diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -10,11310 +10,11318 @@
   "log_history": [
     {
       "epoch": 0.0001239387742455227,
-      "grad_norm": 0.6156698132479959,
+      "grad_norm": 0.36398436718107713,
       "learning_rate": 2.4783147459727386e-07,
-      "loss": 4.9844,
+      "loss": 2.5391,
       "step": 1
     },
     {
       "epoch": 0.0006196938712276135,
-      "grad_norm": 0.6052976200579151,
+      "grad_norm": 0.3553971091320693,
       "learning_rate": 1.2391573729863693e-06,
-      "loss": 4.7617,
+      "loss": 2.3916,
       "step": 5
     },
     {
       "epoch": 0.001239387742455227,
-      "grad_norm": 0.7601239478563812,
+      "grad_norm": 0.33706223881954883,
       "learning_rate": 2.4783147459727386e-06,
-      "loss": 4.8406,
+      "loss": 2.4906,
       "step": 10
     },
     {
       "epoch": 0.0018590816136828407,
-      "grad_norm": 0.7526341301126038,
+      "grad_norm": 0.38128850820849297,
       "learning_rate": 3.717472118959108e-06,
-      "loss": 4.8391,
+      "loss": 2.4898,
       "step": 15
     },
     {
       "epoch": 0.002478775484910454,
-      "grad_norm": 0.6322948184451165,
+      "grad_norm": 0.43552764540168315,
       "learning_rate": 4.956629491945477e-06,
-      "loss": 4.9266,
+      "loss": 2.4367,
       "step": 20
     },
     {
       "epoch": 0.003098469356138068,
-      "grad_norm": 0.6990885844721237,
+      "grad_norm": 0.30493677951675263,
       "learning_rate": 6.195786864931847e-06,
-      "loss": 4.8547,
+      "loss": 2.4398,
       "step": 25
     },
     {
       "epoch": 0.0037181632273656814,
-      "grad_norm": 0.7984745761097714,
+      "grad_norm": 0.40892724849804757,
       "learning_rate": 7.434944237918216e-06,
-      "loss": 4.8336,
+      "loss": 2.3828,
       "step": 30
     },
     {
       "epoch": 0.004337857098593295,
-      "grad_norm": 0.7814803522780867,
+      "grad_norm": 0.41597118556401486,
       "learning_rate": 8.674101610904585e-06,
-      "loss": 4.8438,
+      "loss": 2.459,
       "step": 35
     },
     {
       "epoch": 0.004957550969820908,
-      "grad_norm": 0.7825601910016009,
+      "grad_norm": 0.34011631695052974,
       "learning_rate": 9.913258983890955e-06,
-      "loss": 4.8141,
+      "loss": 2.3883,
       "step": 40
     },
     {
       "epoch": 0.005577244841048522,
-      "grad_norm": 0.7686051064734422,
+      "grad_norm": 0.42036838195034787,
       "learning_rate": 1.1152416356877324e-05,
-      "loss": 4.8469,
+      "loss": 2.407,
       "step": 45
     },
     {
       "epoch": 0.006196938712276136,
-      "grad_norm": 0.7149556352785319,
+      "grad_norm": 0.32938017778181394,
       "learning_rate": 1.2391573729863694e-05,
-      "loss": 4.8094,
+      "loss": 2.4359,
       "step": 50
     },
     {
       "epoch": 0.006816632583503749,
-      "grad_norm": 0.7305536513174533,
+      "grad_norm": 0.522486908526655,
       "learning_rate": 1.3630731102850064e-05,
-      "loss": 4.7266,
+      "loss": 2.3625,
       "step": 55
     },
     {
       "epoch": 0.007436326454731363,
-      "grad_norm": 0.8178979347533347,
+      "grad_norm": 0.39947004413100906,
       "learning_rate": 1.4869888475836432e-05,
-      "loss": 4.7031,
+      "loss": 2.4008,
       "step": 60
     },
     {
       "epoch": 0.008056020325958976,
-      "grad_norm": 0.8334181887763964,
+      "grad_norm": 0.4627809567954664,
       "learning_rate": 1.61090458488228e-05,
-      "loss": 4.6656,
+      "loss": 2.3168,
       "step": 65
     },
     {
       "epoch": 0.00867571419718659,
-      "grad_norm": 0.7695228033324475,
+      "grad_norm": 0.4876882111255118,
       "learning_rate": 1.734820322180917e-05,
-      "loss": 4.7344,
+      "loss": 2.343,
       "step": 70
     },
     {
       "epoch": 0.009295408068414203,
-      "grad_norm": 0.8811868823646412,
+      "grad_norm": 0.3811791732250246,
       "learning_rate": 1.858736059479554e-05,
-      "loss": 4.6906,
+      "loss": 2.3398,
       "step": 75
     },
     {
       "epoch": 0.009915101939641817,
-      "grad_norm": 1.1199523468579367,
+      "grad_norm": 0.4376425547982147,
       "learning_rate": 1.982651796778191e-05,
-      "loss": 4.5859,
+      "loss": 2.3344,
       "step": 80
     },
     {
       "epoch": 0.01053479581086943,
-      "grad_norm": 1.0597694158945299,
+      "grad_norm": 0.3931646393426777,
       "learning_rate": 2.106567534076828e-05,
-      "loss": 4.4813,
+      "loss": 2.332,
       "step": 85
     },
     {
       "epoch": 0.011154489682097044,
-      "grad_norm": 0.8117839754264472,
+      "grad_norm": 0.4461816921220648,
       "learning_rate": 2.230483271375465e-05,
-      "loss": 4.5922,
+      "loss": 2.2609,
       "step": 90
     },
     {
       "epoch": 0.011774183553324657,
-      "grad_norm": 0.8958360055388476,
+      "grad_norm": 0.37584050496240057,
       "learning_rate": 2.3543990086741015e-05,
-      "loss": 4.4766,
+      "loss": 2.2234,
       "step": 95
     },
     {
       "epoch": 0.012393877424552271,
-      "grad_norm": 0.8817057210428043,
+      "grad_norm": 0.33119569793020404,
       "learning_rate": 2.4783147459727388e-05,
-      "loss": 4.5094,
+      "loss": 2.173,
       "step": 100
     },
     {
       "epoch": 0.013013571295779885,
-      "grad_norm": 0.537206137804486,
+      "grad_norm": 0.3705649194998293,
       "learning_rate": 2.6022304832713758e-05,
-      "loss": 4.3797,
+      "loss": 2.0984,
       "step": 105
     },
     {
       "epoch": 0.013633265167007498,
-      "grad_norm": 0.5300195861022883,
+      "grad_norm": 0.2504511318591155,
       "learning_rate": 2.7261462205700128e-05,
-      "loss": 4.3156,
+      "loss": 2.109,
       "step": 110
     },
     {
       "epoch": 0.014252959038235112,
-      "grad_norm": 0.3948064250822983,
+      "grad_norm": 0.23798871106040215,
       "learning_rate": 2.8500619578686494e-05,
-      "loss": 4.257,
+      "loss": 2.1016,
       "step": 115
     },
     {
       "epoch": 0.014872652909462726,
-      "grad_norm": 0.364515312322803,
+      "grad_norm": 0.21122743560448873,
       "learning_rate": 2.9739776951672864e-05,
-      "loss": 4.2406,
+      "loss": 2.1574,
       "step": 120
     },
     {
       "epoch": 0.01549234678069034,
-      "grad_norm": 0.37164041744407617,
