474164608796,
-      "grad_norm": 0.692081868648529,
+      "grad_norm": 0.5516778230667114,
       "learning_rate": 0.0001852900219438693,
-      "loss": 1.7047,
+      "loss": 1.4036,
       "step": 524
     },
     {
       "epoch": 0.5251476977900035,
-      "grad_norm": 0.6639224886894226,
+      "grad_norm": 0.5738380551338196,
       "learning_rate": 0.0001852352489117124,
-      "loss": 1.7448,
+      "loss": 1.5042,
       "step": 525
     },
     {
       "epoch": 0.5261479791191273,
-      "grad_norm": 0.7168188095092773,
+      "grad_norm": 0.6360776424407959,
       "learning_rate": 0.00018518038222143174,
-      "loss": 1.6734,
+      "loss": 1.4101,
       "step": 526
     },
     {
       "epoch": 0.527148260448251,
-      "grad_norm": 0.7671873569488525,
+      "grad_norm": 0.5776675939559937,
       "learning_rate": 0.00018512542193331583,
-      "loss": 1.9392,
+      "loss": 1.6015,
       "step": 527
     },
     {
       "epoch": 0.5281485417773749,
-      "grad_norm": 0.7861583828926086,
+      "grad_norm": 0.5662726759910583,
       "learning_rate": 0.00018507036810775615,
-      "loss": 1.5749,
+      "loss": 1.3186,
       "step": 528
     },
     {
       "epoch": 0.5291488231064987,
-      "grad_norm": 0.6727952361106873,
+      "grad_norm": 0.6518335938453674,
       "learning_rate": 0.00018501522080524688,
-      "loss": 1.7584,
+      "loss": 1.4882,
       "step": 529
     },
     {
       "epoch": 0.5301491044356225,
-      "grad_norm": 0.7287748456001282,
+      "grad_norm": 0.5475590825080872,
       "learning_rate": 0.0001849599800863849,
-      "loss": 1.783,
+      "loss": 1.487,
       "step": 530
     },
     {
       "epoch": 0.5311493857647464,
-      "grad_norm": 0.6883361339569092,
+      "grad_norm": 0.6275209188461304,
       "learning_rate": 0.0001849046460118698,
-      "loss": 1.6104,
+      "loss": 1.3563,
       "step": 531
     },
     {
       "epoch": 0.5321496670938701,
-      "grad_norm": 0.6767789125442505,
+      "grad_norm": 0.5629132390022278,
       "learning_rate": 0.0001848492186425037,
-      "loss": 1.8416,
+      "loss": 1.516,
       "step": 532
     },
     {
       "epoch": 0.533149948422994,
-      "grad_norm": 0.7468088269233704,
+      "grad_norm": 0.5251057744026184,
       "learning_rate": 0.0001847936980391913,
-      "loss": 1.8297,
+      "loss": 1.5254,
       "step": 533
     },
     {
       "epoch": 0.5341502297521178,
-      "grad_norm": 0.7560007572174072,
+      "grad_norm": 0.5635396838188171,
       "learning_rate": 0.00018473808426293964,
-      "loss": 1.5405,
+      "loss": 1.3408,
       "step": 534
     },
     {
       "epoch": 0.5351505110812416,
-      "grad_norm": 0.6900463104248047,
+      "grad_norm": 0.527082622051239,
       "learning_rate": 0.00018468237737485823,
-      "loss": 1.5117,
+      "loss": 1.2664,
       "step": 535
     },
     {
       "epoch": 0.5361507924103655,
-      "grad_norm": 0.8691229820251465,
+      "grad_norm": 0.6555044054985046,
       "learning_rate": 0.00018462657743615888,
-      "loss": 1.724,
+      "loss": 1.464,
       "step": 536
     },
     {
       "epoch": 0.5371510737394892,
-      "grad_norm": 0.7081372141838074,
+      "grad_norm": 0.5468676686286926,
       "learning_rate": 0.00018457068450815562,
-      "loss": 1.6418,
+      "loss": 1.3733,
       "step": 537
     },
     {
       "epoch": 0.538151355068613,
-      "grad_norm": 0.8148525953292847,
+      "grad_norm": 0.5662835836410522,
       "learning_rate": 0.00018451469865226464,
-      "loss": 1.8564,
+      "loss": 1.509,
       "step": 538
     },
     {
       "epoch": 0.5391516363977369,
-      "grad_norm": 0.7306076288223267,
+      "grad_norm": 0.5553548336029053,
       "learning_rate": 0.00018445861993000436,
-      "loss": 1.4974,
+      "loss": 1.2476,
       "step": 539
     },
     {
       "epoch": 0.5401519177268607,
-      "grad_norm": 0.815825343132019,
+      "grad_norm": 0.6240925192832947,
       "learning_rate": 0.00018440244840299506,
-      "loss": 1.8965,
+      "loss": 1.5835,
       "step": 540
     },
     {
       "epoch": 0.5411521990559846,
-      "grad_norm": 0.753034234046936,
+      "grad_norm": 0.6107541918754578,
       "learning_rate": 0.0001843461841329591,
-      "loss": 2.016,
+      "loss": 1.7176,
       "step": 541
     },
     {
       "epoch": 0.5421524803851083,
-      "grad_norm": 0.8658022284507751,
+      "grad_norm": 0.6990326642990112,
       "learning_rate": 0.0001842898271817208,
-      "loss": 1.6697,
+      "loss": 1.4235,
       "step": 542
     },
     {
       "epoch": 0.5431527617142321,
-      "grad_norm": 0.7143135666847229,
+      "grad_norm": 0.583871603012085,
       "learning_rate": 0.00018423337761120618,
-      "loss": 1.8741,
+      "loss": 1.5283,
       "step": 543
     },
     {
       "epoch": 0.544153043043356,
-      "grad_norm": 0.6843370795249939,
+      "grad_norm": 0.5585455894470215,
       "learning_rate": 0.00018417683548344318,
-      "loss": 1.763,
+      "loss": 1.4875,
       "step": 544
     },
     {
       "epoch": 0.5451533243724798,
-      "grad_norm": 0.6699584126472473,
+      "grad_norm": 0.5199955701828003,
       "learning_rate": 0.00018412020086056133,
-      "loss": 1.7126,
+      "loss": 1.3989,
       "step": 545
     },
     {
       "epoch": 0.5461536057016035,
-      "grad_norm": 0.6921600699424744,
+      "grad_norm": 0.5517343878746033,
       "learning_rate": 0.0001840634738047918,
-      "loss": 1.6697,
+      "loss": 1.4073,
       "step": 546
     },
     {
       "epoch": 0.5471538870307274,
-      "grad_norm": 0.822501003742218,
+      "grad_norm": 0.7140716314315796,
       "learning_rate": 0.0001840066543784675,
-      "loss": 1.7045,
+      "loss": 1.4477,
       "step": 547
     },
     {
       "epoch": 0.5481541683598512,
-      "grad_norm": 0.7563886046409607,
+      "grad_norm": 0.548422634601593,
       "learning_rate": 0.00018394974264402257,
-      "loss": 1.6853,
+      "loss": 1.4198,
       "step": 548
     },
     {
       "epoch": 0.549154449688975,
-      "grad_norm": 0.7408218383789062,
+      "grad_norm": 0.5907624363899231,
       "learning_rate": 0.00018389273866399275,
-      "loss": 1.6496,
+      "loss": 1.4033,
       "step": 549
     },
     {
       "epoch": 0.5501547310180989,
-      "grad_norm": 0.6454717516899109,
+      "grad_norm": 0.5327603220939636,
       "learning_rate": 0.00018383564250101512,
-      "loss": 1.5063,
+      "loss": 1.2674,
       "step": 550
     },
     {
       "epoch": 0.5511550123472226,
-      "grad_norm": 0.7033074498176575,
+      "grad_norm": 0.4678132236003876,
       "learning_rate": 0.000183778454217828,
-      "loss": 1.6432,
+      "loss": 1.3644,
       "step": 551
     },
     {
       "epoch": 0.5521552936763465,
-      "grad_norm": 0.768194854259491,
+      "grad_norm": 0.674040675163269,
       "learning_rate": 0.0001837211738772711,
-      "loss": 2.0594,
+      "loss": 1.6942,
       "step": 552
     },
     {
       "epoch": 0.5531555750054703,
-      "grad_norm": 0.7805166244506836,
+      "grad_norm": 0.5374539494514465,
       "learning_rate": 0.000183663801542285,
-      "loss": 1.4317,
+      "loss": 1.1887,
       "step": 553
     },
     {
       "epoch": 0.5541558563345941,
-      "grad_norm": 0.6603556871414185,
+      "grad_norm": 0.5528072118759155,
       "learning_rate": 0.00018360633727591155,
-      "loss": 1.4171,
+      "loss": 1.2,
       "step": 554
     },
     {
       "epoch": 0.555156137663718,
-      "grad_norm": 0.6996607780456543,
+      "grad_norm": 0.6597411632537842,
       "learning_rate": 0.00018354878114129367,
-      "loss": 1.6832,
+      "loss": 1.402,
       "step": 555
     },
     {
       "epoch": 0.5561564189928417,
-      "grad_norm": 0.7861623167991638,
+      "grad_norm": 0.5931501388549805,
       "learning_rate": 0.00018349113320167504,
-      "loss": 1.8425,
+      "loss": 1.5583,
       "step": 556
     },
     {
       "epoch": 0.5571567003219655,
-      "grad_norm": 0.8387210369110107,
+      "grad_norm": 0.6331121921539307,
       "learning_rate": 0.00018343339352040042,
-      "loss": 2.1272,
+      "loss": 1.7882,
       "step": 557
     },
     {
       "epoch": 0.5581569816510894,
-      "grad_norm": 0.829555094242096,
+      "grad_norm": 0.5221824645996094,
       "learning_rate": 0.00018337556216091517,
-      "loss": 1.4835,
+      "loss": 1.2457,
       "step": 558
     },
     {
       "epoch": 0.5591572629802132,
-      "grad_norm": 0.7824863791465759,
+      "grad_norm": 0.6008853912353516,
       "learning_rate": 0.00018331763918676556,
-      "loss": 1.8893,
+      "loss": 1.5916,
       "step": 559
     },
     {
       "epoch": 0.560157544309337,
-      "grad_norm": 0.707683801651001,
+      "grad_norm": 0.5409006476402283,
       "learning_rate": 0.00018325962466159848,
-      "loss": 1.6492,
+      "loss": 1.3457,
       "step": 560
     },
     {
       "epoch": 0.5611578256384608,
-      "grad_norm": 0.775600254535675,
+      "grad_norm": 0.5095859169960022,
       "learning_rate": 0.00018320151864916135,
-      "loss": 1.6542,
+      "loss": 1.3622,
       "step": 561
     },
     {
       "epoch": 0.5621581069675846,
-      "grad_norm": 0.7602002024650574,
+      "grad_norm": 0.5716331005096436,
       "learning_rate": 0.00018314332121330225,
-      "loss": 1.9625,
+      "loss": 1.6168,
       "step": 562
     },
     {
       "epoch": 0.5631583882967085,
-      "grad_norm": 0.7535431385040283,
+      "grad_norm": 0.600307047367096,
       "learning_rate": 0.0001830850324179695,
-      "loss": 1.6407,
+      "loss": 1.4117,
       "step": 563
     },
     {
       "epoch": 0.5641586696258323,
-      "grad_norm": 1.1884644031524658,
+      "grad_norm": 0.7528484463691711,
       "learning_rate": 0.00018302665232721208,
-      "loss": 1.6188,
+      "loss": 1.3418,
       "step": 564
     },
     {
       "epoch": 0.565158950954956,
-      "grad_norm": 0.7576595544815063,
+      "grad_norm": 0.6119087338447571,
       "learning_rate": 0.0001829681810051791,
-      "loss": 1.7739,
+      "loss": 1.4908,
       "step": 565
     },
     {
       "epoch": 0.5661592322840799,
-      "grad_norm": 0.6807442307472229,
+      "grad_norm": 0.6440190672874451,
       "learning_rate": 0.00018290961851611995,
-      "loss": 1.6244,
+      "loss": 1.3511,
       "step": 566
     },
     {
       "epoch": 0.5671595136132037,
-      "grad_norm": 0.7222456932067871,
+      "grad_norm": 0.647294282913208,
       "learning_rate": 0.00018285096492438424,
-      "loss": 1.7809,
+      "loss": 1.5165,
       "step": 567
     },
     {
       "epoch": 0.5681597949423275,
-      "grad_norm": 0.7002213597297668,
+      "grad_norm": 0.5499668717384338,
       "learning_rate": 0.00018279222029442163,
-      "loss": 1.5462,
+      "loss": 1.2876,
       "step": 568
     },
     {
       "epoch": 0.5691600762714514,
-      "grad_norm": 0.8433569669723511,
+      "grad_norm": 0.5629482865333557,
       "learning_rate": 0.00018273338469078186,
-      "loss": 1.5042,
+      "loss": 1.2256,
       "step": 569
     },
     {
       "epoch": 0.5701603576005752,
-      "grad_norm": 0.663144588470459,
+      "grad_norm": 0.48661288619041443,
       "learning_rate": 0.00018267445817811466,
-      "loss": 1.7133,
+      "loss": 1.44,
       "step": 570
     },
     {
       "epoch": 0.5711606389296989,
-      "grad_norm": 0.7298465967178345,
+      "grad_norm": 0.5713567733764648,
       "learning_rate": 0.00018261544082116954,
-      "loss": 2.0201,
+      "loss": 1.741,
       "step": 571
     },
     {
       "epoch": 0.5721609202588228,
-      "grad_norm": 0.7613754868507385,
+      "grad_norm": 0.6130850315093994,
       "learning_rate": 0.00018255633268479595,
-      "loss": 1.8065,
+      "loss": 1.526,
       "step": 572
     },
     {
       "epoch": 0.5731612015879466,
-      "grad_norm": 0.7252177596092224,
+      "grad_norm": 0.5415536761283875,
       "learning_rate": 0.00018249713383394303,
-      "loss": 1.5715,
+      "loss": 1.2405,
       "step": 573
     },
     {
       "epoch": 0.5741614829170705,
-      "grad_norm": 0.783961832523346,
+      "grad_norm": 0.600574791431427,
       "learning_rate": 0.0001824378443336596,
-      "loss": 1.7102,
+      "loss": 1.4534,
       "step": 574
     },
     {
       "epoch": 0.5751617642461943,
-      "grad_norm": 0.8532115817070007,
+      "grad_norm": 0.5479387044906616,
       "learning_rate": 0.00018237846424909413,
-      "loss": 1.7011,
+      "loss": 1.4277,
       "step": 575
     },
     {
       "epoch": 0.576162045575318,
-      "grad_norm": 0.7841559052467346,
+      "grad_norm": 0.5536132454872131,
       "learning_rate": 0.00018231899364549455,
-      "loss": 1.6397,
+      "loss": 1.3918,
       "step": 576
     },
     {
       "epoch": 0.5771623269044419,
-      "grad_norm": 0.7118051648139954,
+      "grad_norm": 0.6228598356246948,
       "learning_rate": 0.00018225943258820833,
-      "loss": 1.7166,
+      "loss": 1.413,
       "step": 577
     },
     {
       "epoch": 0.5781626082335657,
-      "grad_norm": 0.7298933863639832,
+      "grad_norm": 0.5498123168945312,
       "learning_rate": 0.00018219978114268227,
-      "loss": 1.604,
+      "loss": 1.3558,
       "step": 578
     },
     {
       "epoch": 0.5791628895626895,
-      "grad_norm": 0.6644678711891174,
+      "grad_norm": 0.5427498817443848,
       "learning_rate": 0.00018214003937446253,
-      "loss": 1.7673,
+      "loss": 1.509,
       "step": 579
     },
     {
       "epoch": 0.5801631708918134,
-      "grad_norm": 0.6707085371017456,
+      "grad_norm": 0.522285521030426,
       "learning_rate": 0.00018208020734919455,
-      "loss": 1.662,
+      "loss": 1.3847,
       "step": 580
     },
     {
       "epoch": 0.5811634522209371,
-      "grad_norm": 0.7431498765945435,
+      "grad_norm": 0.5963860750198364,
       "learning_rate": 0.00018202028513262288,
-      "loss": 1.7757,
+      "loss": 1.4605,
       "step": 581
     },
     {
       "epoch": 0.5821637335500609,
-      "grad_norm": 0.6936060190200806,
+      "grad_norm": 0.4854499101638794,
       "learning_rate": 0.00018196027279059117,
-      "loss": 1.8464,
+      "loss": 1.4968,
       "step": 582
     },
     {
       "epoch": 0.5831640148791848,
-      "grad_norm": 0.6512508988380432,
+      "grad_norm": 0.503466010093689,
       "learning_rate": 0.00018190017038904215,
-      "loss": 1.5537,
+      "loss": 1.2568,