+      "grad_norm": 0.1813741930176235,
       "learning_rate": 3.0978934324659233e-05,
-      "loss": 4.2383,
+      "loss": 2.1199,
       "step": 125
     },
     {
       "epoch": 0.01611204065191795,
-      "grad_norm": 0.3245015145961911,
+      "grad_norm": 0.1990606894765091,
       "learning_rate": 3.22180916976456e-05,
-      "loss": 4.1914,
+      "loss": 2.1242,
       "step": 130
     },
     {
       "epoch": 0.016731734523145567,
-      "grad_norm": 0.30522684760315766,
+      "grad_norm": 0.15350817933593797,
       "learning_rate": 3.345724907063197e-05,
-      "loss": 4.0875,
+      "loss": 2.1379,
       "step": 135
     },
     {
       "epoch": 0.01735142839437318,
-      "grad_norm": 0.35317894181130866,
+      "grad_norm": 0.15326065617394624,
       "learning_rate": 3.469640644361834e-05,
-      "loss": 4.1297,
+      "loss": 2.1441,
       "step": 140
     },
     {
       "epoch": 0.017971122265600794,
-      "grad_norm": 0.34626796389259323,
+      "grad_norm": 0.1630492804177675,
       "learning_rate": 3.593556381660471e-05,
-      "loss": 4.1219,
+      "loss": 2.0914,
       "step": 145
     },
     {
       "epoch": 0.018590816136828406,
-      "grad_norm": 0.32611609298532446,
+      "grad_norm": 0.17107386161983282,
       "learning_rate": 3.717472118959108e-05,
-      "loss": 4.0687,
+      "loss": 2.0324,
       "step": 150
     },
     {
       "epoch": 0.01921051000805602,
-      "grad_norm": 0.2826877224295346,
+      "grad_norm": 0.16054446888841606,
       "learning_rate": 3.841387856257745e-05,
-      "loss": 4.2211,
+      "loss": 2.1078,
       "step": 155
     },
     {
       "epoch": 0.019830203879283633,
-      "grad_norm": 0.2887063005587996,
+      "grad_norm": 0.14956110781178233,
       "learning_rate": 3.965303593556382e-05,
-      "loss": 4.0383,
+      "loss": 2.0621,
       "step": 160
     },
     {
       "epoch": 0.02044989775051125,
-      "grad_norm": 0.2914285373125258,
+      "grad_norm": 0.13835441017934677,
       "learning_rate": 4.0892193308550185e-05,
-      "loss": 4.107,
+      "loss": 2.059,
       "step": 165
     },
     {
       "epoch": 0.02106959162173886,
-      "grad_norm": 0.2565212414056402,
+      "grad_norm": 0.12666794482225688,
       "learning_rate": 4.213135068153656e-05,
-      "loss": 4.1594,
+      "loss": 2.0848,
       "step": 170
     },
     {
       "epoch": 0.021689285492966476,
-      "grad_norm": 0.2440231811546915,
+      "grad_norm": 0.12997209523884481,
       "learning_rate": 4.337050805452293e-05,
-      "loss": 4.093,
+      "loss": 2.0883,
       "step": 175
     },
     {
       "epoch": 0.022308979364194088,
-      "grad_norm": 0.23545457064479375,
+      "grad_norm": 0.13662868551878915,
       "learning_rate": 4.46096654275093e-05,
-      "loss": 4.1391,
+      "loss": 2.0496,
       "step": 180
     },
     {
       "epoch": 0.022928673235421703,
-      "grad_norm": 0.24355219689682756,
+      "grad_norm": 0.12904027963907533,
       "learning_rate": 4.5848822800495664e-05,
-      "loss": 4.1117,
+      "loss": 2.0391,
       "step": 185
     },
     {
       "epoch": 0.023548367106649315,
-      "grad_norm": 0.25029203671068706,
+      "grad_norm": 0.11673326298638313,
       "learning_rate": 4.708798017348203e-05,
-      "loss": 4.143,
+      "loss": 1.9664,
       "step": 190
     },
     {
       "epoch": 0.02416806097787693,
-      "grad_norm": 0.23034997869099355,
+      "grad_norm": 0.11261296057950389,
       "learning_rate": 4.83271375464684e-05,
-      "loss": 4.0078,
+      "loss": 2.023,
       "step": 195
     },
     {
       "epoch": 0.024787754849104542,
-      "grad_norm": 0.22754977237904275,
+      "grad_norm": 0.12060637138665908,
       "learning_rate": 4.9566294919454776e-05,
-      "loss": 4.0586,
+      "loss": 2.0344,
       "step": 200
     },
     {
       "epoch": 0.025407448720332158,
-      "grad_norm": 0.21627590215778986,
+      "grad_norm": 0.11902127583462105,
       "learning_rate": 5.080545229244115e-05,
-      "loss": 4.057,
+      "loss": 2.0461,
       "step": 205
     },
     {
       "epoch": 0.02602714259155977,
-      "grad_norm": 0.22617748275994737,
+      "grad_norm": 0.11966181981073709,
       "learning_rate": 5.2044609665427516e-05,
-      "loss": 4.0891,
+      "loss": 2.0465,
       "step": 210
     },
     {
       "epoch": 0.02664683646278738,
-      "grad_norm": 0.2210255983769268,
+      "grad_norm": 0.11020817117690564,
       "learning_rate": 5.328376703841388e-05,
-      "loss": 4.0289,
+      "loss": 2.0453,
       "step": 215
     },
     {
       "epoch": 0.027266530334014997,
-      "grad_norm": 0.23704324491853482,
+      "grad_norm": 0.11068723686941234,
       "learning_rate": 5.4522924411400255e-05,
-      "loss": 4.1234,
+      "loss": 2.0457,
       "step": 220
     },
     {
       "epoch": 0.02788622420524261,
-      "grad_norm": 0.21415865510311569,
+      "grad_norm": 0.10772518213779579,
       "learning_rate": 5.576208178438662e-05,
-      "loss": 3.9641,
+      "loss": 2.0211,
       "step": 225
     },
     {
       "epoch": 0.028505918076470224,
-      "grad_norm": 0.20869571114446672,
+      "grad_norm": 0.11617312325311774,
       "learning_rate": 5.700123915737299e-05,
-      "loss": 4.0156,
+      "loss": 2.0199,
       "step": 230
     },
     {
       "epoch": 0.029125611947697836,
-      "grad_norm": 0.2295558494942413,
+      "grad_norm": 0.10687892359246541,
       "learning_rate": 5.8240396530359354e-05,
-      "loss": 4.0008,
+      "loss": 2.0168,
       "step": 235
     },
     {
       "epoch": 0.02974530581892545,
-      "grad_norm": 0.20993210502465767,
+      "grad_norm": 0.10534898334754005,
       "learning_rate": 5.947955390334573e-05,
-      "loss": 4.2164,
+      "loss": 1.9563,
       "step": 240
     },
     {
       "epoch": 0.030364999690153063,
-      "grad_norm": 0.22805360919484266,
+      "grad_norm": 0.11188760033211348,
       "learning_rate": 6.0718711276332094e-05,
-      "loss": 4.0617,
+      "loss": 2.0211,
       "step": 245
     },
     {
       "epoch": 0.03098469356138068,
-      "grad_norm": 0.21679836800491897,
+      "grad_norm": 0.11217525164835511,
       "learning_rate": 6.195786864931847e-05,
-      "loss": 3.9891,
+      "loss": 2.0055,