       "step": 583
     },
     {
       "epoch": 0.5841642962083086,
-      "grad_norm": 0.7541425228118896,
+      "grad_norm": 0.6027483940124512,
       "learning_rate": 0.0001818399779940175,
-      "loss": 1.8552,
+      "loss": 1.5744,
       "step": 584
     },
     {
       "epoch": 0.5851645775374325,
-      "grad_norm": 0.720447301864624,
+      "grad_norm": 0.5450258851051331,
       "learning_rate": 0.0001817796956716578,
-      "loss": 1.5392,
+      "loss": 1.2672,
       "step": 585
     },
     {
       "epoch": 0.5861648588665562,
-      "grad_norm": 0.7526831030845642,
+      "grad_norm": 0.5376724600791931,
       "learning_rate": 0.00018171932348820234,
-      "loss": 1.8224,
+      "loss": 1.5099,
       "step": 586
     },
     {
       "epoch": 0.58716514019568,
-      "grad_norm": 0.6906039714813232,
+      "grad_norm": 0.513921856880188,
       "learning_rate": 0.0001816588615099893,
-      "loss": 1.6498,
+      "loss": 1.3213,
       "step": 587
     },
     {
       "epoch": 0.5881654215248039,
-      "grad_norm": 0.726737380027771,
+      "grad_norm": 0.7540159225463867,
       "learning_rate": 0.00018159830980345548,
-      "loss": 1.5377,
+      "loss": 1.2231,
       "step": 588
     },
     {
       "epoch": 0.5891657028539277,
-      "grad_norm": 0.6792006492614746,
+      "grad_norm": 0.5917702317237854,
       "learning_rate": 0.0001815376684351362,
-      "loss": 1.8905,
+      "loss": 1.6094,
       "step": 589
     },
     {
       "epoch": 0.5901659841830514,
-      "grad_norm": 0.7885284423828125,
+      "grad_norm": 0.5507463216781616,
       "learning_rate": 0.00018147693747166534,
-      "loss": 1.709,
+      "loss": 1.3904,
       "step": 590
     },
     {
       "epoch": 0.5911662655121753,
-      "grad_norm": 0.7270367741584778,
+      "grad_norm": 0.545695960521698,
       "learning_rate": 0.00018141611697977529,
-      "loss": 1.7973,
+      "loss": 1.5172,
       "step": 591
     },
     {
       "epoch": 0.5921665468412991,
-      "grad_norm": 0.7852438688278198,
+      "grad_norm": 0.5876530408859253,
       "learning_rate": 0.00018135520702629675,
-      "loss": 1.6312,
+      "loss": 1.3676,
       "step": 592
     },
     {
       "epoch": 0.5931668281704229,
-      "grad_norm": 0.7881343960762024,
+      "grad_norm": 0.5510894060134888,
       "learning_rate": 0.0001812942076781588,
-      "loss": 1.7581,
+      "loss": 1.4379,
       "step": 593
     },
     {
       "epoch": 0.5941671094995468,
-      "grad_norm": 0.7581279277801514,
+      "grad_norm": 0.5105913877487183,
       "learning_rate": 0.0001812331190023886,
-      "loss": 1.6811,
+      "loss": 1.3687,
       "step": 594
     },
     {
       "epoch": 0.5951673908286705,
-      "grad_norm": 0.7250573039054871,
+      "grad_norm": 0.47876060009002686,
       "learning_rate": 0.0001811719410661116,
-      "loss": 1.6835,
+      "loss": 1.3178,
       "step": 595
     },
     {
       "epoch": 0.5961676721577944,
-      "grad_norm": 0.7383652329444885,
+      "grad_norm": 0.6079074144363403,
       "learning_rate": 0.00018111067393655132,
-      "loss": 1.7804,
+      "loss": 1.4713,
       "step": 596
     },
     {
       "epoch": 0.5971679534869182,
-      "grad_norm": 0.6631885170936584,
+      "grad_norm": 0.5363487601280212,
       "learning_rate": 0.0001810493176810292,
-      "loss": 1.477,
+      "loss": 1.1868,
       "step": 597
     },
     {
       "epoch": 0.598168234816042,
-      "grad_norm": 0.6705698370933533,
+      "grad_norm": 0.5252292156219482,
       "learning_rate": 0.00018098787236696474,
-      "loss": 1.5939,
+      "loss": 1.303,
       "step": 598
     },
     {
       "epoch": 0.5991685161451659,
-      "grad_norm": 0.6646862626075745,
+      "grad_norm": 0.5377137064933777,
       "learning_rate": 0.00018092633806187513,
-      "loss": 1.5903,
+      "loss": 1.3653,
       "step": 599
     },
     {
       "epoch": 0.6001687974742896,
-      "grad_norm": 0.7267604470252991,
+      "grad_norm": 0.5274302363395691,
       "learning_rate": 0.0001808647148333755,
-      "loss": 1.6864,
+      "loss": 1.3693,
       "step": 600
     },
     {
       "epoch": 0.6011690788034134,
-      "grad_norm": 0.672102689743042,
+      "grad_norm": 0.5664658546447754,
       "learning_rate": 0.00018080300274917862,
-      "loss": 1.73,
+      "loss": 1.3807,
       "step": 601
     },
     {
       "epoch": 0.6021693601325373,
-      "grad_norm": 0.7541230320930481,
+      "grad_norm": 0.6609538197517395,
       "learning_rate": 0.00018074120187709495,
-      "loss": 1.7824,
+      "loss": 1.5015,
       "step": 602
     },
     {
       "epoch": 0.6031696414616611,
-      "grad_norm": 0.6589316129684448,
+      "grad_norm": 0.4943195879459381,
       "learning_rate": 0.00018067931228503246,
-      "loss": 1.7007,
+      "loss": 1.4436,
       "step": 603
     },
     {
       "epoch": 0.604169922790785,
-      "grad_norm": 0.7083007097244263,
+      "grad_norm": 0.549712598323822,
       "learning_rate": 0.00018061733404099655,
-      "loss": 1.7703,
+      "loss": 1.455,
       "step": 604
     },
     {
       "epoch": 0.6051702041199087,
-      "grad_norm": 0.6700689196586609,
+      "grad_norm": 0.5765941143035889,
       "learning_rate": 0.00018055526721309016,
-      "loss": 1.5967,
+      "loss": 1.3317,
       "step": 605
     },
     {
       "epoch": 0.6061704854490325,
-      "grad_norm": 0.6766354441642761,
+      "grad_norm": 0.5223068594932556,
       "learning_rate": 0.0001804931118695135,
-      "loss": 1.6208,
+      "loss": 1.3456,
       "step": 606
     },
     {
       "epoch": 0.6071707667781564,
-      "grad_norm": 0.8214102387428284,
+      "grad_norm": 0.5385129451751709,
       "learning_rate": 0.00018043086807856403,
-      "loss": 1.6084,
+      "loss": 1.3388,
       "step": 607
     },
     {
       "epoch": 0.6081710481072802,
-      "grad_norm": 0.736492395401001,
+      "grad_norm": 0.5244528651237488,
       "learning_rate": 0.00018036853590863648,
-      "loss": 1.7254,
+      "loss": 1.398,
       "step": 608
     },
     {
       "epoch": 0.609171329436404,
-      "grad_norm": 0.6791033148765564,
+      "grad_norm": 0.5274112224578857,
       "learning_rate": 0.00018030611542822257,
-      "loss": 1.5837,
+      "loss": 1.3105,
       "step": 609
     },
     {
       "epoch": 0.6101716107655278,
-      "grad_norm": 0.6344060301780701,
+      "grad_norm": 0.5351893305778503,
       "learning_rate": 0.00018024360670591114,
-      "loss": 1.5407,
+      "loss": 1.3128,
       "step": 610
     },
     {
       "epoch": 0.6111718920946516,
-      "grad_norm": 0.9026575088500977,
+      "grad_norm": 0.5729460120201111,
       "learning_rate": 0.00018018100981038798,
-      "loss": 1.6748,
+      "loss": 1.3606,
       "step": 611
     },
     {
       "epoch": 0.6121721734237754,
-      "grad_norm": 0.8029866814613342,
+      "grad_norm": 0.5494408011436462,
       "learning_rate": 0.00018011832481043576,
-      "loss": 1.7547,
+      "loss": 1.4517,
       "step": 612
     },
     {
       "epoch": 0.6131724547528993,
-      "grad_norm": 0.8065117597579956,
+      "grad_norm": 0.5205882787704468,
       "learning_rate": 0.00018005555177493394,
-      "loss": 1.8051,
+      "loss": 1.4943,
       "step": 613
     },
     {
       "epoch": 0.614172736082023,
-      "grad_norm": 0.7858480215072632,
+      "grad_norm": 0.5488479137420654,
       "learning_rate": 0.00017999269077285875,
-      "loss": 1.6728,
+      "loss": 1.3939,
       "step": 614
     },
     {
       "epoch": 0.6151730174111469,
-      "grad_norm": 0.6735272407531738,
+      "grad_norm": 0.5779786109924316,
       "learning_rate": 0.00017992974187328305,
-      "loss": 1.8585,
+      "loss": 1.5744,
       "step": 615
     },
     {
       "epoch": 0.6161732987402707,
-      "grad_norm": 0.7518951892852783,
+      "grad_norm": 0.5576769113540649,
       "learning_rate": 0.00017986670514537627,
-      "loss": 1.5429,
+      "loss": 1.2284,
       "step": 616
     },
     {
       "epoch": 0.6171735800693945,
-      "grad_norm": 0.6952928900718689,
+      "grad_norm": 0.4912784993648529,
       "learning_rate": 0.00017980358065840444,
-      "loss": 1.5982,
+      "loss": 1.292,
       "step": 617
     },
     {
       "epoch": 0.6181738613985184,
-      "grad_norm": 0.8996840119361877,
+      "grad_norm": 0.657666027545929,
       "learning_rate": 0.0001797403684817299,
-      "loss": 1.8164,
+      "loss": 1.4918,
       "step": 618
     },
     {
       "epoch": 0.6191741427276422,
-      "grad_norm": 0.7645425200462341,
+      "grad_norm": 0.5642833113670349,
       "learning_rate": 0.00017967706868481144,
-      "loss": 1.8373,
+      "loss": 1.4718,
       "step": 619
     },
     {
       "epoch": 0.6201744240567659,
-      "grad_norm": 0.8479064106941223,
+      "grad_norm": 0.7243106961250305,
       "learning_rate": 0.00017961368133720407,
-      "loss": 1.6483,
+      "loss": 1.4342,
       "step": 620
     },
     {
       "epoch": 0.6211747053858898,
-      "grad_norm": 0.7806827425956726,
+      "grad_norm": 0.4982456564903259,
       "learning_rate": 0.000179550206508559,
-      "loss": 1.78,
+      "loss": 1.4478,
       "step": 621
     },
     {
       "epoch": 0.6221749867150136,
-      "grad_norm": 0.6476775407791138,
+      "grad_norm": 0.5249592065811157,
       "learning_rate": 0.00017948664426862364,
-      "loss": 1.7712,
+      "loss": 1.485,
       "step": 622
     },
     {
       "epoch": 0.6231752680441374,
-      "grad_norm": 0.8421279788017273,
+      "grad_norm": 0.6167681217193604,
       "learning_rate": 0.00017942299468724134,
-      "loss": 1.7753,
+      "loss": 1.4813,
       "step": 623
     },
     {
       "epoch": 0.6241755493732613,
-      "grad_norm": 0.6706071496009827,
+      "grad_norm": 0.5300460457801819,
       "learning_rate": 0.0001793592578343515,
-      "loss": 1.4093,
+      "loss": 1.1364,
       "step": 624
     },
     {
       "epoch": 0.625175830702385,
-      "grad_norm": 0.8224231004714966,
+      "grad_norm": 0.5908417105674744,
       "learning_rate": 0.0001792954337799894,
-      "loss": 1.7343,
+      "loss": 1.4402,
       "step": 625
     },
     {
       "epoch": 0.6261761120315089,
-      "grad_norm": 0.8398690819740295,
+      "grad_norm": 0.5684035420417786,
       "learning_rate": 0.00017923152259428612,
-      "loss": 1.8017,
+      "loss": 1.4847,
       "step": 626
     },
     {
       "epoch": 0.6271763933606327,
-      "grad_norm": 0.6664738059043884,
+      "grad_norm": 0.5421493053436279,
       "learning_rate": 0.00017916752434746856,
-      "loss": 1.6023,
+      "loss": 1.3348,
       "step": 627
     },
     {
       "epoch": 0.6281766746897565,
-      "grad_norm": 0.9246477484703064,
+      "grad_norm": 0.5295160412788391,
       "learning_rate": 0.0001791034391098591,
-      "loss": 1.7862,
+      "loss": 1.4703,
       "step": 628
     },
     {
       "epoch": 0.6291769560188804,
-      "grad_norm": 0.797835111618042,
+      "grad_norm": 0.5196051001548767,
       "learning_rate": 0.00017903926695187595,
-      "loss": 1.6059,
+      "loss": 1.3478,
       "step": 629
     },
     {
       "epoch": 0.6301772373480041,
-      "grad_norm": 0.613727331161499,
+      "grad_norm": 0.4994469881057739,
       "learning_rate": 0.0001789750079440326,
-      "loss": 1.5086,
+      "loss": 1.2368,
       "step": 630
     },
     {
       "epoch": 0.6311775186771279,
-      "grad_norm": 0.7127765417098999,
+      "grad_norm": 0.5117055177688599,
       "learning_rate": 0.00017891066215693817,
-      "loss": 1.5985,
+      "loss": 1.3429,
       "step": 631
     },
     {
       "epoch": 0.6321778000062518,
-      "grad_norm": 0.6923073530197144,
+      "grad_norm": 0.49438026547431946,
       "learning_rate": 0.00017884622966129695,
-      "loss": 1.5537,
+      "loss": 1.301,
       "step": 632
     },
     {
       "epoch": 0.6331780813353756,
-      "grad_norm": 0.7015733122825623,
+      "grad_norm": 0.6113334894180298,
       "learning_rate": 0.00017878171052790868,
-      "loss": 1.7782,
+      "loss": 1.4636,
       "step": 633
     },
     {
       "epoch": 0.6341783626644993,
-      "grad_norm": 0.6932784914970398,
+      "grad_norm": 0.6063141822814941,
       "learning_rate": 0.00017871710482766817,
-      "loss": 1.4294,
+      "loss": 1.2262,
       "step": 634
     },
     {
       "epoch": 0.6351786439936232,
-      "grad_norm": 0.7054254412651062,
+      "grad_norm": 0.5604403614997864,
       "learning_rate": 0.00017865241263156546,
-      "loss": 1.7227,
+      "loss": 1.4112,
       "step": 635
     },
     {
       "epoch": 0.636178925322747,
-      "grad_norm": 0.6994242072105408,
+      "grad_norm": 0.523415207862854,
       "learning_rate": 0.0001785876340106855,
-      "loss": 1.5998,
+      "loss": 1.3281,
       "step": 636
     },
     {
       "epoch": 0.6371792066518709,
-      "grad_norm": 0.81461501121521,
+      "grad_norm": 0.5602991580963135,
       "learning_rate": 0.0001785227690362083,
-      "loss": 1.7212,
+      "loss": 1.44,
       "step": 637
     },
     {
       "epoch": 0.6381794879809947,
-      "grad_norm": 0.943434476852417,
+      "grad_norm": 0.46946853399276733,
       "learning_rate": 0.00017845781777940878,
-      "loss": 1.5926,
+      "loss": 1.2956,
       "step": 638
     },
     {
       "epoch": 0.6391797693101184,
-      "grad_norm": 0.8455945253372192,
+      "grad_norm": 0.5586503744125366,
       "learning_rate": 0.00017839278031165658,
-      "loss": 1.8511,
+      "loss": 1.5419,
       "step": 639
     },
     {
       "epoch": 0.6401800506392423,
-      "grad_norm": 0.9348243474960327,
+      "grad_norm": 0.5270752310752869,
       "learning_rate": 0.00017832765670441612,
-      "loss": 1.6293,
+      "loss": 1.305,
       "step": 640
     },
     {
       "epoch": 0.6411803319683661,
-      "grad_norm": 0.746127724647522,
+      "grad_norm": 0.57756108045578,
       "learning_rate": 0.0001782624470292465,
-      "loss": 1.4903,
+      "loss": 1.2145,
       "step": 641
     },
     {
       "epoch": 0.6421806132974899,
-      "grad_norm": 0.6215783357620239,
+      "grad_norm": 0.5709058046340942,
       "learning_rate": 0.0001781971513578013,
-      "loss": 1.7806,
+      "loss": 1.4804,
       "step": 642
     },
     {
       "epoch": 0.6431808946266138,
-      "grad_norm": 0.7447994947433472,
+      "grad_norm": 0.505849301815033,
       "learning_rate": 0.00017813176976182873,
-      "loss": 1.7475,
+      "loss": 1.3964,
       "step": 643
     },
     {
       "epoch": 0.6441811759557375,
-      "grad_norm": 0.6916540265083313,
+      "grad_norm": 0.5171617269515991,
       "learning_rate": 0.00017806630231317127,
-      "loss": 1.6401,
+      "loss": 1.3283,
       "step": 644
     },
     {
       "epoch": 0.6451814572848613,
-      "grad_norm": 0.7208524942398071,
+      "grad_norm": 0.5567512512207031,
       "learning_rate": 0.00017800074908376584,
-      "loss": 1.7524,
+      "loss": 1.481,
       "step": 645
     },
     {
       "epoch": 0.6461817386139852,
-      "grad_norm": 0.7548331618309021,
+      "grad_norm": 0.5000666379928589,
       "learning_rate": 0.00017793511014564358,
-      "loss": 1.5644,
+      "loss": 1.2856,
       "step": 646
     },
     {
       "epoch": 0.647182019943109,
-      "grad_norm": 0.7919667959213257,
+      "grad_norm": 0.49550777673721313,
       "learning_rate": 0.00017786938557092983,
-      "loss": 1.6758,
+      "loss": 1.3447,
       "step": 647
     },
     {
       "epoch": 0.6481823012722329,
-      "grad_norm": 0.700618326663971,
+      "grad_norm": 0.5904624462127686,
       "learning_rate": 0.00017780357543184397,
-      "loss": 1.5213,
+      "loss": 1.241,
       "step": 648
     },
     {
       "epoch": 0.6491825826013566,
-      "grad_norm": 0.6646535992622375,
+      "grad_norm": 0.4615901708602905,
       "learning_rate": 0.00017773767980069945,
-      "loss": 1.6487,
+      "loss": 1.3436,
       "step": 649
     },
     {
       "epoch": 0.6501828639304804,
-      "grad_norm": 0.6486669182777405,
+      "grad_norm": 0.48083069920539856,
       "learning_rate": 0.0001776716987499037,
-      "loss": 1.6556,
+      "loss": 1.3906,
       "step": 650
     },
     {
       "epoch": 0.6511831452596043,
-      "grad_norm": 0.657747745513916,
+      "grad_norm": 0.4525931775569916,
       "learning_rate": 0.0001776056323519579,
-      "loss": 1.5943,
+      "loss": 1.3417,
       "step": 651
     },
     {
       "epoch": 0.6521834265887281,
-      "grad_norm": 0.7777379751205444,
+      "grad_norm": 0.6179555058479309,
       "learning_rate": 0.00017753948067945712,
-      "loss": 1.6069,
+      "loss": 1.3438,
       "step": 652
     },
     {
       "epoch": 0.6531837079178519,
-      "grad_norm": 0.772153913974762,
+      "grad_norm": 0.5525293946266174,
       "learning_rate": 0.00017747324380509006,
-      "loss": 1.7065,
+      "loss": 1.4551,
       "step": 653
     },
     {
       "epoch": 0.6541839892469757,
-      "grad_norm": 0.6984367966651917,
+      "grad_norm": 0.533028781414032,
       "learning_rate": 0.00017740692180163908,
-      "loss": 1.7122,
+      "loss": 1.4396,
       "step": 654
     },
     {
       "epoch": 0.6551842705760995,
-      "grad_norm": 0.8033855557441711,
+      "grad_norm": 0.5196881890296936,
       "learning_rate": 0.00017734051474198003,
-      "loss": 1.6095,
+      "loss": 1.3032,
       "step": 655
     },
     {
       "epoch": 0.6561845519052233,
-      "grad_norm": 0.7568691372871399,
+      "grad_norm": 0.5190469622612,
       "learning_rate": 0.0001772740226990823,
-      "loss": 1.6783,
+      "loss": 1.4049,
       "step": 656
     },
     {
       "epoch": 0.6571848332343472,
-      "grad_norm": 0.7288162708282471,
+      "grad_norm": 0.49517175555229187,
       "learning_rate": 0.00017720744574600863,
-      "loss": 1.695,
+      "loss": 1.3696,
       "step": 657
     },
     {
       "epoch": 0.658185114563471,
-      "grad_norm": 0.6898120045661926,
+      "grad_norm": 0.5165138244628906,
       "learning_rate": 0.00017714078395591502,
-      "loss": 1.6539,
+      "loss": 1.3667,
       "step": 658
     },
     {
       "epoch": 0.6591853958925948,
-      "grad_norm": 0.6977367997169495,
+      "grad_norm": 0.5624507665634155,
       "learning_rate": 0.00017707403740205071,
-      "loss": 1.4558,
+      "loss": 1.2109,
       "step": 659
     },
     {
       "epoch": 0.6601856772217186,
-      "grad_norm": 0.6594682335853577,
+      "grad_norm": 0.45942649245262146,
       "learning_rate": 0.00017700720615775812,
-      "loss": 1.56,
+      "loss": 1.259,
       "step": 660
     },
     {
       "epoch": 0.6611859585508424,
-      "grad_norm": 0.6146736741065979,
+      "grad_norm": 0.5019019842147827,
       "learning_rate": 0.0001769402902964727,
-      "loss": 1.7014,
+      "loss": 1.3739,
       "step": 661
     },
     {
       "epoch": 0.6621862398799663,
-      "grad_norm": 0.7182234525680542,
+      "grad_norm": 0.4661652743816376,
       "learning_rate": 0.00017687328989172288,
-      "loss": 1.5655,
+      "loss": 1.2606,
       "step": 662
     },
     {
       "epoch": 0.66318652120909,
-      "grad_norm": 0.6940692067146301,
+      "grad_norm": 0.5310545563697815,
       "learning_rate": 0.00017680620501712996,
-      "loss": 1.6177,
+      "loss": 1.3406,
       "step": 663
     },
     {
       "epoch": 0.6641868025382138,
-      "grad_norm": 0.7672961950302124,
+      "grad_norm": 0.5190532207489014,
       "learning_rate": 0.00017673903574640814,
-      "loss": 1.559,
+      "loss": 1.3052,
       "step": 664
     },
     {
       "epoch": 0.6651870838673377,
-      "grad_norm": 0.654500424861908,
+      "grad_norm": 0.5265533328056335,
       "learning_rate": 0.00017667178215336423,
-      "loss": 1.5024,
+      "loss": 1.2326,
       "step": 665
     },
     {
       "epoch": 0.6661873651964615,
-      "grad_norm": 0.8137261867523193,
+      "grad_norm": 0.5971291065216064,
       "learning_rate": 0.0001766044443118978,
-      "loss": 1.7865,
+      "loss": 1.4291,
       "step": 666
     },
     {
       "epoch": 0.6671876465255854,
-      "grad_norm": 0.806624710559845,
+      "grad_norm": 0.5295760631561279,
       "learning_rate": 0.000176537022296001,
-      "loss": 1.4944,
+      "loss": 1.2781,
       "step": 667
     },
     {
       "epoch": 0.6681879278547092,
-      "grad_norm": 0.7952747941017151,
+      "grad_norm": 0.5124595761299133,
       "learning_rate": 0.00017646951617975837,
-      "loss": 1.5371,
+      "loss": 1.318,
       "step": 668
     },
     {
       "epoch": 0.6691882091838329,
-      "grad_norm": 0.6380738615989685,
+      "grad_norm": 0.5968078970909119,
       "learning_rate": 0.00017640192603734692,
-      "loss": 1.3117,
+      "loss": 1.1483,
       "step": 669
     },
     {
       "epoch": 0.6701884905129568,
-      "grad_norm": 0.6559002995491028,
+      "grad_norm": 0.6211404204368591,
       "learning_rate": 0.00017633425194303606,
-      "loss": 1.3662,
+      "loss": 1.1164,
       "step": 670
     },
     {
       "epoch": 0.6711887718420806,
-      "grad_norm": 0.715826153755188,
+      "grad_norm": 0.5539883375167847,
       "learning_rate": 0.00017626649397118734,
-      "loss": 1.7271,
+      "loss": 1.453,
       "step": 671
     },
     {
       "epoch": 0.6721890531712044,
-      "grad_norm": 0.6719872355461121,
+      "grad_norm": 0.5188294649124146,
       "learning_rate": 0.00017619865219625452,
-      "loss": 1.747,
+      "loss": 1.5201,
       "step": 672
     },
     {
       "epoch": 0.6731893345003283,
-      "grad_norm": 0.6901715397834778,
+      "grad_norm": 0.531973659992218,
       "learning_rate": 0.00017613072669278343,
-      "loss": 1.6438,
+      "loss": 1.3176,
       "step": 673
     },
     {
       "epoch": 0.674189615829452,
-      "grad_norm": 0.6601479649543762,
+      "grad_norm": 0.5878707766532898,
       "learning_rate": 0.00017606271753541192,
-      "loss": 1.8191,
+      "loss": 1.5326,
       "step": 674
     },
     {
       "epoch": 0.6751898971585758,
-      "grad_norm": 0.8059187531471252,
+      "grad_norm": 0.595443844795227,
       "learning_rate": 0.00017599462479886974,
-      "loss": 1.6946,
+      "loss": 1.4033,
       "step": 675
     },
     {
       "epoch": 0.6761901784876997,
-      "grad_norm": 0.6966856718063354,
+      "grad_norm": 0.5093846321105957,
       "learning_rate": 0.00017592644855797854,
-      "loss": 1.5551,
+      "loss": 1.2995,
       "step": 676
     },
     {
       "epoch": 0.6771904598168235,
-      "grad_norm": 0.7306144833564758,
+      "grad_norm": 0.5521978735923767,
       "learning_rate": 0.00017585818888765168,
-      "loss": 1.5429,
+      "loss": 1.2912,
       "step": 677
     },
     {
       "epoch": 0.6781907411459474,
-      "grad_norm": 0.572907030582428,
+      "grad_norm": 0.4612530469894409,
       "learning_rate": 0.0001757898458628941,
-      "loss": 1.4437,
+      "loss": 1.1902,
       "step": 678
     },
     {
       "epoch": 0.6791910224750711,
-      "grad_norm": 0.6807466149330139,
+      "grad_norm": 0.4973600506782532,
       "learning_rate": 0.00017572141955880252,
-      "loss": 1.6307,
+      "loss": 1.3547,
       "step": 679
     },
     {
       "epoch": 0.6801913038041949,
-      "grad_norm": 0.7529204487800598,
+      "grad_norm": 0.606407105922699,
       "learning_rate": 0.00017565291005056504,
-      "loss": 1.631,
+      "loss": 1.371,
       "step": 680
     },
     {
       "epoch": 0.6811915851333188,
-      "grad_norm": 0.6292940378189087,
+      "grad_norm": 0.5027814507484436,
       "learning_rate": 0.00017558431741346122,
-      "loss": 1.7512,
+      "loss": 1.4551,
       "step": 681
     },
     {
       "epoch": 0.6821918664624426,
-      "grad_norm": 0.7981480956077576,
+      "grad_norm": 0.5732039213180542,
       "learning_rate": 0.00017551564172286197,
-      "loss": 1.7704,
+      "loss": 1.4181,
       "step": 682
     },
     {
       "epoch": 0.6831921477915663,
-      "grad_norm": 0.7816259860992432,
+      "grad_norm": 0.6327995657920837,
       "learning_rate": 0.00017544688305422943,
-      "loss": 1.4954,
+      "loss": 1.237,
       "step": 683
     },
     {
       "epoch": 0.6841924291206902,
-      "grad_norm": 0.6866456866264343,
+      "grad_norm": 0.5779625177383423,
       "learning_rate": 0.00017537804148311695,
-      "loss": 1.7986,
+      "loss": 1.5356,
       "step": 684
     },
     {
       "epoch": 0.685192710449814,
-      "grad_norm": 0.7499064803123474,
+      "grad_norm": 0.6031951308250427,
       "learning_rate": 0.00017530911708516902,
-      "loss": 1.6472,
+      "loss": 1.3776,
       "step": 685
     },
     {
       "epoch": 0.6861929917789378,
-      "grad_norm": 0.5923457145690918,
+      "grad_norm": 0.4811258018016815,
       "learning_rate": 0.00017524010993612098,
-      "loss": 1.4866,
+      "loss": 1.185,
       "step": 686
     },
     {
       "epoch": 0.6871932731080617,
-      "grad_norm": 0.6991822719573975,
+      "grad_norm": 0.5048002600669861,
       "learning_rate": 0.00017517102011179933,
-      "loss": 1.605,
+      "loss": 1.3335,
       "step": 687
     },
     {
       "epoch": 0.6881935544371854,
-      "grad_norm": 0.7880247235298157,
+      "grad_norm": 0.5963343977928162,
       "learning_rate": 0.0001751018476881212,
-      "loss": 1.641,
+      "loss": 1.4326,
       "step": 688
     },
     {
       "epoch": 0.6891938357663093,
-      "grad_norm": 0.7848097085952759,
+      "grad_norm": 0.4770168960094452,
       "learning_rate": 0.00017503259274109464,
-      "loss": 1.7505,
+      "loss": 1.4664,
       "step": 689
     },
     {
       "epoch": 0.6901941170954331,
-      "grad_norm": 0.693678081035614,
+      "grad_norm": 0.5020537376403809,
       "learning_rate": 0.00017496325534681825,
-      "loss": 1.6565,
+      "loss": 1.349,
       "step": 690
     },
     {
       "epoch": 0.6911943984245569,
-      "grad_norm": 0.8232877254486084,
+      "grad_norm": 0.5567785501480103,
       "learning_rate": 0.00017489383558148136,
-      "loss": 1.7664,
+      "loss": 1.452,
       "step": 691
     },
     {
       "epoch": 0.6921946797536808,
-      "grad_norm": 0.7834855914115906,
+      "grad_norm": 0.5167350769042969,
       "learning_rate": 0.00017482433352136365,
-      "loss": 1.4381,
+      "loss": 1.1148,
       "step": 692
     },
     {
       "epoch": 0.6931949610828045,
-      "grad_norm": 0.6186713576316833,
+      "grad_norm": 0.6030716300010681,
       "learning_rate": 0.00017475474924283536,
-      "loss": 1.6482,
+      "loss": 1.3473,
       "step": 693
     },
     {
       "epoch": 0.6941952424119283,
-      "grad_norm": 0.7511133551597595,
+      "grad_norm": 0.5643062591552734,
       "learning_rate": 0.00017468508282235704,
-      "loss": 1.6186,
+      "loss": 1.3476,
       "step": 694
     },
     {
       "epoch": 0.6951955237410522,
-      "grad_norm": 0.8017745614051819,
+      "grad_norm": 0.5124102234840393,
       "learning_rate": 0.00017461533433647946,
-      "loss": 1.6597,
+      "loss": 1.339,
       "step": 695
     },
     {
       "epoch": 0.696195805070176,
-      "grad_norm": 0.8190794587135315,
+      "grad_norm": 0.5690215229988098,
       "learning_rate": 0.00017454550386184362,
-      "loss": 1.6602,
+      "loss": 1.3816,
       "step": 696
     },
     {
       "epoch": 0.6971960863992998,
-      "grad_norm": 0.7479042410850525,
+      "grad_norm": 0.5938367247581482,
       "learning_rate": 0.00017447559147518055,
-      "loss": 1.77,
+      "loss": 1.4554,
       "step": 697
     },
     {
       "epoch": 0.6981963677284236,
-      "grad_norm": 0.7239962816238403,
+      "grad_norm": 0.5288996696472168,
       "learning_rate": 0.00017440559725331135,
-      "loss": 1.5838,
+      "loss": 1.2904,
       "step": 698
     },
     {
       "epoch": 0.6991966490575474,
-      "grad_norm": 0.7252762317657471,
+      "grad_norm": 0.5047140121459961,
       "learning_rate": 0.000174335521273147,
-      "loss": 1.5462,
+      "loss": 1.2362,
       "step": 699
     },
     {
       "epoch": 0.7001969303866713,
-      "grad_norm": 0.9383960962295532,
+      "grad_norm": 0.5563321709632874,
       "learning_rate": 0.00017426536361168834,
-      "loss": 1.5104,
+      "loss": 1.2863,
       "step": 700
     },
     {
       "epoch": 0.7011972117157951,
-      "grad_norm": 0.6944159269332886,
+      "grad_norm": 0.48857688903808594,
       "learning_rate": 0.00017419512434602594,
-      "loss": 1.6382,
+      "loss": 1.3387,
       "step": 701
     },
     {
       "epoch": 0.7021974930449189,
-      "grad_norm": 0.6809273362159729,
+      "grad_norm": 0.5205016732215881,
       "learning_rate": 0.00017412480355334005,
-      "loss": 1.725,
+      "loss": 1.3874,
       "step": 702
     },
     {
       "epoch": 0.7031977743740427,
-      "grad_norm": 0.7521125674247742,