       "step": 250
     },
     {
       "epoch": 0.031604387432608294,
-      "grad_norm": 0.20555177272110325,
+      "grad_norm": 0.1123864808685792,
       "learning_rate": 6.319702602230483e-05,
-      "loss": 3.9547,
+      "loss": 2.0078,
       "step": 255
     },
     {
       "epoch": 0.0322240813038359,
-      "grad_norm": 0.22756620407217162,
+      "grad_norm": 0.10988735365168292,
       "learning_rate": 6.44361833952912e-05,
-      "loss": 4.0219,
+      "loss": 1.9512,
       "step": 260
     },
     {
       "epoch": 0.03284377517506352,
-      "grad_norm": 0.21772033945390115,
+      "grad_norm": 0.11717299164220939,
       "learning_rate": 6.567534076827757e-05,
-      "loss": 4.0328,
+      "loss": 2.0352,
       "step": 265
     },
     {
       "epoch": 0.03346346904629113,
-      "grad_norm": 0.22974604776495353,
+      "grad_norm": 0.11127165660011876,
       "learning_rate": 6.691449814126395e-05,
-      "loss": 3.8969,
+      "loss": 2.0063,
       "step": 270
     },
     {
       "epoch": 0.03408316291751875,
-      "grad_norm": 0.22447015222263075,
+      "grad_norm": 0.11434862645507657,
       "learning_rate": 6.815365551425031e-05,
-      "loss": 3.957,
+      "loss": 1.8934,
       "step": 275
     },
     {
       "epoch": 0.03470285678874636,
-      "grad_norm": 0.2126380837798873,
+      "grad_norm": 0.11029266191053742,
       "learning_rate": 6.939281288723668e-05,
-      "loss": 3.9594,
+      "loss": 1.9535,
       "step": 280
     },
     {
       "epoch": 0.03532255065997397,
-      "grad_norm": 0.228888586034849,
+      "grad_norm": 0.1118252814826643,
       "learning_rate": 7.063197026022306e-05,
-      "loss": 3.982,
+      "loss": 2.0055,
       "step": 285
     },
     {
       "epoch": 0.03594224453120159,
-      "grad_norm": 0.22194460293105397,
+      "grad_norm": 0.1158412859452615,
       "learning_rate": 7.187112763320942e-05,
-      "loss": 4.0539,
+      "loss": 1.9695,
       "step": 290
     },
     {
       "epoch": 0.0365619384024292,
-      "grad_norm": 0.22114527588166774,
+      "grad_norm": 0.10966650026564005,
       "learning_rate": 7.311028500619579e-05,
-      "loss": 4.0297,
+      "loss": 1.9957,
       "step": 295
     },
     {
       "epoch": 0.03718163227365681,
-      "grad_norm": 0.2253633278202601,
+      "grad_norm": 0.11337717400230966,
       "learning_rate": 7.434944237918216e-05,
-      "loss": 3.9047,
+      "loss": 1.923,
       "step": 300
     },
     {
       "epoch": 0.03780132614488443,
-      "grad_norm": 0.21472709323547942,
+      "grad_norm": 0.10903076207822042,
       "learning_rate": 7.558859975216854e-05,
-      "loss": 3.9953,
+      "loss": 1.9684,
       "step": 305
     },
     {
       "epoch": 0.03842102001611204,
-      "grad_norm": 0.21240689967710813,
+      "grad_norm": 0.10798533162202754,
       "learning_rate": 7.68277571251549e-05,
-      "loss": 3.9609,
+      "loss": 1.9453,
       "step": 310
     },
     {
       "epoch": 0.03904071388733965,
-      "grad_norm": 0.2138973933351751,
+      "grad_norm": 0.11175148400384678,
       "learning_rate": 7.806691449814127e-05,
-      "loss": 3.932,
+      "loss": 1.9535,
       "step": 315
     },
     {
       "epoch": 0.039660407758567266,
-      "grad_norm": 0.21095964738376724,
+      "grad_norm": 0.10703314633558067,
       "learning_rate": 7.930607187112764e-05,
-      "loss": 3.9281,
+      "loss": 1.9563,
       "step": 320
     },
     {
       "epoch": 0.04028010162979488,
-      "grad_norm": 0.23767132219880815,
+      "grad_norm": 0.10454396198544379,
       "learning_rate": 8.0545229244114e-05,
-      "loss": 3.9227,
+      "loss": 1.9418,
       "step": 325
     },
     {
       "epoch": 0.0408997955010225,
-      "grad_norm": 0.21145595313668683,
+      "grad_norm": 0.10965164879246644,
       "learning_rate": 8.178438661710037e-05,
-      "loss": 3.8734,
+      "loss": 1.9852,
       "step": 330
     },
     {
       "epoch": 0.041519489372250105,
-      "grad_norm": 0.22018542729191426,
+      "grad_norm": 0.11046491057161609,
       "learning_rate": 8.302354399008675e-05,
-      "loss": 4.0695,
+      "loss": 1.9461,
       "step": 335
     },
     {
       "epoch": 0.04213918324347772,
-      "grad_norm": 0.2223712124211028,
+      "grad_norm": 0.10873277373716497,
       "learning_rate": 8.426270136307312e-05,
-      "loss": 3.9984,
+      "loss": 1.8918,
       "step": 340
     },
     {
       "epoch": 0.042758877114705336,
-      "grad_norm": 0.22870383890024354,
+      "grad_norm": 0.11520873752551357,
       "learning_rate": 8.550185873605948e-05,
-      "loss": 3.9656,
+      "loss": 1.9738,
       "step": 345
     },
     {
       "epoch": 0.04337857098593295,
-      "grad_norm": 0.21918918087091638,
+      "grad_norm": 0.11319565385145965,
       "learning_rate": 8.674101610904586e-05,
-      "loss": 3.9547,
+      "loss": 1.8652,
       "step": 350
     },
     {
       "epoch": 0.04399826485716056,
-      "grad_norm": 0.24597902739695576,
+      "grad_norm": 0.11118195705926588,
       "learning_rate": 8.798017348203223e-05,
-      "loss": 3.8828,
+      "loss": 1.9871,
       "step": 355
     },
     {
       "epoch": 0.044617958728388175,
-      "grad_norm": 0.23311556933668512,
+      "grad_norm": 0.11282384360301383,
       "learning_rate": 8.92193308550186e-05,
-      "loss": 3.9781,
+      "loss": 2.0211,
       "step": 360
     },
     {
       "epoch": 0.04523765259961579,
-      "grad_norm": 0.2335730208221731,
+      "grad_norm": 0.12014079628091061,
       "learning_rate": 9.045848822800496e-05,
-      "loss": 3.932,
+      "loss": 1.9898,
       "step": 365
     },
     {
       "epoch": 0.045857346470843406,
-      "grad_norm": 0.22758589754195266,
+      "grad_norm": 0.12069690241752978,
       "learning_rate": 9.169764560099133e-05,
-      "loss": 3.9328,
+      "loss": 1.9766,
       "step": 370
     },
     {
       "epoch": 0.046477040342071015,
-      "grad_norm": 0.21887801195676554,
+      "grad_norm": 0.11312380995936393,
       "learning_rate": 9.29368029739777e-05,
-      "loss": 3.8961,
+      "loss": 1.9609,
       "step": 375
     },
     {
       "epoch": 0.04709673421329863,
-      "grad_norm": 0.23691795117542777,
+      "grad_norm": 0.10660555317363303,