+      "grad_norm": 0.5850381851196289,
       "learning_rate": 0.00017405440131090048,
-      "loss": 1.8499,
+      "loss": 1.5369,
       "step": 703
     },
     {
       "epoch": 0.7041980557031665,
-      "grad_norm": 0.6854100227355957,
+      "grad_norm": 0.5708681344985962,
       "learning_rate": 0.00017398391769606658,
-      "loss": 1.6648,
+      "loss": 1.3622,
       "step": 704
     },
     {
       "epoch": 0.7051983370322903,
-      "grad_norm": 0.7382327318191528,
+      "grad_norm": 0.5743641257286072,
       "learning_rate": 0.00017391335278628712,
-      "loss": 1.5806,
+      "loss": 1.2946,
       "step": 705
     },
     {
       "epoch": 0.7061986183614142,
-      "grad_norm": 0.7387582063674927,
+      "grad_norm": 0.5376024842262268,
       "learning_rate": 0.00017384270665910014,
-      "loss": 1.5563,
+      "loss": 1.2952,
       "step": 706
     },
     {
       "epoch": 0.707198899690538,
-      "grad_norm": 0.7698972821235657,
+      "grad_norm": 0.6123641133308411,
       "learning_rate": 0.000173771979392133,
-      "loss": 1.6626,
+      "loss": 1.4239,
       "step": 707
     },
     {
       "epoch": 0.7081991810196617,
-      "grad_norm": 0.7639899849891663,
+      "grad_norm": 0.5639240741729736,
       "learning_rate": 0.00017370117106310214,
-      "loss": 1.6725,
+      "loss": 1.3627,
       "step": 708
     },
     {
       "epoch": 0.7091994623487856,
-      "grad_norm": 0.6684393286705017,
+      "grad_norm": 0.5551653504371643,
       "learning_rate": 0.0001736302817498131,
-      "loss": 1.64,
+      "loss": 1.3435,
       "step": 709
     },
     {
       "epoch": 0.7101997436779094,
-      "grad_norm": 0.6329504251480103,
+      "grad_norm": 0.4746958911418915,
       "learning_rate": 0.00017355931153016044,
-      "loss": 1.4472,
+      "loss": 1.2402,
       "step": 710
     },
     {
       "epoch": 0.7112000250070333,
-      "grad_norm": 0.8133587837219238,
+      "grad_norm": 0.4722553491592407,
       "learning_rate": 0.0001734882604821276,
-      "loss": 1.7971,
+      "loss": 1.3962,
       "step": 711
     },
     {
       "epoch": 0.712200306336157,
-      "grad_norm": 0.6524143218994141,
+      "grad_norm": 0.5038101077079773,
       "learning_rate": 0.0001734171286837868,
-      "loss": 1.5366,
+      "loss": 1.3261,
       "step": 712
     },
     {
       "epoch": 0.7132005876652808,
-      "grad_norm": 0.6714311242103577,
+      "grad_norm": 0.5004639625549316,
       "learning_rate": 0.00017334591621329906,
-      "loss": 1.841,
+      "loss": 1.4943,
       "step": 713
     },
     {
       "epoch": 0.7142008689944047,
-      "grad_norm": 0.6690782904624939,
+      "grad_norm": 0.5141516327857971,
       "learning_rate": 0.00017327462314891402,
-      "loss": 1.623,
+      "loss": 1.2754,
       "step": 714
     },
     {
       "epoch": 0.7152011503235285,
-      "grad_norm": 0.650442361831665,
+      "grad_norm": 0.5491873025894165,
       "learning_rate": 0.00017320324956896977,
-      "loss": 1.6124,
+      "loss": 1.3052,
       "step": 715
     },
     {
       "epoch": 0.7162014316526523,
-      "grad_norm": 0.7075713276863098,
+      "grad_norm": 0.49937358498573303,
       "learning_rate": 0.00017313179555189306,
-      "loss": 1.5154,
+      "loss": 1.2277,
       "step": 716
     },
     {
       "epoch": 0.7172017129817762,
-      "grad_norm": 0.729060173034668,
+      "grad_norm": 0.6419594287872314,
       "learning_rate": 0.00017306026117619889,
-      "loss": 1.7072,
+      "loss": 1.4844,
       "step": 717
     },
     {
       "epoch": 0.7182019943108999,
-      "grad_norm": 0.8547433614730835,
+      "grad_norm": 0.521108090877533,
       "learning_rate": 0.0001729886465204906,
-      "loss": 1.6237,
+      "loss": 1.2917,
       "step": 718
     },
     {
       "epoch": 0.7192022756400237,
-      "grad_norm": 0.6729336380958557,
+      "grad_norm": 0.532421886920929,
       "learning_rate": 0.0001729169516634598,
-      "loss": 1.7769,
+      "loss": 1.4555,
       "step": 719
     },
     {
       "epoch": 0.7202025569691476,
-      "grad_norm": 0.7437167167663574,
+      "grad_norm": 0.5168073177337646,
       "learning_rate": 0.0001728451766838861,
-      "loss": 1.5056,
+      "loss": 1.2116,
       "step": 720
     },
     {
       "epoch": 0.7212028382982714,
-      "grad_norm": 0.6573147177696228,
+      "grad_norm": 0.5593972206115723,
       "learning_rate": 0.00017277332166063726,
-      "loss": 1.7694,
+      "loss": 1.4345,
       "step": 721
     },
     {
       "epoch": 0.7222031196273953,
-      "grad_norm": 0.6767126321792603,
+      "grad_norm": 0.5317432284355164,
       "learning_rate": 0.00017270138667266894,
-      "loss": 1.6014,
+      "loss": 1.2987,
       "step": 722
     },
     {
       "epoch": 0.723203400956519,
-      "grad_norm": 0.7488179206848145,
+      "grad_norm": 0.6262248158454895,
       "learning_rate": 0.00017262937179902472,
-      "loss": 1.573,
+      "loss": 1.2591,
       "step": 723
     },
     {
       "epoch": 0.7242036822856428,
-      "grad_norm": 0.6491002440452576,
+      "grad_norm": 0.5377100110054016,
       "learning_rate": 0.00017255727711883588,
-      "loss": 1.6705,
+      "loss": 1.366,
       "step": 724
     },
     {
       "epoch": 0.7252039636147667,
-      "grad_norm": 0.764090359210968,
+      "grad_norm": 0.5637168288230896,
       "learning_rate": 0.00017248510271132144,
-      "loss": 1.6761,
+      "loss": 1.4593,
       "step": 725
     },
     {
       "epoch": 0.7262042449438905,
-      "grad_norm": 0.7116997838020325,
+      "grad_norm": 0.5360320210456848,
       "learning_rate": 0.00017241284865578802,
-      "loss": 1.7435,
+      "loss": 1.4797,
       "step": 726
     },
     {
       "epoch": 0.7272045262730142,
-      "grad_norm": 0.6367645859718323,
+      "grad_norm": 0.48500168323516846,
       "learning_rate": 0.00017234051503162978,
-      "loss": 1.7061,
+      "loss": 1.3875,
       "step": 727
     },
     {
       "epoch": 0.7282048076021381,
-      "grad_norm": 0.7232155203819275,
+      "grad_norm": 0.5666176080703735,
       "learning_rate": 0.0001722681019183283,
-      "loss": 1.8142,
+      "loss": 1.4683,
       "step": 728
     },
     {
       "epoch": 0.7292050889312619,
-      "grad_norm": 0.7533649802207947,
+      "grad_norm": 0.5710940361022949,
       "learning_rate": 0.00017219560939545246,
-      "loss": 1.8202,
+      "loss": 1.5538,
       "step": 729
     },
     {
       "epoch": 0.7302053702603858,
-      "grad_norm": 0.6923018097877502,
+      "grad_norm": 0.5658044219017029,
       "learning_rate": 0.00017212303754265843,
-      "loss": 1.4925,
+      "loss": 1.248,
       "step": 730
     },
     {
       "epoch": 0.7312056515895096,
-      "grad_norm": 0.7326932549476624,
+      "grad_norm": 0.5355331301689148,
       "learning_rate": 0.0001720503864396896,
-      "loss": 1.5192,
+      "loss": 1.259,
       "step": 731
     },
     {
       "epoch": 0.7322059329186333,
-      "grad_norm": 0.7220762968063354,
+      "grad_norm": 0.5683363676071167,
       "learning_rate": 0.00017197765616637636,
-      "loss": 1.7601,
+      "loss": 1.4242,
       "step": 732
     },
     {
       "epoch": 0.7332062142477572,
-      "grad_norm": 0.605725884437561,
+      "grad_norm": 0.488972932100296,
       "learning_rate": 0.0001719048468026361,
-      "loss": 1.6309,
+      "loss": 1.3442,
       "step": 733
     },
     {
       "epoch": 0.734206495576881,
-      "grad_norm": 0.6728388667106628,
+      "grad_norm": 0.45563748478889465,
       "learning_rate": 0.00017183195842847322,
-      "loss": 1.5993,
+      "loss": 1.3236,
       "step": 734
     },
     {
       "epoch": 0.7352067769060048,
-      "grad_norm": 0.7035244703292847,
+      "grad_norm": 0.5114185214042664,
       "learning_rate": 0.0001717589911239788,
-      "loss": 1.6031,
+      "loss": 1.3071,
       "step": 735
     },
     {
       "epoch": 0.7362070582351287,
-      "grad_norm": 0.7473010420799255,
+      "grad_norm": 0.558686375617981,
       "learning_rate": 0.00017168594496933074,
-      "loss": 1.5833,
+      "loss": 1.2889,
       "step": 736
     },
     {
       "epoch": 0.7372073395642524,
-      "grad_norm": 0.6310701370239258,
+      "grad_norm": 0.49099281430244446,
       "learning_rate": 0.00017161282004479351,
-      "loss": 1.4328,
+      "loss": 1.1701,
       "step": 737
     },
     {
       "epoch": 0.7382076208933762,
-      "grad_norm": 0.6805673837661743,
+      "grad_norm": 0.549524188041687,
       "learning_rate": 0.0001715396164307182,
-      "loss": 1.5429,
+      "loss": 1.2853,
       "step": 738
     },
     {
       "epoch": 0.7392079022225001,
-      "grad_norm": 0.747222900390625,
+      "grad_norm": 0.5683863162994385,
       "learning_rate": 0.0001714663342075424,
-      "loss": 1.7696,
+      "loss": 1.4201,
       "step": 739
     },
     {
       "epoch": 0.7402081835516239,
-      "grad_norm": 0.8214403390884399,
+      "grad_norm": 0.5957104563713074,
       "learning_rate": 0.00017139297345578994,
-      "loss": 1.5997,
+      "loss": 1.3406,
       "step": 740
     },
     {
       "epoch": 0.7412084648807478,
-      "grad_norm": 0.6722521781921387,
+      "grad_norm": 0.4645147919654846,
       "learning_rate": 0.00017131953425607104,
-      "loss": 1.5287,
+      "loss": 1.2344,
       "step": 741
     },
     {
       "epoch": 0.7422087462098715,
-      "grad_norm": 0.6937971115112305,
+      "grad_norm": 0.4981783330440521,
       "learning_rate": 0.00017124601668908212,
-      "loss": 1.7263,
+      "loss": 1.422,
       "step": 742
     },
     {
       "epoch": 0.7432090275389953,
-      "grad_norm": 0.7590844631195068,
+      "grad_norm": 0.5426530838012695,
       "learning_rate": 0.00017117242083560568,
-      "loss": 1.7263,
+      "loss": 1.4275,
       "step": 743
     },
     {
       "epoch": 0.7442093088681192,
-      "grad_norm": 0.7913306355476379,
+      "grad_norm": 0.5585354566574097,
       "learning_rate": 0.00017109874677651024,
-      "loss": 1.7646,
+      "loss": 1.5049,
       "step": 744
     },
     {
       "epoch": 0.745209590197243,
-      "grad_norm": 0.7123669385910034,
+      "grad_norm": 0.5639151930809021,
       "learning_rate": 0.0001710249945927503,
-      "loss": 1.6768,
+      "loss": 1.4019,
       "step": 745
     },
     {
       "epoch": 0.7462098715263668,
-      "grad_norm": 0.8426288366317749,
+      "grad_norm": 0.8334717750549316,
       "learning_rate": 0.00017095116436536612,
-      "loss": 1.8496,
+      "loss": 1.5607,
       "step": 746
     },
     {
       "epoch": 0.7472101528554906,
-      "grad_norm": 0.6152015328407288,
+      "grad_norm": 0.513970673084259,
       "learning_rate": 0.00017087725617548385,
-      "loss": 1.4527,
+      "loss": 1.1967,
       "step": 747
     },
     {
       "epoch": 0.7482104341846144,
-      "grad_norm": 0.8348223567008972,
+      "grad_norm": 0.6200702786445618,
       "learning_rate": 0.00017080327010431513,
-      "loss": 1.4847,
+      "loss": 1.2298,
       "step": 748
     },
     {
       "epoch": 0.7492107155137382,
-      "grad_norm": 0.7883800268173218,
+      "grad_norm": 0.54522305727005,
       "learning_rate": 0.00017072920623315734,
-      "loss": 1.5941,
+      "loss": 1.3214,
       "step": 749
     },
     {
       "epoch": 0.7502109968428621,
-      "grad_norm": 0.6957768201828003,
+      "grad_norm": 0.6682360172271729,
       "learning_rate": 0.00017065506464339326,
-      "loss": 1.7543,
+      "loss": 1.4631,
       "step": 750
     },
     {
       "epoch": 0.7512112781719859,
-      "grad_norm": 0.5898700952529907,
+      "grad_norm": 0.5061535239219666,
       "learning_rate": 0.00017058084541649106,
-      "loss": 1.7859,
+      "loss": 1.5062,
       "step": 751
     },
     {
       "epoch": 0.7522115595011097,
-      "grad_norm": 0.6882239580154419,
+      "grad_norm": 0.5790627598762512,
       "learning_rate": 0.00017050654863400429,
-      "loss": 1.3233,
+      "loss": 1.1371,
       "step": 752
     },
     {
       "epoch": 0.7532118408302335,
-      "grad_norm": 0.7327316999435425,
+      "grad_norm": 0.6058077216148376,
       "learning_rate": 0.00017043217437757164,
-      "loss": 1.5067,
+      "loss": 1.2185,
       "step": 753
     },
     {
       "epoch": 0.7542121221593573,
-      "grad_norm": 0.9257964491844177,
+      "grad_norm": 0.5494515895843506,
       "learning_rate": 0.00017035772272891702,
-      "loss": 1.503,
+      "loss": 1.2468,
       "step": 754
     },
     {
       "epoch": 0.7552124034884812,
-      "grad_norm": 0.7924116253852844,
+      "grad_norm": 0.5687912106513977,
       "learning_rate": 0.00017028319376984928,
-      "loss": 1.8975,
+      "loss": 1.5621,
       "step": 755
     },
     {
       "epoch": 0.756212684817605,
-      "grad_norm": 0.6651099920272827,
+      "grad_norm": 0.5341185927391052,
       "learning_rate": 0.00017020858758226229,
-      "loss": 1.649,
+      "loss": 1.3598,
       "step": 756
     },
     {
       "epoch": 0.7572129661467287,
-      "grad_norm": 0.7257362604141235,
+      "grad_norm": 0.5373026132583618,
       "learning_rate": 0.0001701339042481347,
-      "loss": 1.6919,
+      "loss": 1.4185,
       "step": 757
     },
     {
       "epoch": 0.7582132474758526,
-      "grad_norm": 0.8733739852905273,
+      "grad_norm": 0.46508973836898804,
       "learning_rate": 0.00017005914384953007,
-      "loss": 1.5929,
+      "loss": 1.2962,
       "step": 758
     },
     {
       "epoch": 0.7592135288049764,
-      "grad_norm": 0.6347383856773376,
+      "grad_norm": 0.4580937325954437,
       "learning_rate": 0.00016998430646859654,
-      "loss": 1.3341,
+      "loss": 1.0707,
       "step": 759
     },
     {
       "epoch": 0.7602138101341002,
-      "grad_norm": 0.6915012001991272,
+      "grad_norm": 0.5277093052864075,
       "learning_rate": 0.00016990939218756683,
-      "loss": 1.4971,
+      "loss": 1.2529,
       "step": 760
     },
     {
       "epoch": 0.761214091463224,
-      "grad_norm": 0.7862069606781006,
+      "grad_norm": 0.5356671214103699,
       "learning_rate": 0.0001698344010887582,
-      "loss": 1.7468,
+      "loss": 1.4032,
       "step": 761
     },
     {
       "epoch": 0.7622143727923478,
-      "grad_norm": 0.7318029403686523,
+      "grad_norm": 0.6881769299507141,
       "learning_rate": 0.0001697593332545723,
-      "loss": 1.8143,
+      "loss": 1.4885,
       "step": 762
     },
     {
       "epoch": 0.7632146541214717,
-      "grad_norm": 0.6758155226707458,
+      "grad_norm": 0.5370383262634277,
       "learning_rate": 0.0001696841887674951,
-      "loss": 1.6652,
+      "loss": 1.3271,
       "step": 763
     },
     {
       "epoch": 0.7642149354505955,
-      "grad_norm": 0.6853237748146057,
+      "grad_norm": 0.4792316257953644,
       "learning_rate": 0.00016960896771009684,
-      "loss": 1.5176,
+      "loss": 1.2274,
       "step": 764
     },
     {
       "epoch": 0.7652152167797193,
-      "grad_norm": 0.9686934351921082,
+      "grad_norm": 0.5276592373847961,
       "learning_rate": 0.00016953367016503182,
-      "loss": 1.5366,
+      "loss": 1.2399,
       "step": 765
     },
     {
       "epoch": 0.7662154981088432,
-      "grad_norm": 0.7232028841972351,
+      "grad_norm": 0.4789050221443176,
       "learning_rate": 0.00016945829621503838,
-      "loss": 1.6932,
+      "loss": 1.4002,
       "step": 766
     },
     {
       "epoch": 0.7672157794379669,
-      "grad_norm": 0.6606596112251282,
+      "grad_norm": 0.492712140083313,
       "learning_rate": 0.00016938284594293897,
-      "loss": 1.7051,
+      "loss": 1.3897,
       "step": 767
     },
     {
       "epoch": 0.7682160607670907,
-      "grad_norm": 0.6337714195251465,
+      "grad_norm": 0.5009675621986389,
       "learning_rate": 0.00016930731943163972,
-      "loss": 1.6505,
+      "loss": 1.3797,
       "step": 768
     },
     {
       "epoch": 0.7692163420962146,
-      "grad_norm": 0.6292264461517334,
+      "grad_norm": 0.4863432049751282,
       "learning_rate": 0.00016923171676413063,
-      "loss": 1.7207,
+      "loss": 1.4251,
       "step": 769
     },
     {
       "epoch": 0.7702166234253384,
-      "grad_norm": 0.7183407545089722,
+      "grad_norm": 0.5190616846084595,
       "learning_rate": 0.00016915603802348535,
-      "loss": 1.7025,
+      "loss": 1.4265,
       "step": 770
     },
     {
       "epoch": 0.7712169047544621,
-      "grad_norm": 0.805107593536377,
+      "grad_norm": 0.5603469610214233,
       "learning_rate": 0.00016908028329286112,
-      "loss": 1.592,
+      "loss": 1.2852,
       "step": 771
     },
     {
       "epoch": 0.772217186083586,
-      "grad_norm": 0.725777804851532,
+      "grad_norm": 0.5128753185272217,
       "learning_rate": 0.0001690044526554987,
-      "loss": 1.6714,
+      "loss": 1.3324,
       "step": 772
     },
     {
       "epoch": 0.7732174674127098,
-      "grad_norm": 0.6801775097846985,
+      "grad_norm": 0.4992072284221649,
       "learning_rate": 0.00016892854619472223,
-      "loss": 1.5047,
+      "loss": 1.2498,
       "step": 773
     },
     {
       "epoch": 0.7742177487418337,
-      "grad_norm": 0.7701449990272522,
+      "grad_norm": 0.6128407716751099,
       "learning_rate": 0.00016885256399393924,
-      "loss": 1.5506,
+      "loss": 1.2967,
       "step": 774
     },
     {
       "epoch": 0.7752180300709575,
-      "grad_norm": 0.6954746842384338,
+      "grad_norm": 0.5186858177185059,
       "learning_rate": 0.00016877650613664034,
-      "loss": 1.4859,
+      "loss": 1.2654,
       "step": 775
     },
     {
       "epoch": 0.7762183114000812,
-      "grad_norm": 0.7431885004043579,
+      "grad_norm": 0.5207421183586121,
       "learning_rate": 0.00016870037270639942,
-      "loss": 1.6087,
+      "loss": 1.2994,
       "step": 776
     },
     {
       "epoch": 0.7772185927292051,
-      "grad_norm": 0.687329113483429,
+      "grad_norm": 0.509912371635437,
       "learning_rate": 0.0001686241637868734,
-      "loss": 1.7038,
+      "loss": 1.3971,
       "step": 777
     },
     {
       "epoch": 0.7782188740583289,
-      "grad_norm": 0.6656787395477295,
+      "grad_norm": 0.47703370451927185,
       "learning_rate": 0.00016854787946180198,
-      "loss": 1.5691,
+      "loss": 1.282,
       "step": 778
     },
     {
       "epoch": 0.7792191553874527,
-      "grad_norm": 0.7476064562797546,
+      "grad_norm": 0.5404442548751831,
       "learning_rate": 0.00016847151981500789,
-      "loss": 1.4972,
+      "loss": 1.1986,
       "step": 779
     },
     {
       "epoch": 0.7802194367165766,
-      "grad_norm": 0.7320332527160645,
+      "grad_norm": 0.541050136089325,
       "learning_rate": 0.00016839508493039657,
-      "loss": 1.7326,
+      "loss": 1.4478,
       "step": 780
     },
     {
       "epoch": 0.7812197180457003,
-      "grad_norm": 0.6432293057441711,
+      "grad_norm": 0.46520569920539856,
       "learning_rate": 0.00016831857489195618,
-      "loss": 1.542,
+      "loss": 1.2385,
       "step": 781
     },
     {
       "epoch": 0.7822199993748241,
-      "grad_norm": 0.6751729846000671,
+      "grad_norm": 0.5150445699691772,
       "learning_rate": 0.00016824198978375736,
-      "loss": 1.6864,
+      "loss": 1.3695,
       "step": 782
     },
     {
       "epoch": 0.783220280703948,
-      "grad_norm": 0.770193338394165,
+      "grad_norm": 0.5754334926605225,
       "learning_rate": 0.00016816532968995328,
-      "loss": 1.5318,
+      "loss": 1.3026,
       "step": 783
     },
     {
       "epoch": 0.7842205620330718,
-      "grad_norm": 0.6820619106292725,
+      "grad_norm": 0.5335776209831238,
       "learning_rate": 0.0001680885946947796,
-      "loss": 1.6004,
+      "loss": 1.3391,
       "step": 784
     },
     {
       "epoch": 0.7852208433621957,
-      "grad_norm": 0.9120951294898987,
+      "grad_norm": 0.6596659421920776,
       "learning_rate": 0.00016801178488255413,
-      "loss": 1.6506,
+      "loss": 1.3224,
       "step": 785
     },
     {
       "epoch": 0.7862211246913194,
-      "grad_norm": 0.7819542288780212,
+      "grad_norm": 0.5251991748809814,
       "learning_rate": 0.00016793490033767698,
-      "loss": 1.5292,
+      "loss": 1.1744,
       "step": 786
     },
     {
       "epoch": 0.7872214060204432,
-      "grad_norm": 0.6647278666496277,
+      "grad_norm": 0.5112204551696777,
       "learning_rate": 0.00016785794114463037,
-      "loss": 1.5941,
+      "loss": 1.2455,
       "step": 787
     },
     {
       "epoch": 0.7882216873495671,
-      "grad_norm": 0.6874713897705078,
+      "grad_norm": 0.532893717288971,
       "learning_rate": 0.00016778090738797853,
-      "loss": 1.5543,
+      "loss": 1.2437,
       "step": 788
     },
     {
       "epoch": 0.7892219686786909,
-      "grad_norm": 0.7759424448013306,
+      "grad_norm": 0.5534240007400513,
       "learning_rate": 0.00016770379915236766,
-      "loss": 1.6788,
+      "loss": 1.396,
       "step": 789
     },
     {
       "epoch": 0.7902222500078147,
-      "grad_norm": 0.724583625793457,
+      "grad_norm": 0.5164292454719543,
       "learning_rate": 0.00016762661652252567,
-      "loss": 1.5998,
+      "loss": 1.3138,
       "step": 790
     },
     {
       "epoch": 0.7912225313369385,
-      "grad_norm": 0.7921720743179321,
+      "grad_norm": 0.5660764575004578,
       "learning_rate": 0.00016754935958326244,
-      "loss": 1.5956,
+      "loss": 1.3014,
       "step": 791
     },
     {
       "epoch": 0.7922228126660623,
-      "grad_norm": 0.6484968662261963,
+      "grad_norm": 0.5137651562690735,
       "learning_rate": 0.00016747202841946928,
-      "loss": 1.5708,
+      "loss": 1.2834,
       "step": 792
     },
     {
       "epoch": 0.7932230939951862,
-      "grad_norm": 0.6372153759002686,
+      "grad_norm": 0.5546874403953552,
       "learning_rate": 0.00016739462311611919,
-      "loss": 1.5213,
+      "loss": 1.2841,
       "step": 793
     },
     {
       "epoch": 0.79422337532431,
-      "grad_norm": 0.7025095224380493,
+      "grad_norm": 0.5112007260322571,
       "learning_rate": 0.00016731714375826657,
-      "loss": 1.4701,
+      "loss": 1.1873,
       "step": 794
     },
     {
       "epoch": 0.7952236566534338,
-      "grad_norm": 0.681094765663147,
+      "grad_norm": 0.5462679862976074,
       "learning_rate": 0.00016723959043104728,
-      "loss": 1.5101,
+      "loss": 1.2602,
       "step": 795
     },
     {
       "epoch": 0.7962239379825576,
-      "grad_norm": 0.7129995822906494,
+      "grad_norm": 0.5083702802658081,
       "learning_rate": 0.00016716196321967832,
-      "loss": 1.6038,
+      "loss": 1.334,
       "step": 796
     },
     {
       "epoch": 0.7972242193116814,
-      "grad_norm": 0.7403759360313416,
+      "grad_norm": 0.5491913557052612,
       "learning_rate": 0.00016708426220945802,
-      "loss": 1.5906,
+      "loss": 1.335,
       "step": 797
     },
     {
       "epoch": 0.7982245006408052,
-      "grad_norm": 0.6562372446060181,
+      "grad_norm": 0.5257419943809509,
       "learning_rate": 0.00016700648748576574,
-      "loss": 1.6469,
+      "loss": 1.374,
       "step": 798
     },
     {
       "epoch": 0.7992247819699291,
-      "grad_norm": 0.839885413646698,
+      "grad_norm": 0.5252013206481934,
       "learning_rate": 0.0001669286391340618,
-      "loss": 1.5385,
+      "loss": 1.281,
       "step": 799
     },
     {
       "epoch": 0.8002250632990529,
-      "grad_norm": 0.8687535524368286,
+      "grad_norm": 0.5784058570861816,
       "learning_rate": 0.00016685071723988748,
-      "loss": 1.6759,
+      "loss": 1.385,
       "step": 800
     },
     {
       "epoch": 0.8012253446281766,
-      "grad_norm": 0.6825409531593323,
+      "grad_norm": 0.5508819818496704,
       "learning_rate": 0.00016677272188886483,
-      "loss": 1.841,
+      "loss": 1.5138,
       "step": 801
     },
     {
       "epoch": 0.8022256259573005,
-      "grad_norm": 0.6831037402153015,
+      "grad_norm": 0.5943104028701782,
       "learning_rate": 0.00016669465316669667,
-      "loss": 1.5476,
+      "loss": 1.2341,
       "step": 802
     },
     {
       "epoch": 0.8032259072864243,
-      "grad_norm": 0.6906002759933472,
+      "grad_norm": 0.5109750032424927,
       "learning_rate": 0.00016661651115916642,
-      "loss": 1.6866,
+      "loss": 1.361,
       "step": 803
     },
     {
       "epoch": 0.8042261886155482,
-      "grad_norm": 0.7675560116767883,
+      "grad_norm": 0.5322972536087036,
       "learning_rate": 0.00016653829595213794,
-      "loss": 1.5663,
+      "loss": 1.3383,
       "step": 804
     },
     {
       "epoch": 0.805226469944672,
-      "grad_norm": 0.6594063639640808,
+      "grad_norm": 0.4870489537715912,
       "learning_rate": 0.00016646000763155568,
-      "loss": 1.5247,
+      "loss": 1.2932,
       "step": 805
     },
     {
       "epoch": 0.8062267512737957,
-      "grad_norm": 0.7470384836196899,
+      "grad_norm": 0.6070749163627625,
       "learning_rate": 0.00016638164628344425,
-      "loss": 1.6468,
+      "loss": 1.3517,
       "step": 806
     },
     {
       "epoch": 0.8072270326029196,
-      "grad_norm": 0.6874479651451111,
+      "grad_norm": 0.5695485472679138,
       "learning_rate": 0.00016630321199390867,
-      "loss": 1.5948,
+      "loss": 1.295,
       "step": 807
     },
     {
       "epoch": 0.8082273139320434,
-      "grad_norm": 0.7301204204559326,
+      "grad_norm": 0.49092933535575867,
       "learning_rate": 0.00016622470484913406,
-      "loss": 1.3922,
+      "loss": 1.1708,
       "step": 808
     },
     {
       "epoch": 0.8092275952611672,
-      "grad_norm": 0.6781039834022522,
+      "grad_norm": 0.5488709807395935,
       "learning_rate": 0.00016614612493538551,
-      "loss": 1.6054,
+      "loss": 1.3101,
       "step": 809
     },
     {
       "epoch": 0.810227876590291,
-      "grad_norm": 0.6913226246833801,
+      "grad_norm": 0.6875150799751282,
       "learning_rate": 0.00016606747233900815,
-      "loss": 1.5754,
+      "loss": 1.3,
       "step": 810
     },
     {
       "epoch": 0.8112281579194148,
-      "grad_norm": 0.667425811290741,
+      "grad_norm": 0.5599775910377502,
       "learning_rate": 0.00016598874714642697,
-      "loss": 1.8492,
+      "loss": 1.5711,
       "step": 811
     },
     {
       "epoch": 0.8122284392485386,
-      "grad_norm": 0.7662241458892822,
+      "grad_norm": 0.7102994322776794,
       "learning_rate": 0.00016590994944414678,
-      "loss": 1.8034,
+      "loss": 1.4553,
       "step": 812
     },
     {
       "epoch": 0.8132287205776625,
-      "grad_norm": 0.7574827075004578,
+      "grad_norm": 0.5191233158111572,
       "learning_rate": 0.00016583107931875192,
-      "loss": 1.7435,
+      "loss": 1.4292,
       "step": 813
     },
     {
       "epoch": 0.8142290019067863,
-      "grad_norm": 0.9005519151687622,
+      "grad_norm": 0.4739600718021393,
       "learning_rate": 0.0001657521368569064,
-      "loss": 1.6769,
+      "loss": 1.3776,
       "step": 814
     },
     {
       "epoch": 0.8152292832359102,
-      "grad_norm": 0.6895585656166077,
+      "grad_norm": 0.5282078981399536,
       "learning_rate": 0.0001656731221453537,
-      "loss": 1.7562,
+      "loss": 1.4359,
       "step": 815
     },
     {
       "epoch": 0.8162295645650339,
-      "grad_norm": 0.7573346495628357,
+      "grad_norm": 0.690367579460144,
       "learning_rate": 0.00016559403527091675,
-      "loss": 1.4748,
+      "loss": 1.1747,
       "step": 816
     },
     {
       "epoch": 0.8172298458941577,
-      "grad_norm": 0.7698647975921631,
+      "grad_norm": 0.5715120434761047,
       "learning_rate": 0.0001655148763204977,
-      "loss": 1.6174,
+      "loss": 1.3289,
       "step": 817
     },
     {
       "epoch": 0.8182301272232816,
-      "grad_norm": 0.7975410223007202,
+      "grad_norm": 0.7024423480033875,
       "learning_rate": 0.00016543564538107797,
-      "loss": 1.7924,
+      "loss": 1.4758,
       "step": 818
     },
     {
       "epoch": 0.8192304085524054,
-      "grad_norm": 0.9687625169754028,
+      "grad_norm": 0.5568886399269104,
       "learning_rate": 0.00016535634253971794,
-      "loss": 1.7725,
+      "loss": 1.5172,
       "step": 819
     },
     {
       "epoch": 0.8202306898815291,
-      "grad_norm": 0.6777274012565613,
+      "grad_norm": 0.5847441554069519,
       "learning_rate": 0.00016527696788355714,
-      "loss": 1.5018,
+      "loss": 1.1993,
       "step": 820
     },
     {
       "epoch": 0.821230971210653,
-      "grad_norm": 0.6990464329719543,
+      "grad_norm": 0.5402149558067322,
       "learning_rate": 0.00016519752149981397,
-      "loss": 1.5804,
+      "loss": 1.2921,
       "step": 821
     },
     {
       "epoch": 0.8222312525397768,
-      "grad_norm": 0.8445940613746643,
+      "grad_norm": 0.6050311326980591,
       "learning_rate": 0.0001651180034757856,