       "learning_rate": 9.417596034696406e-05,
-      "loss": 3.9922,
+      "loss": 1.9391,
       "step": 380
     },
     {
       "epoch": 0.047716428084526245,
-      "grad_norm": 0.22314356605972202,
+      "grad_norm": 0.11034212285475851,
       "learning_rate": 9.541511771995044e-05,
-      "loss": 3.9156,
+      "loss": 1.9402,
       "step": 385
     },
     {
       "epoch": 0.04833612195575386,
-      "grad_norm": 0.22883014470157503,
+      "grad_norm": 0.10700147884403914,
       "learning_rate": 9.66542750929368e-05,
-      "loss": 3.6133,
+      "loss": 1.9547,
       "step": 390
     },
     {
       "epoch": 0.04895581582698147,
-      "grad_norm": 0.2261978377608961,
+      "grad_norm": 0.11111784345239176,
       "learning_rate": 9.789343246592317e-05,
-      "loss": 4.0008,
+      "loss": 1.9352,
       "step": 395
     },
     {
       "epoch": 0.049575509698209085,
-      "grad_norm": 0.23167762207515435,
+      "grad_norm": 0.1156370481856559,
       "learning_rate": 9.913258983890955e-05,
-      "loss": 3.8664,
+      "loss": 1.9441,
       "step": 400
     },
     {
       "epoch": 0.0501952035694367,
-      "grad_norm": 0.23913876095025915,
+      "grad_norm": 0.1138918955121605,
       "learning_rate": 0.0001003717472118959,
-      "loss": 3.8797,
+      "loss": 1.9289,
       "step": 405
     },
     {
       "epoch": 0.050814897440664315,
-      "grad_norm": 0.21300252960890317,
+      "grad_norm": 0.11700414573434677,
       "learning_rate": 0.0001016109045848823,
-      "loss": 3.9016,
+      "loss": 1.9359,
       "step": 410
     },
     {
       "epoch": 0.051434591311891924,
-      "grad_norm": 0.24940978228703184,
+      "grad_norm": 0.11419554025531545,
       "learning_rate": 0.00010285006195786867,
-      "loss": 3.7727,
+      "loss": 1.9324,
       "step": 415
     },
     {
       "epoch": 0.05205428518311954,
-      "grad_norm": 0.23257277242036176,
+      "grad_norm": 0.11511324040424104,
       "learning_rate": 0.00010408921933085503,
-      "loss": 3.9555,
+      "loss": 1.9391,
       "step": 420
     },
     {
       "epoch": 0.052673979054347154,
-      "grad_norm": 0.22550835138918388,
+      "grad_norm": 0.11598251416819207,
       "learning_rate": 0.0001053283767038414,
-      "loss": 3.8078,
+      "loss": 1.8605,
       "step": 425
     },
     {
       "epoch": 0.05329367292557476,
-      "grad_norm": 0.2214992349435567,
+      "grad_norm": 0.11945825058588103,
       "learning_rate": 0.00010656753407682776,
-      "loss": 3.8711,
+      "loss": 1.9219,
       "step": 430
     },
     {
       "epoch": 0.05391336679680238,
-      "grad_norm": 0.2385884965327802,
+      "grad_norm": 0.12186744920368066,
       "learning_rate": 0.00010780669144981412,
-      "loss": 3.8664,
+      "loss": 1.9664,
       "step": 435
     },
     {
       "epoch": 0.054533060668029994,
-      "grad_norm": 0.2393533535483792,
+      "grad_norm": 0.11978444755418904,
       "learning_rate": 0.00010904584882280051,
-      "loss": 3.9273,
+      "loss": 1.9582,
       "step": 440
     },
     {
       "epoch": 0.05515275453925761,
-      "grad_norm": 0.23892895926770114,
+      "grad_norm": 0.11151838879430113,
       "learning_rate": 0.00011028500619578688,
-      "loss": 3.9016,
+      "loss": 1.9141,
       "step": 445
     },
     {
       "epoch": 0.05577244841048522,
-      "grad_norm": 0.25293116707319535,
+      "grad_norm": 0.11753526250283612,
       "learning_rate": 0.00011152416356877324,
-      "loss": 3.8977,
+      "loss": 1.9355,
       "step": 450
     },
     {
       "epoch": 0.05639214228171283,
-      "grad_norm": 0.2339751975680064,
+      "grad_norm": 0.12027058315054732,
       "learning_rate": 0.00011276332094175961,
-      "loss": 3.8297,
+      "loss": 1.9434,
       "step": 455
     },
     {
       "epoch": 0.05701183615294045,
-      "grad_norm": 0.22040331210242547,
+      "grad_norm": 0.12114137216676991,
       "learning_rate": 0.00011400247831474598,
-      "loss": 3.875,
+      "loss": 1.8844,
       "step": 460
     },
     {
       "epoch": 0.057631530024168064,
-      "grad_norm": 0.22773047199756685,
+      "grad_norm": 0.11303473278613534,
       "learning_rate": 0.00011524163568773234,
-      "loss": 3.8602,
+      "loss": 1.9352,
       "step": 465
     },
     {
       "epoch": 0.05825122389539567,
-      "grad_norm": 0.21974486332656235,
+      "grad_norm": 0.12224817657479514,
       "learning_rate": 0.00011648079306071871,
-      "loss": 3.8344,
+      "loss": 1.957,
       "step": 470
     },
     {
       "epoch": 0.05887091776662329,
-      "grad_norm": 0.24683791311372597,
+      "grad_norm": 0.11169054867070344,
       "learning_rate": 0.00011771995043370509,
-      "loss": 3.8414,
+      "loss": 1.9418,
       "step": 475
     },
     {
       "epoch": 0.0594906116378509,
-      "grad_norm": 0.22784797807715068,
+      "grad_norm": 0.11248246802350605,
       "learning_rate": 0.00011895910780669145,
-      "loss": 3.7875,
+      "loss": 1.9105,
       "step": 480
     },
     {
       "epoch": 0.06011030550907852,
-      "grad_norm": 0.22112074000806836,
+      "grad_norm": 0.11732018140367607,
       "learning_rate": 0.00012019826517967782,
-      "loss": 3.8953,
+      "loss": 1.9449,
       "step": 485
     },
     {
       "epoch": 0.06072999938030613,
-      "grad_norm": 0.22438071094190917,
+      "grad_norm": 0.12552929938535534,
       "learning_rate": 0.00012143742255266419,
-      "loss": 4.0109,
+      "loss": 1.9254,
       "step": 490
     },
     {
       "epoch": 0.06134969325153374,
-      "grad_norm": 0.24426703173088538,
+      "grad_norm": 0.11437642863096387,
       "learning_rate": 0.00012267657992565055,
-      "loss": 3.8891,
+      "loss": 1.8961,
       "step": 495
     },
     {
       "epoch": 0.06196938712276136,
-      "grad_norm": 0.24796530136608588,
+      "grad_norm": 0.1260600647441986,
       "learning_rate": 0.00012391573729863693,
-      "loss": 3.9898,
+      "loss": 1.8875,
       "step": 500
     },
     {
       "epoch": 0.06258908099398897,
-      "grad_norm": 0.2504226663983639,
+      "grad_norm": 0.11848912394998189,
       "learning_rate": 0.00012515489467162331,
-      "loss": 3.9109,
+      "loss": 1.8762,
       "step": 505
     },
     {
       "epoch": 0.06320877486521659,
-      "grad_norm": 0.24728151359797979,
+      "grad_norm": 0.12197039580850588,
       "learning_rate": 0.00012639405204460967,
-      "loss": 3.7414,
+      "loss": 1.8574,
       "step": 510
     },
     {
       "epoch": 0.0638284687364442,
-      "grad_norm": 0.2568748492336261,
+      "grad_norm": 0.11905836258470184,
       "learning_rate": 0.00012763320941759605,
-      "loss": 3.85,
+      "loss": 1.9691,
       "step": 515
     },
     {
       "epoch": 0.0644481626076718,
-      "grad_norm": 0.2311918499284196,
+      "grad_norm": 0.11404810658049691,
       "learning_rate": 0.0001288723667905824,
-      "loss": 3.8188,
+      "loss": 1.9176,
       "step": 520
     },
     {
       "epoch": 0.06506785647889943,
-      "grad_norm": 0.25292533490820934,
+      "grad_norm": 0.11921939557929893,
       "learning_rate": 0.00013011152416356878,
-      "loss": 3.7531,
+      "loss": 1.9133,
       "step": 525
     },
     {
       "epoch": 0.06568755035012704,
-      "grad_norm": 0.23091615388429254,
+      "grad_norm": 0.1161201458568179,
       "learning_rate": 0.00013135068153655513,
-      "loss": 3.818,
+      "loss": 1.9879,
       "step": 530
     },
     {
       "epoch": 0.06630724422135464,
-      "grad_norm": 0.2423671174824872,
+      "grad_norm": 0.12386141326873432,
       "learning_rate": 0.0001325898389095415,
-      "loss": 3.9406,
+      "loss": 1.9461,
       "step": 535
     },
     {
       "epoch": 0.06692693809258227,
-      "grad_norm": 0.24386526601080538,
+      "grad_norm": 0.12126274245647896,
       "learning_rate": 0.0001338289962825279,
-      "loss": 3.7711,
+      "loss": 1.9969,
       "step": 540
     },
     {
       "epoch": 0.06754663196380987,
-      "grad_norm": 0.2436193360083945,
+      "grad_norm": 0.11855513780674662,
       "learning_rate": 0.00013506815365551427,
-      "loss": 3.7484,
+      "loss": 1.9379,
       "step": 545
     },
     {
       "epoch": 0.0681663258350375,
-      "grad_norm": 0.24440315661081868,
+      "grad_norm": 0.12454701841587497,
       "learning_rate": 0.00013630731102850062,
-      "loss": 3.8039,
+      "loss": 1.893,
       "step": 550
     },
     {
       "epoch": 0.0687860197062651,
-      "grad_norm": 0.24658694689728877,
+      "grad_norm": 0.13585017663988036,
       "learning_rate": 0.000137546468401487,
-      "loss": 3.9062,
+      "loss": 1.9676,
       "step": 555
     },
     {
       "epoch": 0.06940571357749271,
-      "grad_norm": 0.24860076091533706,
+      "grad_norm": 0.12286666020259437,
       "learning_rate": 0.00013878562577447336,
-      "loss": 3.9148,
+      "loss": 1.9734,
       "step": 560
     },
     {
       "epoch": 0.07002540744872034,
-      "grad_norm": 0.24643140253653378,
+      "grad_norm": 0.12842932394891735,
       "learning_rate": 0.00014002478314745974,
-      "loss": 3.7945,
+      "loss": 1.9359,
       "step": 565
     },
     {
       "epoch": 0.07064510131994794,
-      "grad_norm": 0.24897544182458003,
+      "grad_norm": 0.12453124899243173,
       "learning_rate": 0.00014126394052044612,
-      "loss": 3.8984,
+      "loss": 1.8824,
       "step": 570
     },
     {
       "epoch": 0.07126479519117555,
-      "grad_norm": 0.2553723832902909,
+      "grad_norm": 0.12198945086913854,
       "learning_rate": 0.00014250309789343247,
-      "loss": 3.9109,
+      "loss": 1.9359,
       "step": 575
     },
     {
       "epoch": 0.07188448906240318,
-      "grad_norm": 0.25975637237924604,
+      "grad_norm": 0.11984988586999516,
       "learning_rate": 0.00014374225526641885,
-      "loss": 3.8664,
+      "loss": 1.9258,
       "step": 580
     },
     {
       "epoch": 0.07250418293363078,
-      "grad_norm": 0.2436903470260683,
+      "grad_norm": 0.11931391810243619,
       "learning_rate": 0.0001449814126394052,
-      "loss": 3.8789,
+      "loss": 1.9586,
       "step": 585
     },
     {
       "epoch": 0.0731238768048584,
-      "grad_norm": 0.24798071594725046,
+      "grad_norm": 0.1218417705429113,
       "learning_rate": 0.00014622057001239158,
-      "loss": 3.8836,
+      "loss": 1.882,
       "step": 590
     },
     {
       "epoch": 0.07374357067608601,
-      "grad_norm": 0.2649972707608675,
+      "grad_norm": 0.12362223035885297,
       "learning_rate": 0.00014745972738537794,
-      "loss": 3.7984,
+      "loss": 1.9336,
       "step": 595
     },
     {
       "epoch": 0.07436326454731362,
-      "grad_norm": 0.2647217197239789,
+      "grad_norm": 0.12709267111450237,
       "learning_rate": 0.00014869888475836432,
-      "loss": 3.7625,
+      "loss": 1.9195,
       "step": 600
     },
     {
       "epoch": 0.07498295841854125,
-      "grad_norm": 0.24856131885729846,
+      "grad_norm": 0.12352240140821688,
       "learning_rate": 0.0001499380421313507,
-      "loss": 3.9086,
+      "loss": 1.9566,
       "step": 605
     },
     {
       "epoch": 0.07560265228976885,
-      "grad_norm": 0.2703748729511876,
+      "grad_norm": 0.11695893050509344,
       "learning_rate": 0.00015117719950433707,
-      "loss": 3.8547,
+      "loss": 1.9055,
       "step": 610
     },
     {
       "epoch": 0.07622234616099646,
-      "grad_norm": 0.24654173035745675,
+      "grad_norm": 0.12061523502751974,
       "learning_rate": 0.00015241635687732343,
-      "loss": 3.8789,
+      "loss": 1.8879,
       "step": 615
     },
     {
       "epoch": 0.07684204003222408,
-      "grad_norm": 0.2692972261734959,
+      "grad_norm": 0.13093586226295037,
       "learning_rate": 0.0001536555142503098,
-      "loss": 3.8945,
+      "loss": 1.9043,
       "step": 620
     },
     {
       "epoch": 0.0774617339034517,
-      "grad_norm": 0.27594609338449094,
+      "grad_norm": 0.12506688122473264,
       "learning_rate": 0.00015489467162329616,
-      "loss": 3.8836,
+      "loss": 1.9359,
       "step": 625
     },
     {
       "epoch": 0.0780814277746793,
-      "grad_norm": 0.25550613767047675,
+      "grad_norm": 0.12410187155175814,
       "learning_rate": 0.00015613382899628254,
-      "loss": 3.8148,
+      "loss": 1.9324,
       "step": 630
     },
     {
       "epoch": 0.07870112164590692,
-      "grad_norm": 0.25399112427320353,
+      "grad_norm": 0.1272473564630399,