-      "loss": 1.8591,
+      "loss": 1.59,
       "step": 822
     },
     {
       "epoch": 0.8232315338689006,
-      "grad_norm": 0.8462644815444946,
+      "grad_norm": 0.6215486526489258,
       "learning_rate": 0.00016503841389884798,
-      "loss": 1.7582,
+      "loss": 1.4562,
       "step": 823
     },
     {
       "epoch": 0.8242318151980245,
-      "grad_norm": 0.7679311037063599,
+      "grad_norm": 0.6507789492607117,
       "learning_rate": 0.00016495875285645566,
-      "loss": 1.5971,
+      "loss": 1.349,
       "step": 824
     },
     {
       "epoch": 0.8252320965271482,
-      "grad_norm": 0.7734447717666626,
+      "grad_norm": 0.5273147225379944,
       "learning_rate": 0.00016487902043614173,
-      "loss": 1.714,
+      "loss": 1.4016,
       "step": 825
     },
     {
       "epoch": 0.8262323778562721,
-      "grad_norm": 0.7890239953994751,
+      "grad_norm": 0.579987645149231,
       "learning_rate": 0.0001647992167255177,
-      "loss": 1.6876,
+      "loss": 1.4077,
       "step": 826
     },
     {
       "epoch": 0.8272326591853959,
-      "grad_norm": 0.8530203104019165,
+      "grad_norm": 0.5068405270576477,
       "learning_rate": 0.0001647193418122734,
-      "loss": 1.9096,
+      "loss": 1.5075,
       "step": 827
     },
     {
       "epoch": 0.8282329405145197,
-      "grad_norm": 0.7828260064125061,
+      "grad_norm": 0.519982099533081,
       "learning_rate": 0.00016463939578417692,
-      "loss": 1.5518,
+      "loss": 1.2721,
       "step": 828
     },
     {
       "epoch": 0.8292332218436436,
-      "grad_norm": 0.7015512585639954,
+      "grad_norm": 0.5181561708450317,
       "learning_rate": 0.0001645593787290745,
-      "loss": 1.49,
+      "loss": 1.2299,
       "step": 829
     },
     {
       "epoch": 0.8302335031727673,
-      "grad_norm": 0.694771409034729,
+      "grad_norm": 0.47413337230682373,
       "learning_rate": 0.0001644792907348904,
-      "loss": 1.5506,
+      "loss": 1.2462,
       "step": 830
     },
     {
       "epoch": 0.8312337845018911,
-      "grad_norm": 0.8167857527732849,
+      "grad_norm": 0.5426570773124695,
       "learning_rate": 0.00016439913188962685,
-      "loss": 1.7798,
+      "loss": 1.4496,
       "step": 831
     },
     {
       "epoch": 0.832234065831015,
-      "grad_norm": 0.6682108044624329,
+      "grad_norm": 0.5744379758834839,
       "learning_rate": 0.0001643189022813639,
-      "loss": 1.6107,
+      "loss": 1.3284,
       "step": 832
     },
     {
       "epoch": 0.8332343471601388,
-      "grad_norm": 0.8347259163856506,
+      "grad_norm": 0.49693235754966736,
       "learning_rate": 0.0001642386019982594,
-      "loss": 1.7672,
+      "loss": 1.4082,
       "step": 833
     },
     {
       "epoch": 0.8342346284892626,
-      "grad_norm": 0.6620945334434509,
+      "grad_norm": 0.5346773862838745,
       "learning_rate": 0.00016415823112854883,
-      "loss": 1.6975,
+      "loss": 1.4238,
       "step": 834
     },
     {
       "epoch": 0.8352349098183864,
-      "grad_norm": 0.7286327481269836,
+      "grad_norm": 0.6201802492141724,
       "learning_rate": 0.00016407778976054526,
-      "loss": 1.5956,
+      "loss": 1.3288,
       "step": 835
     },
     {
       "epoch": 0.8362351911475102,
-      "grad_norm": 0.6344440579414368,
+      "grad_norm": 0.5161807537078857,
       "learning_rate": 0.0001639972779826392,
-      "loss": 1.6455,
+      "loss": 1.3798,
       "step": 836
     },
     {
       "epoch": 0.8372354724766341,
-      "grad_norm": 0.6607793569564819,
+      "grad_norm": 0.4670160412788391,
       "learning_rate": 0.0001639166958832985,
-      "loss": 1.6739,
+      "loss": 1.3765,
       "step": 837
     },
     {
       "epoch": 0.8382357538057579,
-      "grad_norm": 0.6973574161529541,
+      "grad_norm": 0.6492543816566467,
       "learning_rate": 0.00016383604355106837,
-      "loss": 1.8042,
+      "loss": 1.5277,
       "step": 838
     },
     {
       "epoch": 0.8392360351348817,
-      "grad_norm": 0.7744210958480835,
+      "grad_norm": 0.5766328573226929,
       "learning_rate": 0.00016375532107457108,
-      "loss": 1.528,
+      "loss": 1.2481,
       "step": 839
     },
     {
       "epoch": 0.8402363164640055,
-      "grad_norm": 0.6944973468780518,
+      "grad_norm": 0.6431072950363159,
       "learning_rate": 0.00016367452854250603,
-      "loss": 1.5498,
+      "loss": 1.2755,
       "step": 840
     },
     {
       "epoch": 0.8412365977931293,
-      "grad_norm": 0.6730696558952332,
+      "grad_norm": 0.5121828317642212,
       "learning_rate": 0.00016359366604364972,
-      "loss": 1.5849,
+      "loss": 1.2927,
       "step": 841
     },
     {
       "epoch": 0.8422368791222531,
-      "grad_norm": 0.7051465511322021,
+      "grad_norm": 0.5222392678260803,
       "learning_rate": 0.00016351273366685526,
-      "loss": 1.5972,
+      "loss": 1.2626,
       "step": 842
     },
     {
       "epoch": 0.843237160451377,
-      "grad_norm": 0.7309426069259644,
+      "grad_norm": 0.5536903142929077,
       "learning_rate": 0.00016343173150105278,
-      "loss": 1.4612,
+      "loss": 1.1892,
       "step": 843
     },
     {
       "epoch": 0.8442374417805008,
-      "grad_norm": 0.7830431461334229,
+      "grad_norm": 0.5569381713867188,
       "learning_rate": 0.00016335065963524897,
-      "loss": 1.7208,
+      "loss": 1.4263,
       "step": 844
     },
     {
       "epoch": 0.8452377231096245,
-      "grad_norm": 0.8609834909439087,
+      "grad_norm": 0.6490715742111206,
       "learning_rate": 0.0001632695181585272,
-      "loss": 1.8229,
+      "loss": 1.452,
       "step": 845
     },
     {
       "epoch": 0.8462380044387484,
-      "grad_norm": 0.7489060759544373,
+      "grad_norm": 0.5965350270271301,
       "learning_rate": 0.00016318830716004722,
-      "loss": 1.6955,
+      "loss": 1.4189,
       "step": 846
     },
     {
       "epoch": 0.8472382857678722,
-      "grad_norm": 0.636900782585144,
+      "grad_norm": 0.45904603600502014,
       "learning_rate": 0.00016310702672904528,
-      "loss": 1.6664,
+      "loss": 1.4024,
       "step": 847
     },
     {
       "epoch": 0.8482385670969961,
-      "grad_norm": 0.6423529386520386,
+      "grad_norm": 0.4320334494113922,
       "learning_rate": 0.00016302567695483382,
-      "loss": 1.5356,
+      "loss": 1.2105,
       "step": 848
     },
     {
       "epoch": 0.8492388484261199,
-      "grad_norm": 0.7380033731460571,
+      "grad_norm": 0.527032196521759,
       "learning_rate": 0.0001629442579268016,
-      "loss": 1.4482,
+      "loss": 1.1996,
       "step": 849
     },
     {
       "epoch": 0.8502391297552436,
-      "grad_norm": 0.8258544206619263,
+      "grad_norm": 0.6317036747932434,
       "learning_rate": 0.00016286276973441333,
-      "loss": 1.7058,
+      "loss": 1.4811,
       "step": 850
     },
     {
       "epoch": 0.8512394110843675,
-      "grad_norm": 0.6473391056060791,
+      "grad_norm": 0.5726277828216553,
       "learning_rate": 0.00016278121246720987,
-      "loss": 1.5374,
+      "loss": 1.3249,
       "step": 851
     },
     {
       "epoch": 0.8522396924134913,
-      "grad_norm": 0.7097072005271912,
+      "grad_norm": 0.4624577462673187,
       "learning_rate": 0.00016269958621480788,
-      "loss": 1.6786,
+      "loss": 1.3291,
       "step": 852
     },
     {
       "epoch": 0.8532399737426151,
-      "grad_norm": 0.724993884563446,
+      "grad_norm": 0.5774461627006531,
       "learning_rate": 0.0001626178910668998,
-      "loss": 1.6022,
+      "loss": 1.2891,
       "step": 853
     },
     {
       "epoch": 0.854240255071739,
-      "grad_norm": 0.6800474524497986,
+      "grad_norm": 0.503584086894989,
       "learning_rate": 0.00016253612711325386,
-      "loss": 1.6382,
+      "loss": 1.3048,
       "step": 854
     },
     {
       "epoch": 0.8552405364008627,
-      "grad_norm": 0.6339759826660156,
+      "grad_norm": 0.4560583233833313,
       "learning_rate": 0.0001624542944437139,
-      "loss": 1.5641,
+      "loss": 1.2658,
       "step": 855
     },
     {
       "epoch": 0.8562408177299866,
-      "grad_norm": 0.6792349219322205,
+      "grad_norm": 0.49611610174179077,
       "learning_rate": 0.00016237239314819917,
-      "loss": 1.3713,
+      "loss": 1.1017,
       "step": 856
     },
     {
       "epoch": 0.8572410990591104,
-      "grad_norm": 0.6544696688652039,
+      "grad_norm": 0.5600405931472778,
       "learning_rate": 0.0001622904233167044,
-      "loss": 1.5639,
+      "loss": 1.3274,
       "step": 857
     },
     {
       "epoch": 0.8582413803882342,
-      "grad_norm": 0.7736073732376099,
+      "grad_norm": 0.5849353075027466,
       "learning_rate": 0.0001622083850392996,
-      "loss": 1.5454,
+      "loss": 1.274,
       "step": 858
     },
     {
       "epoch": 0.859241661717358,
-      "grad_norm": 0.8642422556877136,
+      "grad_norm": 0.5781377553939819,
       "learning_rate": 0.00016212627840613003,
-      "loss": 1.6852,
+      "loss": 1.4157,
       "step": 859
     },
     {
       "epoch": 0.8602419430464818,
-      "grad_norm": 0.6520773768424988,
+      "grad_norm": 0.4908173680305481,
       "learning_rate": 0.000162044103507416,
-      "loss": 1.5335,
+      "loss": 1.3,
       "step": 860
     },
     {
       "epoch": 0.8612422243756056,
-      "grad_norm": 0.7647336721420288,
+      "grad_norm": 0.5844553112983704,
       "learning_rate": 0.00016196186043345288,
-      "loss": 1.5578,
+      "loss": 1.2325,
       "step": 861
     },
     {
       "epoch": 0.8622425057047295,
-      "grad_norm": 0.9621163010597229,
+      "grad_norm": 0.5381117463111877,
       "learning_rate": 0.00016187954927461093,
-      "loss": 1.6976,
+      "loss": 1.41,
       "step": 862
     },
     {
       "epoch": 0.8632427870338533,
-      "grad_norm": 0.6847056746482849,
+      "grad_norm": 0.5468165278434753,
       "learning_rate": 0.00016179717012133521,
-      "loss": 1.7118,
+      "loss": 1.4272,
       "step": 863
     },
     {
       "epoch": 0.864243068362977,
-      "grad_norm": 0.7482467889785767,
+      "grad_norm": 0.5702970027923584,
       "learning_rate": 0.00016171472306414554,
-      "loss": 1.6601,
+      "loss": 1.3624,
       "step": 864
     },
     {
       "epoch": 0.8652433496921009,
-      "grad_norm": 0.7760444283485413,
+      "grad_norm": 0.5430637001991272,
       "learning_rate": 0.00016163220819363628,
-      "loss": 1.5587,
+      "loss": 1.2555,
       "step": 865
     },
     {
       "epoch": 0.8662436310212247,
-      "grad_norm": 0.8380980491638184,
+      "grad_norm": 0.5266844034194946,
       "learning_rate": 0.00016154962560047643,
-      "loss": 1.7171,
+      "loss": 1.3743,
       "step": 866
     },
     {
       "epoch": 0.8672439123503486,
-      "grad_norm": 0.6927618384361267,
+      "grad_norm": 0.5201333165168762,
       "learning_rate": 0.00016146697537540924,
-      "loss": 1.7244,
+      "loss": 1.3959,
       "step": 867
     },
     {
       "epoch": 0.8682441936794724,
-      "grad_norm": 0.7855746746063232,
+      "grad_norm": 0.44362199306488037,
       "learning_rate": 0.0001613842576092524,
-      "loss": 1.5848,
+      "loss": 1.2661,
       "step": 868
     },
     {
       "epoch": 0.8692444750085961,
-      "grad_norm": 0.6743006110191345,
+      "grad_norm": 0.5465226769447327,
       "learning_rate": 0.00016130147239289778,
-      "loss": 1.6969,
+      "loss": 1.3688,
       "step": 869
     },
     {
       "epoch": 0.87024475633772,
-      "grad_norm": 0.7060980200767517,
+      "grad_norm": 0.5353460907936096,
       "learning_rate": 0.00016121861981731135,
-      "loss": 1.5632,
+      "loss": 1.2327,
       "step": 870
     },
     {
       "epoch": 0.8712450376668438,
-      "grad_norm": 0.7673144340515137,
+      "grad_norm": 0.5463739633560181,
       "learning_rate": 0.00016113569997353312,
-      "loss": 1.5687,
+      "loss": 1.2994,
       "step": 871
     },
     {
       "epoch": 0.8722453189959676,
-      "grad_norm": 0.8105847239494324,
+      "grad_norm": 0.5219647288322449,
       "learning_rate": 0.000161052712952677,
-      "loss": 1.6074,
+      "loss": 1.3916,
       "step": 872
     },
     {
       "epoch": 0.8732456003250915,
-      "grad_norm": 0.6536850333213806,
+      "grad_norm": 0.4675636887550354,
       "learning_rate": 0.0001609696588459307,
-      "loss": 1.5842,
+      "loss": 1.2786,
       "step": 873
     },
     {
       "epoch": 0.8742458816542152,
-      "grad_norm": 0.6653574705123901,
+      "grad_norm": 0.48863986134529114,
       "learning_rate": 0.00016088653774455568,
-      "loss": 1.4652,
+      "loss": 1.1762,
       "step": 874
     },
     {
       "epoch": 0.875246162983339,
-      "grad_norm": 0.7202721238136292,
+      "grad_norm": 0.48759785294532776,
       "learning_rate": 0.00016080334973988695,
-      "loss": 1.5212,
+      "loss": 1.2107,
       "step": 875
     },
     {
       "epoch": 0.8762464443124629,
-      "grad_norm": 0.8218807578086853,
+      "grad_norm": 0.7353807687759399,
       "learning_rate": 0.00016072009492333318,
-      "loss": 1.803,
+      "loss": 1.4855,
       "step": 876
     },
     {
       "epoch": 0.8772467256415867,
-      "grad_norm": 0.6170400381088257,
+      "grad_norm": 0.4878953993320465,
       "learning_rate": 0.0001606367733863763,
-      "loss": 1.5313,
+      "loss": 1.2343,
       "step": 877
     },
     {
       "epoch": 0.8782470069707106,
-      "grad_norm": 0.6750448346138,
+      "grad_norm": 0.4764840304851532,
       "learning_rate": 0.00016055338522057158,
-      "loss": 1.6183,
+      "loss": 1.3159,
       "step": 878
     },
     {
       "epoch": 0.8792472882998343,
-      "grad_norm": 0.6602128148078918,
+      "grad_norm": 0.5289160013198853,
       "learning_rate": 0.00016046993051754756,
-      "loss": 1.6669,
+      "loss": 1.3298,
       "step": 879
     },
     {
       "epoch": 0.8802475696289581,
-      "grad_norm": 0.7064031958580017,
+      "grad_norm": 0.5421459078788757,
       "learning_rate": 0.00016038640936900586,
-      "loss": 1.7458,
+      "loss": 1.4081,
       "step": 880
     },
     {
       "epoch": 0.881247850958082,
-      "grad_norm": 0.5916783809661865,
+      "grad_norm": 0.5096681118011475,
       "learning_rate": 0.00016030282186672116,
-      "loss": 1.4966,
+      "loss": 1.2406,
       "step": 881
     },
     {
       "epoch": 0.8822481322872058,
-      "grad_norm": 0.7189202904701233,
+      "grad_norm": 0.5783627033233643,