       "learning_rate": 0.00015737298636926892,
-      "loss": 3.8023,
+      "loss": 1.9199,
       "step": 635
     },
     {
       "epoch": 0.07932081551713453,
-      "grad_norm": 0.26411453154811365,
+      "grad_norm": 0.13031962614082446,
       "learning_rate": 0.00015861214374225527,
-      "loss": 3.8914,
+      "loss": 1.8988,
       "step": 640
     },
     {
       "epoch": 0.07994050938836215,
-      "grad_norm": 0.2681270955278968,
+      "grad_norm": 0.14034276494108713,
       "learning_rate": 0.00015985130111524165,
-      "loss": 3.7773,
+      "loss": 1.9531,
       "step": 645
     },
     {
       "epoch": 0.08056020325958976,
-      "grad_norm": 0.25446306197588175,
+      "grad_norm": 0.1240075132613875,
       "learning_rate": 0.000161090458488228,
-      "loss": 3.8664,
+      "loss": 1.8977,
       "step": 650
     },
     {
       "epoch": 0.08117989713081737,
-      "grad_norm": 0.24587567906822394,
+      "grad_norm": 0.12295043509444326,
       "learning_rate": 0.00016232961586121439,
-      "loss": 3.9359,
+      "loss": 1.9316,
       "step": 655
     },
     {
       "epoch": 0.081799591002045,
-      "grad_norm": 0.2418693517722617,
+      "grad_norm": 0.1307165114400111,
       "learning_rate": 0.00016356877323420074,
-      "loss": 3.8328,
+      "loss": 1.9141,
       "step": 660
     },
     {
       "epoch": 0.0824192848732726,
-      "grad_norm": 0.2506664149033651,
+      "grad_norm": 0.12576570368395784,
       "learning_rate": 0.00016480793060718712,
-      "loss": 3.8836,
+      "loss": 1.9578,
       "step": 665
     },
     {
       "epoch": 0.08303897874450021,
-      "grad_norm": 0.26233517731779643,
+      "grad_norm": 0.1279849537637525,
       "learning_rate": 0.0001660470879801735,
-      "loss": 3.7844,
+      "loss": 1.8535,
       "step": 670
     },
     {
       "epoch": 0.08365867261572783,
-      "grad_norm": 0.2677004498186573,
+      "grad_norm": 0.1277752146528915,
       "learning_rate": 0.00016728624535315988,
-      "loss": 3.8125,
+      "loss": 1.9305,
       "step": 675
     },
     {
       "epoch": 0.08427836648695544,
-      "grad_norm": 0.2398641506201725,
+      "grad_norm": 0.13382769963247415,
       "learning_rate": 0.00016852540272614623,
-      "loss": 3.8055,
+      "loss": 1.8758,
       "step": 680
     },
     {
       "epoch": 0.08489806035818306,
-      "grad_norm": 0.2602146016824983,
+      "grad_norm": 0.1338785424771461,
       "learning_rate": 0.00016976456009913258,
-      "loss": 3.9375,
+      "loss": 1.8848,
       "step": 685
     },
     {
       "epoch": 0.08551775422941067,
-      "grad_norm": 0.266805223760884,
+      "grad_norm": 0.13378100950887833,
       "learning_rate": 0.00017100371747211896,
-      "loss": 3.9594,
+      "loss": 1.9797,
       "step": 690
     },
     {
       "epoch": 0.08613744810063828,
-      "grad_norm": 0.2500429973976006,
+      "grad_norm": 0.12759853308304364,
       "learning_rate": 0.00017224287484510532,
-      "loss": 3.6695,
+      "loss": 1.927,
       "step": 695
     },
     {
       "epoch": 0.0867571419718659,
-      "grad_norm": 0.2603605758250552,
+      "grad_norm": 0.13003091028679756,
       "learning_rate": 0.00017348203221809172,
-      "loss": 3.8656,
+      "loss": 1.9258,
       "step": 700
     },
     {
       "epoch": 0.08737683584309351,
-      "grad_norm": 0.2817796768485939,
+      "grad_norm": 0.1256819224648164,
       "learning_rate": 0.00017472118959107808,
-      "loss": 3.7898,
+      "loss": 1.8852,
       "step": 705
     },
     {
       "epoch": 0.08799652971432112,
-      "grad_norm": 0.26066467074037064,
+      "grad_norm": 0.12540250359127755,
       "learning_rate": 0.00017596034696406446,
-      "loss": 3.757,
+      "loss": 1.9125,
       "step": 710
     },
     {
       "epoch": 0.08861622358554874,
-      "grad_norm": 0.264331246410504,
+      "grad_norm": 0.1262469778474483,
       "learning_rate": 0.0001771995043370508,
-      "loss": 3.8305,
+      "loss": 1.9309,
       "step": 715
     },
     {
       "epoch": 0.08923591745677635,
-      "grad_norm": 0.26208453033679957,
+      "grad_norm": 0.12256263065722443,
       "learning_rate": 0.0001784386617100372,
-      "loss": 3.8859,
+      "loss": 1.8797,
       "step": 720
     },
     {
       "epoch": 0.08985561132800396,
-      "grad_norm": 0.2608654719771367,
+      "grad_norm": 0.12933452681614585,
       "learning_rate": 0.00017967781908302354,
-      "loss": 3.8828,
+      "loss": 1.902,
       "step": 725
     },
     {
       "epoch": 0.09047530519923158,
-      "grad_norm": 0.26424596197452516,
+      "grad_norm": 0.12866605933389477,
       "learning_rate": 0.00018091697645600992,
-      "loss": 3.6992,
+      "loss": 1.9633,
       "step": 730
     },
     {
       "epoch": 0.09109499907045919,
-      "grad_norm": 0.2554376625702158,
+      "grad_norm": 0.11929142114827074,
       "learning_rate": 0.0001821561338289963,
-      "loss": 3.9094,
+      "loss": 1.9008,
       "step": 735
     },
     {
       "epoch": 0.09171469294168681,
-      "grad_norm": 0.25635680373399056,
+      "grad_norm": 0.12559495279922095,
       "learning_rate": 0.00018339529120198265,
-      "loss": 3.7016,
+      "loss": 1.898,
       "step": 740
     },
     {
       "epoch": 0.09233438681291442,
-      "grad_norm": 0.24261097763834139,
+      "grad_norm": 0.1351676419101842,
       "learning_rate": 0.00018463444857496903,
-      "loss": 3.7203,
+      "loss": 1.9105,
       "step": 745
     },
     {
       "epoch": 0.09295408068414203,
-      "grad_norm": 0.26970318223299494,
+      "grad_norm": 0.12357127136653635,
       "learning_rate": 0.0001858736059479554,
-      "loss": 3.8492,
+      "loss": 1.8855,
       "step": 750
     },
     {
       "epoch": 0.09357377455536965,
-      "grad_norm": 0.30498817244113025,
+      "grad_norm": 0.1256351267334546,
       "learning_rate": 0.00018711276332094177,
-      "loss": 3.9031,
+      "loss": 1.8832,
       "step": 755
     },
     {
       "epoch": 0.09419346842659726,
-      "grad_norm": 0.25720828519533,
+      "grad_norm": 0.1263964187356334,
       "learning_rate": 0.00018835192069392812,
-      "loss": 3.8156,
+      "loss": 1.9742,
       "step": 760
     },
     {
       "epoch": 0.09481316229782487,
-      "grad_norm": 0.29468196401306584,
+      "grad_norm": 0.125010680336928,
       "learning_rate": 0.00018959107806691453,
-      "loss": 3.7273,
+      "loss": 1.9094,
       "step": 765
     },
     {
       "epoch": 0.09543285616905249,
-      "grad_norm": 0.2684612791109344,
+      "grad_norm": 0.12818194425097587,
       "learning_rate": 0.00019083023543990088,
-      "loss": 3.8953,
+      "loss": 1.9152,
       "step": 770
     },
     {
       "epoch": 0.0960525500402801,
-      "grad_norm": 0.2667571911932907,
+      "grad_norm": 0.1340819661726965,
       "learning_rate": 0.00019206939281288726,
-      "loss": 3.7086,
+      "loss": 1.8914,
       "step": 775
     },
     {
       "epoch": 0.09667224391150772,
-      "grad_norm": 0.2456038107295753,
+      "grad_norm": 0.12618688913153653,
       "learning_rate": 0.0001933085501858736,
-      "loss": 3.8047,
+      "loss": 1.9062,
       "step": 780
     },
     {
       "epoch": 0.09729193778273533,
-      "grad_norm": 0.2621406327829109,
+      "grad_norm": 0.12816103290715988,
       "learning_rate": 0.00019454770755886,
-      "loss": 3.6516,
+      "loss": 1.868,
       "step": 785
     },
     {
       "epoch": 0.09791163165396294,
-      "grad_norm": 0.2620837502334776,
+      "grad_norm": 0.1344741373451341,
       "learning_rate": 0.00019578686493184635,
-      "loss": 3.8641,
+      "loss": 1.9312,
       "step": 790
     },
     {
       "epoch": 0.09853132552519056,
-      "grad_norm": 0.25330053822524057,
+      "grad_norm": 0.12832878658024055,
       "learning_rate": 0.00019702602230483272,
-      "loss": 3.7797,
+      "loss": 1.8883,
       "step": 795
     },
     {
       "epoch": 0.09915101939641817,
-      "grad_norm": 0.2529835083853911,
+      "grad_norm": 0.13311964726869252,
       "learning_rate": 0.0001982651796778191,
-      "loss": 3.7844,
+      "loss": 1.9547,
       "step": 800
     },
     {
       "epoch": 0.09977071326764578,
-      "grad_norm": 0.27582627311828656,
+      "grad_norm": 0.12760395325139873,
       "learning_rate": 0.00019950433705080546,
-      "loss": 3.8797,
+      "loss": 1.9055,
       "step": 805
     },
     {
       "epoch": 0.1003904071388734,
-      "grad_norm": 0.24831818922645793,
+      "grad_norm": 0.14141498619951107,
       "learning_rate": 0.00019999991575981254,
-      "loss": 3.7563,
+      "loss": 1.8609,
       "step": 810
     },
     {
       "epoch": 0.10101010101010101,
-      "grad_norm": 0.24819054534630788,
+      "grad_norm": 0.12959126389640468,
       "learning_rate": 0.00019999940095918086,
-      "loss": 3.8828,
+      "loss": 1.9059,
       "step": 815
     },
     {
       "epoch": 0.10162979488132863,
-      "grad_norm": 0.28265968045148,
+      "grad_norm": 0.12650382383931852,
       "learning_rate": 0.000199998418160428,
-      "loss": 3.8594,
+      "loss": 1.9281,
       "step": 820
     },
     {
       "epoch": 0.10224948875255624,
-      "grad_norm": 0.2733982142303492,
+      "grad_norm": 0.12221495910665915,
       "learning_rate": 0.00019999696736815346,
-      "loss": 3.8406,
+      "loss": 1.9273,
       "step": 825
     },
     {
       "epoch": 0.10286918262378385,
-      "grad_norm": 0.29382519994780215,
+      "grad_norm": 0.1287913010486784,
       "learning_rate": 0.000199995048589147,
-      "loss": 3.6219,
+      "loss": 1.8945,
       "step": 830
     },
     {
       "epoch": 0.10348887649501147,
-      "grad_norm": 0.2854445476699885,
+      "grad_norm": 0.1311248902853389,
       "learning_rate": 0.00019999266183238847,
-      "loss": 3.7812,
+      "loss": 1.8992,
       "step": 835
     },
     {
       "epoch": 0.10410857036623908,
-      "grad_norm": 0.26006686205278173,
+      "grad_norm": 0.1349754560323098,
       "learning_rate": 0.00019998980710904794,
-      "loss": 3.7742,
+      "loss": 1.9187,
       "step": 840
     },
     {
       "epoch": 0.10472826423746669,
-      "grad_norm": 0.25601656542470685,
+      "grad_norm": 0.13835648010235793,
       "learning_rate": 0.00019998648443248556,
-      "loss": 3.7938,
+      "loss": 1.9113,
       "step": 845
     },
     {
       "epoch": 0.10534795810869431,
-      "grad_norm": 0.23820163447829,
+      "grad_norm": 0.12332364542564311,
       "learning_rate": 0.00019998269381825147,
-      "loss": 3.6883,
+      "loss": 1.8734,
       "step": 850
     },
     {
       "epoch": 0.10596765197992192,
-      "grad_norm": 0.24260416655510744,
+      "grad_norm": 0.13003332916776286,
       "learning_rate": 0.00019997843528408576,
-      "loss": 3.7914,
+      "loss": 1.8773,
       "step": 855
     },
     {
       "epoch": 0.10658734585114953,
-      "grad_norm": 0.25457045020460534,
+      "grad_norm": 0.1375059879901149,
       "learning_rate": 0.00019997370884991842,
-      "loss": 3.8531,
+      "loss": 1.9059,
       "step": 860
     },
     {
       "epoch": 0.10720703972237715,
-      "grad_norm": 0.24724606814852315,
+      "grad_norm": 0.1248324688392207,
       "learning_rate": 0.0001999685145378692,
-      "loss": 3.668,
+      "loss": 1.8902,
       "step": 865
     },
     {
       "epoch": 0.10782673359360476,
-      "grad_norm": 0.24533597733438564,
+      "grad_norm": 0.13819337572100693,
       "learning_rate": 0.00019996285237224758,
-      "loss": 3.6961,
+      "loss": 1.857,
       "step": 870
     },
     {
       "epoch": 0.10844642746483238,
-      "grad_norm": 0.24640097398276353,
+      "grad_norm": 0.1311935728170451,
       "learning_rate": 0.00019995672237955246,
-      "loss": 3.7758,
+      "loss": 1.9184,
       "step": 875
     },
     {
       "epoch": 0.10906612133605999,
-      "grad_norm": 0.2576016534107627,
+      "grad_norm": 0.12800957952899197,
       "learning_rate": 0.00019995012458847233,
-      "loss": 3.8813,
+      "loss": 1.9172,
       "step": 880
     },
     {
       "epoch": 0.1096858152072876,
-      "grad_norm": 0.26762877144872255,
+      "grad_norm": 0.13668641206418503,
       "learning_rate": 0.00019994305902988488,
-      "loss": 3.7875,
+      "loss": 1.8477,
       "step": 885
     },
     {
       "epoch": 0.11030550907851522,
-      "grad_norm": 0.2828605598611797,
+      "grad_norm": 0.13119404407174387,
       "learning_rate": 0.00019993552573685703,
-      "loss": 3.7727,
+      "loss": 1.952,