       "learning_rate": 0.00016021916810254097,
-      "loss": 1.5812,
+      "loss": 1.3505,
       "step": 882
     },
     {
       "epoch": 0.8832484136163296,
-      "grad_norm": 0.7760966420173645,
+      "grad_norm": 0.5718142986297607,
       "learning_rate": 0.00016013544816838565,
-      "loss": 1.6709,
+      "loss": 1.4106,
       "step": 883
     },
     {
       "epoch": 0.8842486949454534,
-      "grad_norm": 0.6894650459289551,
+      "grad_norm": 0.551607072353363,
       "learning_rate": 0.00016005166215624827,
-      "loss": 1.6255,
+      "loss": 1.3474,
       "step": 884
     },
     {
       "epoch": 0.8852489762745772,
-      "grad_norm": 0.6777058839797974,
+      "grad_norm": 0.5464247465133667,
       "learning_rate": 0.0001599678101581945,
-      "loss": 1.7479,
+      "loss": 1.4443,
       "step": 885
     },
     {
       "epoch": 0.886249257603701,
-      "grad_norm": 0.7056024670600891,
+      "grad_norm": 0.5075456500053406,
       "learning_rate": 0.00015988389226636253,
-      "loss": 1.7896,
+      "loss": 1.4919,
       "step": 886
     },
     {
       "epoch": 0.8872495389328249,
-      "grad_norm": 0.6465604305267334,
+      "grad_norm": 0.48557186126708984,
       "learning_rate": 0.00015979990857296295,
-      "loss": 1.7363,
+      "loss": 1.4225,
       "step": 887
     },
     {
       "epoch": 0.8882498202619487,
-      "grad_norm": 0.6703017950057983,
+      "grad_norm": 0.5385611653327942,
       "learning_rate": 0.00015971585917027862,
-      "loss": 1.6617,
+      "loss": 1.2937,
       "step": 888
     },
     {
       "epoch": 0.8892501015910725,
-      "grad_norm": 0.7116142511367798,
+      "grad_norm": 0.6477749943733215,
       "learning_rate": 0.00015963174415066468,
-      "loss": 1.8232,
+      "loss": 1.5628,
       "step": 889
     },
     {
       "epoch": 0.8902503829201963,
-      "grad_norm": 0.7552229762077332,
+      "grad_norm": 0.6205973029136658,
       "learning_rate": 0.0001595475636065483,
-      "loss": 1.7847,
+      "loss": 1.4902,
       "step": 890
     },
     {
       "epoch": 0.8912506642493201,
-      "grad_norm": 0.70728999376297,
+      "grad_norm": 0.45717301964759827,
       "learning_rate": 0.00015946331763042867,
-      "loss": 1.5665,
+      "loss": 1.1998,
       "step": 891
     },
     {
       "epoch": 0.892250945578444,
-      "grad_norm": 0.6701356768608093,
+      "grad_norm": 0.5279855132102966,
       "learning_rate": 0.00015937900631487686,
-      "loss": 1.3572,
+      "loss": 1.0668,
       "step": 892
     },
     {
       "epoch": 0.8932512269075678,
-      "grad_norm": 0.6960388422012329,
+      "grad_norm": 0.5207269787788391,
       "learning_rate": 0.00015929462975253585,
-      "loss": 1.5815,
+      "loss": 1.2774,
       "step": 893
     },
     {
       "epoch": 0.8942515082366915,
-      "grad_norm": 0.6505674719810486,
+      "grad_norm": 0.5200834274291992,
       "learning_rate": 0.00015921018803612014,
-      "loss": 1.7499,
+      "loss": 1.4316,
       "step": 894
     },
     {
       "epoch": 0.8952517895658154,
-      "grad_norm": 0.604205310344696,
+      "grad_norm": 0.48317649960517883,
       "learning_rate": 0.0001591256812584159,
-      "loss": 1.6838,
+      "loss": 1.4101,
       "step": 895
     },
     {
       "epoch": 0.8962520708949392,
-      "grad_norm": 0.5875198841094971,
+      "grad_norm": 0.475483775138855,
       "learning_rate": 0.00015904110951228082,
-      "loss": 1.5147,
+      "loss": 1.2011,
       "step": 896
     },
     {
       "epoch": 0.897252352224063,
-      "grad_norm": 0.6970433592796326,
+      "grad_norm": 0.6542660593986511,
       "learning_rate": 0.00015895647289064396,
-      "loss": 1.7767,
+      "loss": 1.56,
       "step": 897
     },
     {
       "epoch": 0.8982526335531869,
-      "grad_norm": 0.7364515066146851,
+      "grad_norm": 0.5154829621315002,
       "learning_rate": 0.00015887177148650564,
-      "loss": 1.6672,
+      "loss": 1.3748,
       "step": 898
     },
     {
       "epoch": 0.8992529148823106,
-      "grad_norm": 0.7843589186668396,
+      "grad_norm": 0.5744799375534058,
       "learning_rate": 0.0001587870053929374,
-      "loss": 1.689,
+      "loss": 1.4072,
       "step": 899
     },
     {
       "epoch": 0.9002531962114345,
-      "grad_norm": 0.6405196189880371,
+      "grad_norm": 0.4835909307003021,
       "learning_rate": 0.00015870217470308188,
-      "loss": 1.5917,
+      "loss": 1.3037,
       "step": 900
     },
     {
       "epoch": 0.9012534775405583,
-      "grad_norm": 0.7019757628440857,
+      "grad_norm": 0.5292366743087769,
       "learning_rate": 0.0001586172795101526,
-      "loss": 1.5497,
+      "loss": 1.2395,
       "step": 901
     },
     {
       "epoch": 0.9022537588696821,
-      "grad_norm": 0.8048270344734192,
+      "grad_norm": 0.5905430912971497,
       "learning_rate": 0.00015853231990743406,
-      "loss": 1.5821,
+      "loss": 1.29,
       "step": 902
     },
     {
       "epoch": 0.903254040198806,
-      "grad_norm": 0.6245777606964111,
+      "grad_norm": 0.4918007254600525,
       "learning_rate": 0.0001584472959882815,
-      "loss": 1.5688,
+      "loss": 1.2593,
       "step": 903
     },
     {
       "epoch": 0.9042543215279297,
-      "grad_norm": 0.6584132313728333,
+      "grad_norm": 0.4735652208328247,
       "learning_rate": 0.00015836220784612085,
-      "loss": 1.4555,
+      "loss": 1.1669,
       "step": 904
     },
     {
       "epoch": 0.9052546028570535,
-      "grad_norm": 0.7710773944854736,
+      "grad_norm": 0.6272550821304321,
       "learning_rate": 0.00015827705557444852,
-      "loss": 1.6416,
+      "loss": 1.3692,
       "step": 905
     },
     {
       "epoch": 0.9062548841861774,
-      "grad_norm": 0.6738126277923584,
+      "grad_norm": 0.5333564877510071,
       "learning_rate": 0.00015819183926683153,
-      "loss": 1.6272,
+      "loss": 1.3672,
       "step": 906
     },
     {
       "epoch": 0.9072551655153012,
-      "grad_norm": 0.6698735356330872,
+      "grad_norm": 0.44029948115348816,
       "learning_rate": 0.00015810655901690715,
-      "loss": 1.4778,
+      "loss": 1.2124,
       "step": 907
     },
     {
       "epoch": 0.9082554468444249,
-      "grad_norm": 1.0088928937911987,
+      "grad_norm": 0.5636379718780518,
       "learning_rate": 0.00015802121491838297,
-      "loss": 1.6854,
+      "loss": 1.3507,
       "step": 908
     },
     {
       "epoch": 0.9092557281735488,
-      "grad_norm": 0.6948708891868591,
+      "grad_norm": 0.4394778907299042,
       "learning_rate": 0.0001579358070650367,
-      "loss": 1.5673,
+      "loss": 1.3159,
       "step": 909
     },
     {
       "epoch": 0.9102560095026726,
-      "grad_norm": 0.6728948950767517,
+      "grad_norm": 0.5382723212242126,
       "learning_rate": 0.00015785033555071616,
-      "loss": 1.6646,
+      "loss": 1.3733,
       "step": 910
     },
     {
       "epoch": 0.9112562908317965,
-      "grad_norm": 0.8096952438354492,
+      "grad_norm": 0.5251659750938416,
       "learning_rate": 0.00015776480046933905,
-      "loss": 1.4675,
+      "loss": 1.2253,
       "step": 911
     },
     {
       "epoch": 0.9122565721609203,
-      "grad_norm": 0.6625403761863708,
+      "grad_norm": 0.4791383743286133,
       "learning_rate": 0.000157679201914893,
-      "loss": 1.4793,
+      "loss": 1.2341,
       "step": 912
     },
     {
       "epoch": 0.913256853490044,
-      "grad_norm": 0.7129424810409546,
+      "grad_norm": 0.5058613419532776,
       "learning_rate": 0.00015759353998143528,
-      "loss": 1.574,
+      "loss": 1.2717,
       "step": 913
     },
     {
       "epoch": 0.9142571348191679,
-      "grad_norm": 0.6151349544525146,
+      "grad_norm": 0.46837320923805237,
       "learning_rate": 0.00015750781476309288,
-      "loss": 1.5631,
+      "loss": 1.2484,
       "step": 914
     },
     {
       "epoch": 0.9152574161482917,
-      "grad_norm": 0.7185074687004089,
+      "grad_norm": 0.524444580078125,
       "learning_rate": 0.00015742202635406235,
-      "loss": 1.8382,
+      "loss": 1.5512,
       "step": 915
     },
     {
       "epoch": 0.9162576974774155,
-      "grad_norm": 0.7076066732406616,
+      "grad_norm": 0.6169744729995728,
       "learning_rate": 0.00015733617484860963,
-      "loss": 1.5394,
+      "loss": 1.271,
       "step": 916
     },
     {
       "epoch": 0.9172579788065394,
-      "grad_norm": 0.7286276817321777,
+      "grad_norm": 0.48883670568466187,
       "learning_rate": 0.00015725026034106996,
-      "loss": 1.8139,
+      "loss": 1.4779,
       "step": 917
     },
     {
       "epoch": 0.9182582601356631,
-      "grad_norm": 0.757075846195221,
+      "grad_norm": 0.5408645272254944,
       "learning_rate": 0.00015716428292584787,
-      "loss": 1.6768,
+      "loss": 1.3574,
       "step": 918
     },
     {
       "epoch": 0.919258541464787,
-      "grad_norm": 0.6926739811897278,
+      "grad_norm": 0.5622221231460571,
       "learning_rate": 0.00015707824269741702,
-      "loss": 1.4541,
+      "loss": 1.2146,
       "step": 919
     },
     {
       "epoch": 0.9202588227939108,
-      "grad_norm": 0.6489847898483276,
+      "grad_norm": 0.477328896522522,
       "learning_rate": 0.00015699213975031996,
-      "loss": 1.4725,
+      "loss": 1.162,
       "step": 920
     },
     {
       "epoch": 0.9212591041230346,
-      "grad_norm": 0.7668707966804504,
+      "grad_norm": 0.503027081489563,
       "learning_rate": 0.0001569059741791684,
-      "loss": 1.4239,
+      "loss": 1.1674,
       "step": 921
     },
     {
       "epoch": 0.9222593854521585,
-      "grad_norm": 0.736863911151886,
+      "grad_norm": 0.5951637625694275,
       "learning_rate": 0.0001568197460786426,
-      "loss": 1.6117,
+      "loss": 1.3737,
       "step": 922
     },
     {
       "epoch": 0.9232596667812822,
-      "grad_norm": 0.8462884426116943,
+      "grad_norm": 0.5276626348495483,
       "learning_rate": 0.0001567334555434917,
-      "loss": 1.5025,
+      "loss": 1.2494,
       "step": 923
     },
     {
       "epoch": 0.924259948110406,
-      "grad_norm": 0.7481950521469116,
+      "grad_norm": 0.6354761123657227,
       "learning_rate": 0.0001566471026685334,
-      "loss": 1.5024,
+      "loss": 1.2052,
       "step": 924
     },
     {
       "epoch": 0.9252602294395299,
-      "grad_norm": 0.6457516551017761,
+      "grad_norm": 0.4227287471294403,
       "learning_rate": 0.00015656068754865387,
-      "loss": 1.4526,
+      "loss": 1.1446,
       "step": 925
     },
     {
       "epoch": 0.9262605107686537,
-      "grad_norm": 0.809140682220459,
+      "grad_norm": 0.5290839076042175,
       "learning_rate": 0.00015647421027880772,
-      "loss": 1.4449,
+      "loss": 1.2057,
       "step": 926
     },
     {
       "epoch": 0.9272607920977775,
-      "grad_norm": 0.6967790126800537,
+      "grad_norm": 0.4961225986480713,
       "learning_rate": 0.0001563876709540178,
-      "loss": 1.5552,
+      "loss": 1.2788,
       "step": 927
     },
     {
       "epoch": 0.9282610734269013,
-      "grad_norm": 0.6858595609664917,
+      "grad_norm": 0.5095213651657104,
       "learning_rate": 0.0001563010696693752,
-      "loss": 1.6202,
+      "loss": 1.2751,
       "step": 928
     },
     {
       "epoch": 0.9292613547560251,
-      "grad_norm": 0.7033559679985046,
+      "grad_norm": 0.5027223825454712,
       "learning_rate": 0.00015621440652003907,
-      "loss": 1.7186,
+      "loss": 1.3653,
       "step": 929
     },
     {
       "epoch": 0.930261636085149,
-      "grad_norm": 0.6527283787727356,
+      "grad_norm": 0.49251896142959595,
       "learning_rate": 0.00015612768160123652,
-      "loss": 1.5028,
+      "loss": 1.1556,
       "step": 930
     },
     {
       "epoch": 0.9312619174142728,
-      "grad_norm": 0.7243602275848389,
+      "grad_norm": 0.5187139511108398,
       "learning_rate": 0.00015604089500826257,
-      "loss": 1.6729,
+      "loss": 1.3623,
       "step": 931
     },
     {
       "epoch": 0.9322621987433966,
-      "grad_norm": 0.6734297275543213,
+      "grad_norm": 0.5004428029060364,
       "learning_rate": 0.00015595404683648,
-      "loss": 1.4731,
+      "loss": 1.185,
       "step": 932
     },
     {
       "epoch": 0.9332624800725204,
-      "grad_norm": 0.7641247510910034,
+      "grad_norm": 0.5750531554222107,
       "learning_rate": 0.00015586713718131922,
-      "loss": 1.5851,
+      "loss": 1.2999,
       "step": 933
     },
     {
       "epoch": 0.9342627614016442,
-      "grad_norm": 0.7062788009643555,
+      "grad_norm": 0.482732355594635,
       "learning_rate": 0.0001557801661382782,
-      "loss": 1.5735,
+      "loss": 1.2635,
       "step": 934
     },
     {
       "epoch": 0.935263042730768,
-      "grad_norm": 0.6413556337356567,
+      "grad_norm": 0.47854143381118774,
       "learning_rate": 0.00015569313380292248,
-      "loss": 1.5854,
+      "loss": 1.2833,
       "step": 935
     },
     {
       "epoch": 0.9362633240598919,
-      "grad_norm": 0.645720362663269,
+      "grad_norm": 0.49382665753364563,
       "learning_rate": 0.00015560604027088477,
-      "loss": 1.5072,
+      "loss": 1.2327,
       "step": 936
     },
     {
       "epoch": 0.9372636053890157,
-      "grad_norm": 0.6726225018501282,
+      "grad_norm": 0.5009885430335999,
       "learning_rate": 0.00015551888563786515,
-      "loss": 1.587,
+      "loss": 1.2967,
       "step": 937
     },
     {
       "epoch": 0.9382638867181394,
-      "grad_norm": 0.7043680548667908,
+      "grad_norm": 0.5012707114219666,
       "learning_rate": 0.00015543166999963076,
-      "loss": 1.6577,
+      "loss": 1.3231,
       "step": 938
     },
     {
       "epoch": 0.9392641680472633,
-      "grad_norm": 0.7049617767333984,
+      "grad_norm": 0.6908506751060486,
       "learning_rate": 0.0001553443934520159,
-      "loss": 1.7624,
+      "loss": 1.4055,
       "step": 939
     },
     {
       "epoch": 0.9402644493763871,
-      "grad_norm": 0.7060776352882385,
+      "grad_norm": 0.7104817032814026,
       "learning_rate": 0.00015525705609092157,
-      "loss": 1.6208,
+      "loss": 1.3435,
       "step": 940
     },
     {
       "epoch": 0.941264730705511,
-      "grad_norm": 0.6215025186538696,
+      "grad_norm": 0.49263522028923035,
       "learning_rate": 0.00015516965801231586,
-      "loss": 1.4645,
+      "loss": 1.2259,
       "step": 941
     },
     {
       "epoch": 0.9422650120346348,
-      "grad_norm": 0.7021099328994751,
+      "grad_norm": 0.5337693691253662,
       "learning_rate": 0.0001550821993122334,
-      "loss": 1.566,
+      "loss": 1.2863,
       "step": 942
     },
     {
       "epoch": 0.9432652933637585,
-      "grad_norm": 0.6451042294502258,
+      "grad_norm": 0.5506749153137207,
       "learning_rate": 0.0001549946800867755,
-      "loss": 1.7491,
+      "loss": 1.4061,
       "step": 943
     },
     {
       "epoch": 0.9442655746928824,
-      "grad_norm": 0.7288572192192078,
+      "grad_norm": 0.5121364593505859,
       "learning_rate": 0.00015490710043210997,
-      "loss": 1.6302,
+      "loss": 1.3567,
       "step": 944
     },
     {
       "epoch": 0.9452658560220062,
-      "grad_norm": 0.7850833535194397,
+      "grad_norm": 0.5326678156852722,
       "learning_rate": 0.00015481946044447099,
-      "loss": 1.5673,
+      "loss": 1.2719,
       "step": 945
     },
     {
       "epoch": 0.94626613735113,
-      "grad_norm": 0.7459181547164917,
+      "grad_norm": 0.6023722290992737,
       "learning_rate": 0.00015473176022015906,
-      "loss": 1.4529,
+      "loss": 1.1512,
       "step": 946
     },
     {
       "epoch": 0.9472664186802539,
-      "grad_norm": 0.7002627849578857,
+      "grad_norm": 0.4953387975692749,
       "learning_rate": 0.0001546439998555409,
-      "loss": 1.8814,
+      "loss": 1.556,
       "step": 947
     },
     {
       "epoch": 0.9482667000093776,
-      "grad_norm": 0.6664572358131409,
+      "grad_norm": 0.5187799334526062,
       "learning_rate": 0.0001545561794470492,
-      "loss": 1.5337,
+      "loss": 1.279,
       "step": 948
     },
     {
       "epoch": 0.9492669813385014,
-      "grad_norm": 0.757116973400116,
+      "grad_norm": 0.5788894295692444,
       "learning_rate": 0.00015446829909118275,
-      "loss": 1.5775,
+      "loss": 1.3246,
       "step": 949
     },
     {
       "epoch": 0.9502672626676253,
-      "grad_norm": 0.7456643581390381,
+      "grad_norm": 0.5551681518554688,
       "learning_rate": 0.00015438035888450623,
-      "loss": 1.525,
+      "loss": 1.2231,
       "step": 950
     },
     {
       "epoch": 0.9512675439967491,
-      "grad_norm": 0.6722500920295715,
+      "grad_norm": 0.4898390471935272,
       "learning_rate": 0.00015429235892364994,
-      "loss": 1.5059,
+      "loss": 1.2036,
       "step": 951
     },
     {
       "epoch": 0.952267825325873,
-      "grad_norm": 0.7431210279464722,
+      "grad_norm": 0.5427507162094116,
       "learning_rate": 0.00015420429930530996,
-      "loss": 1.6867,
+      "loss": 1.3614,
       "step": 952
     },
     {
       "epoch": 0.9532681066549967,
-      "grad_norm": 0.751015305519104,
+      "grad_norm": 0.557054340839386,
       "learning_rate": 0.00015411618012624786,
-      "loss": 1.7371,
+      "loss": 1.4249,
       "step": 953
     },
     {
       "epoch": 0.9542683879841205,
-      "grad_norm": 0.807579517364502,
+      "grad_norm": 0.5793543457984924,
       "learning_rate": 0.00015402800148329071,
-      "loss": 1.7353,
+      "loss": 1.4341,
       "step": 954
     },
     {
       "epoch": 0.9552686693132444,
-      "grad_norm": 0.608161449432373,
+      "grad_norm": 0.5993456244468689,
       "learning_rate": 0.00015393976347333088,
-      "loss": 1.3074,
+      "loss": 1.0259,
       "step": 955
     },
     {
       "epoch": 0.9562689506423682,
-      "grad_norm": 0.7092815637588501,
+      "grad_norm": 0.554904580116272,
       "learning_rate": 0.00015385146619332596,
-      "loss": 1.676,
+      "loss": 1.3558,
       "step": 956
     },
     {
       "epoch": 0.9572692319714919,
-      "grad_norm": 0.7639429569244385,
+      "grad_norm": 0.5488478541374207,
       "learning_rate": 0.00015376310974029873,
-      "loss": 1.6452,
+      "loss": 1.358,
       "step": 957
     },
     {
       "epoch": 0.9582695133006158,
-      "grad_norm": 0.7333659529685974,
+      "grad_norm": 0.5108879208564758,
       "learning_rate": 0.00015367469421133695,
-      "loss": 1.6821,
+      "loss": 1.3865,
       "step": 958
     },
     {
       "epoch": 0.9592697946297396,
-      "grad_norm": 0.7246838212013245,
+      "grad_norm": 0.4606814682483673,
       "learning_rate": 0.00015358621970359325,
-      "loss": 1.5078,
+      "loss": 1.2055,
       "step": 959
     },
     {
       "epoch": 0.9602700759588634,
-      "grad_norm": 0.7209622859954834,
+      "grad_norm": 0.4974004328250885,
       "learning_rate": 0.00015349768631428519,
-      "loss": 1.5617,
+      "loss": 1.2541,
       "step": 960
     },
     {
       "epoch": 0.9612703572879873,
-      "grad_norm": 0.7034916877746582,
+      "grad_norm": 0.5107241272926331,
       "learning_rate": 0.00015340909414069488,
-      "loss": 1.4711,
+      "loss": 1.1624,
       "step": 961
     },
     {
       "epoch": 0.962270638617111,
-      "grad_norm": 0.7311360239982605,
+      "grad_norm": 0.5587212443351746,
       "learning_rate": 0.00015332044328016914,
-      "loss": 1.6488,
+      "loss": 1.349,
       "step": 962
     },
     {
       "epoch": 0.9632709199462349,
-      "grad_norm": 0.6668992638587952,
+      "grad_norm": 0.5209497809410095,
       "learning_rate": 0.0001532317338301192,
-      "loss": 1.6804,
+      "loss": 1.3695,
       "step": 963
     },
     {
       "epoch": 0.9642712012753587,
-      "grad_norm": 0.6265329122543335,
+      "grad_norm": 0.4985620677471161,
       "learning_rate": 0.00015314296588802076,
-      "loss": 1.8169,
+      "loss": 1.4597,
       "step": 964
     },
     {
       "epoch": 0.9652714826044825,
-      "grad_norm": 0.6945448517799377,
+      "grad_norm": 0.5065100789070129,
       "learning_rate": 0.00015305413955141365,
-      "loss": 1.8041,
+      "loss": 1.4225,
       "step": 965
     },
     {
       "epoch": 0.9662717639336064,
-      "grad_norm": 0.6718643307685852,
+      "grad_norm": 0.5079792737960815,
       "learning_rate": 0.00015296525491790205,
-      "loss": 1.3486,
+      "loss": 1.057,
       "step": 966
     },
     {
       "epoch": 0.9672720452627301,
-      "grad_norm": 0.6232700943946838,
+      "grad_norm": 0.4673600196838379,
       "learning_rate": 0.00015287631208515406,
-      "loss": 1.5672,
+      "loss": 1.2531,
       "step": 967
     },
     {
       "epoch": 0.9682723265918539,
-      "grad_norm": 0.7481172680854797,
+      "grad_norm": 0.5309945344924927,
       "learning_rate": 0.00015278731115090171,
-      "loss": 1.5992,
+      "loss": 1.374,
       "step": 968
     },
     {
       "epoch": 0.9692726079209778,
-      "grad_norm": 0.6585466861724854,
+      "grad_norm": 0.4792092442512512,
       "learning_rate": 0.00015269825221294098,
-      "loss": 1.6403,
+      "loss": 1.3018,
       "step": 969
     },
     {
       "epoch": 0.9702728892501016,
-      "grad_norm": 0.7587956786155701,
+      "grad_norm": 0.5222868323326111,
       "learning_rate": 0.00015260913536913154,
-      "loss": 1.7991,
+      "loss": 1.4063,
       "step": 970
     },
     {
       "epoch": 0.9712731705792254,
-      "grad_norm": 0.672698974609375,
+      "grad_norm": 0.5373417139053345,
       "learning_rate": 0.00015251996071739664,
-      "loss": 1.4311,
+      "loss": 1.2183,
       "step": 971
     },
     {
       "epoch": 0.9722734519083492,
-      "grad_norm": 0.7597199082374573,
+      "grad_norm": 0.5624721050262451,
       "learning_rate": 0.00015243072835572318,
-      "loss": 1.5692,
+      "loss": 1.2696,
       "step": 972
     },
     {
       "epoch": 0.973273733237473,
-      "grad_norm": 0.7342745661735535,
+      "grad_norm": 0.46938082575798035,
       "learning_rate": 0.0001523414383821613,
-      "loss": 1.6364,
+      "loss": 1.3544,
       "step": 973
     },
     {
       "epoch": 0.9742740145665969,
-      "grad_norm": 0.6640815138816833,
+      "grad_norm": 0.45348694920539856,
       "learning_rate": 0.00015225209089482462,
-      "loss": 1.5113,
+      "loss": 1.2078,
       "step": 974
     },
     {
       "epoch": 0.9752742958957207,
-      "grad_norm": 0.6298378109931946,
+      "grad_norm": 0.48000606894493103,
       "learning_rate": 0.0001521626859918898,
-      "loss": 1.4822,
+      "loss": 1.1914,
       "step": 975
     },
     {
       "epoch": 0.9762745772248445,
-      "grad_norm": 0.6862055659294128,
+      "grad_norm": 0.5106796622276306,
       "learning_rate": 0.00015207322377159668,
-      "loss": 1.6159,
+      "loss": 1.3249,
       "step": 976
     },
     {
       "epoch": 0.9772748585539683,
-      "grad_norm": 0.6377236843109131,
+      "grad_norm": 0.49865373969078064,
       "learning_rate": 0.00015198370433224805,
-      "loss": 1.6046,
+      "loss": 1.2876,
       "step": 977
     },
     {
       "epoch": 0.9782751398830921,
-      "grad_norm": 0.620070219039917,
+      "grad_norm": 0.5271755456924438,
       "learning_rate": 0.00015189412777220958,
-      "loss": 1.589,
+      "loss": 1.3049,
       "step": 978
     },
     {
       "epoch": 0.9792754212122159,
-      "grad_norm": 0.7776119112968445,
+      "grad_norm": 0.49824708700180054,
       "learning_rate": 0.00015180449418990976,
-      "loss": 1.485,
+      "loss": 1.1614,
       "step": 979
     },
     {
       "epoch": 0.9802757025413398,
-      "grad_norm": 0.8258413076400757,
+      "grad_norm": 0.7327549457550049,
       "learning_rate": 0.00015171480368383964,
-      "loss": 1.5615,
+      "loss": 1.2923,
       "step": 980
     },
     {
       "epoch": 0.9812759838704636,
-      "grad_norm": 0.7297958135604858,
+      "grad_norm": 0.5170425176620483,
       "learning_rate": 0.00015162505635255287,
-      "loss": 1.5408,
+      "loss": 1.3097,
       "step": 981
     },
     {
       "epoch": 0.9822762651995874,
-      "grad_norm": 0.5848103165626526,
+      "grad_norm": 0.47041526436805725,
       "learning_rate": 0.00015153525229466555,
-      "loss": 1.6821,
+      "loss": 1.3508,
       "step": 982
     },
     {
       "epoch": 0.9832765465287112,
-      "grad_norm": 0.7375655174255371,
+      "grad_norm": 0.4670693278312683,
       "learning_rate": 0.00015144539160885613,
-      "loss": 1.7568,
+      "loss": 1.3974,
       "step": 983
     },
     {
       "epoch": 0.984276827857835,
-      "grad_norm": 0.7466885447502136,
+      "grad_norm": 0.5745754837989807,
       "learning_rate": 0.00015135547439386516,
-      "loss": 1.5805,
+      "loss": 1.2977,
       "step": 984
     },
     {
       "epoch": 0.9852771091869589,
-      "grad_norm": 0.6645593047142029,
+      "grad_norm": 0.5845474004745483,
       "learning_rate": 0.0001512655007484955,
-      "loss": 1.6776,
+      "loss": 1.3384,
       "step": 985
     },
     {
       "epoch": 0.9862773905160827,
-      "grad_norm": 0.7973874807357788,
+      "grad_norm": 0.5627439618110657,
       "learning_rate": 0.00015117547077161185,
-      "loss": 1.4931,
+      "loss": 1.1756,
       "step": 986
     },
     {
       "epoch": 0.9872776718452064,
-      "grad_norm": 0.685391902923584,
+      "grad_norm": 0.6411226987838745,
       "learning_rate": 0.0001510853845621409,
-      "loss": 1.6254,
+      "loss": 1.3441,
       "step": 987
     },
     {
       "epoch": 0.9882779531743303,
-      "grad_norm": 0.6562414765357971,
+      "grad_norm": 0.545659601688385,
       "learning_rate": 0.00015099524221907107,
-      "loss": 1.6677,
+      "loss": 1.3766,
       "step": 988
     },
     {
       "epoch": 0.9892782345034541,
-      "grad_norm": 0.6216359734535217,
+      "grad_norm": 0.5058498382568359,
       "learning_rate": 0.0001509050438414525,
-      "loss": 1.6107,
+      "loss": 1.3171,
       "step": 989
     },
     {
       "epoch": 0.9902785158325779,
-      "grad_norm": 0.7108810544013977,
+      "grad_norm": 0.6247567534446716,
       "learning_rate": 0.00015081478952839693,
-      "loss": 1.5268,
+      "loss": 1.2141,
       "step": 990
     },
     {
       "epoch": 0.9912787971617018,
-      "grad_norm": 0.7076026797294617,
+      "grad_norm": 0.5492308139801025,
       "learning_rate": 0.00015072447937907753,
-      "loss": 1.3716,
+      "loss": 1.1626,
       "step": 991
     },
     {
       "epoch": 0.9922790784908255,
-      "grad_norm": 0.6056272983551025,
+      "grad_norm": 0.4795534908771515,
       "learning_rate": 0.00015063411349272877,
-      "loss": 1.4931,
+      "loss": 1.218,
       "step": 992
     },
     {
       "epoch": 0.9932793598199494,
-      "grad_norm": 0.726671576499939,
+      "grad_norm": 0.5527793169021606,
       "learning_rate": 0.00015054369196864644,
-      "loss": 1.6409,
+      "loss": 1.3816,
       "step": 993
     },
     {
       "epoch": 0.9942796411490732,
-      "grad_norm": 0.7019214630126953,
+      "grad_norm": 0.5297475457191467,
       "learning_rate": 0.00015045321490618748,
-      "loss": 1.4476,
+      "loss": 1.2515,
       "step": 994
     },
     {
       "epoch": 0.995279922478197,
-      "grad_norm": 0.755043625831604,
+      "grad_norm": 0.518803596496582,
       "learning_rate": 0.00015036268240476978,
-      "loss": 1.6674,
+      "loss": 1.3631,
       "step": 995
     },
     {
       "epoch": 0.9962802038073209,
-      "grad_norm": 0.7450313568115234,
+      "grad_norm": 0.47196391224861145,
       "learning_rate": 0.00015027209456387218,
-      "loss": 1.3706,
+      "loss": 1.0932,
       "step": 996
     },
     {
       "epoch": 0.9972804851364446,
-      "grad_norm": 0.6804680228233337,
+      "grad_norm": 0.5369086861610413,
       "learning_rate": 0.00015018145148303438,
-      "loss": 1.3878,
+      "loss": 1.1181,
       "step": 997
     },
     {
       "epoch": 0.9982807664655684,
-      "grad_norm": 0.7353954315185547,
+      "grad_norm": 0.5940788388252258,
       "learning_rate": 0.00015009075326185667,
-      "loss": 1.8656,
+      "loss": 1.561,
       "step": 998
     },
     {
       "epoch": 0.9992810477946923,
-      "grad_norm": 0.7213340401649475,
+      "grad_norm": 0.5340734124183655,
       "learning_rate": 0.00015000000000000001,
-      "loss": 1.6031,
+      "loss": 1.2909,
       "step": 999
     },
     {
       "epoch": 1.0002813291238162,
-      "grad_norm": 0.7066403031349182,
+      "grad_norm": 0.5133704543113708,
       "learning_rate": 0.00014990919179718584,
-      "loss": 1.3663,
+      "loss": 1.0441,
       "step": 1000
     }
   ],
@@ -7026,7 +7026,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 8104306689589248.0,
+  "total_flos": 6.479774965540454e+16,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null