       "step": 890
     },
     {
       "epoch": 0.11092520294974283,
-      "grad_norm": 0.2530807405713651,
+      "grad_norm": 0.12829852963816493,
       "learning_rate": 0.00019992752474464463,
-      "loss": 3.7836,
+      "loss": 1.8984,
       "step": 895
     },
     {
       "epoch": 0.11154489682097044,
-      "grad_norm": 0.24995839965943753,
+      "grad_norm": 0.13471553863362062,
       "learning_rate": 0.00019991905609069237,
-      "loss": 3.7469,
+      "loss": 1.9227,
       "step": 900
     },
     {
       "epoch": 0.11216459069219806,
-      "grad_norm": 0.2445922348449931,
+      "grad_norm": 0.13020408574344738,
       "learning_rate": 0.0001999101198146337,
-      "loss": 3.8023,
+      "loss": 1.8969,
       "step": 905
     },
     {
       "epoch": 0.11278428456342567,
-      "grad_norm": 0.25886682005652195,
+      "grad_norm": 0.13094295669195982,
       "learning_rate": 0.0001999007159582904,
-      "loss": 3.7359,
+      "loss": 1.8957,
       "step": 910
     },
     {
       "epoch": 0.11340397843465329,
-      "grad_norm": 0.2596083110218009,
+      "grad_norm": 0.1500020819120754,
       "learning_rate": 0.00019989084456567267,
-      "loss": 3.9406,
+      "loss": 1.8844,
       "step": 915
     },
     {
       "epoch": 0.1140236723058809,
-      "grad_norm": 0.27691730769188033,
+      "grad_norm": 0.13558321660460868,
       "learning_rate": 0.00019988050568297866,
-      "loss": 3.7781,
+      "loss": 1.909,
       "step": 920
     },
     {
       "epoch": 0.1146433661771085,
-      "grad_norm": 0.24898319031007313,
+      "grad_norm": 0.134435880311451,
       "learning_rate": 0.0001998696993585945,
-      "loss": 3.7016,
+      "loss": 1.9082,
       "step": 925
     },
     {
       "epoch": 0.11526306004833613,
-      "grad_norm": 0.26070691986935773,
+      "grad_norm": 0.13545407410456206,
       "learning_rate": 0.00019985842564309382,
-      "loss": 3.7398,
+      "loss": 1.9605,
       "step": 930
     },
     {
       "epoch": 0.11588275391956374,
-      "grad_norm": 0.2680878780430638,
+      "grad_norm": 0.12233937748981194,
       "learning_rate": 0.00019984668458923775,
-      "loss": 3.7141,
+      "loss": 1.8906,
       "step": 935
     },
     {
       "epoch": 0.11650244779079134,
-      "grad_norm": 0.26826550613584305,
+      "grad_norm": 0.12657415594685154,
       "learning_rate": 0.00019983447625197457,
-      "loss": 3.7281,
+      "loss": 1.8859,
       "step": 940
     },
     {
       "epoch": 0.11712214166201897,
-      "grad_norm": 0.24978764564648406,
+      "grad_norm": 0.12294879566199171,
       "learning_rate": 0.0001998218006884393,
-      "loss": 3.8539,
+      "loss": 1.9539,
       "step": 945
     },
     {
       "epoch": 0.11774183553324657,
-      "grad_norm": 0.24266170387051966,
+      "grad_norm": 0.12925893885107417,
       "learning_rate": 0.0001998086579579538,
-      "loss": 3.8109,
+      "loss": 1.8777,
       "step": 950
     },
     {
       "epoch": 0.11836152940447418,
-      "grad_norm": 0.2456266645553349,
+      "grad_norm": 0.12420792452753447,
       "learning_rate": 0.00019979504812202612,
-      "loss": 3.8133,
+      "loss": 1.907,
       "step": 955
     },
     {
       "epoch": 0.1189812232757018,
-      "grad_norm": 0.257918004226877,
+      "grad_norm": 0.1341600688997077,
       "learning_rate": 0.00019978097124435042,
-      "loss": 3.7789,
+      "loss": 1.9879,
       "step": 960
     },
     {
       "epoch": 0.11960091714692941,
-      "grad_norm": 0.25054565001999723,
+      "grad_norm": 0.12957070872516174,
       "learning_rate": 0.00019976642739080665,
-      "loss": 3.8383,
+      "loss": 1.8902,
       "step": 965
     },
     {
       "epoch": 0.12022061101815704,
-      "grad_norm": 0.24754360275279963,
+      "grad_norm": 0.12730629525810772,
       "learning_rate": 0.00019975141662946014,
-      "loss": 3.832,
+      "loss": 1.8961,
       "step": 970
     },
     {
       "epoch": 0.12084030488938464,
-      "grad_norm": 0.2649378383036184,
+      "grad_norm": 0.13670751436447726,
       "learning_rate": 0.0001997359390305614,
-      "loss": 3.7023,
+      "loss": 1.8367,
       "step": 975
     },
     {
       "epoch": 0.12145999876061225,
-      "grad_norm": 0.24620912612433915,
+      "grad_norm": 0.12099186401629365,
       "learning_rate": 0.00019971999466654577,
-      "loss": 3.6609,
+      "loss": 1.8691,
       "step": 980
     },
     {
       "epoch": 0.12207969263183988,
-      "grad_norm": 0.2664106735221588,
+      "grad_norm": 0.12888102009545702,
       "learning_rate": 0.000199703583612033,
-      "loss": 3.793,
+      "loss": 1.8719,
       "step": 985
     },
     {
       "epoch": 0.12269938650306748,
-      "grad_norm": 0.2525567461568162,
+      "grad_norm": 0.12535560588591485,
       "learning_rate": 0.00019968670594382694,
-      "loss": 3.9133,
+      "loss": 1.918,
       "step": 990
     },
     {
       "epoch": 0.12331908037429509,
-      "grad_norm": 0.24961131643058485,
+      "grad_norm": 0.13738798941417166,
       "learning_rate": 0.00019966936174091527,
-      "loss": 3.7922,
+      "loss": 1.8742,
       "step": 995
     },
     {
       "epoch": 0.12393877424552271,
-      "grad_norm": 0.25735420432831124,
+      "grad_norm": 0.12749783789851546,
       "learning_rate": 0.00019965155108446906,
-      "loss": 3.8508,
+      "loss": 1.9262,
       "step": 1000
     },
     {
       "epoch": 0.12455846811675032,
-      "grad_norm": 0.2465006690918658,
+      "grad_norm": 0.1297736019911043,
       "learning_rate": 0.00019963327405784226,
-      "loss": 3.7047,
+      "loss": 1.9117,
       "step": 1005
     },
     {
       "epoch": 0.12517816198797793,
-      "grad_norm": 0.2624010766660747,
+      "grad_norm": 0.14543122470335054,
       "learning_rate": 0.0001996145307465716,
-      "loss": 3.8375,
+      "loss": 1.9094,
       "step": 1010
     },
     {
       "epoch": 0.12579785585920555,
-      "grad_norm": 0.2794925439442463,
+      "grad_norm": 0.12860903642292762,
       "learning_rate": 0.00019959532123837588,
-      "loss": 3.8727,
+      "loss": 1.923,
       "step": 1015
     },
     {
       "epoch": 0.12641754973043318,
-      "grad_norm": 0.25253463553227007,
+      "grad_norm": 0.13001166660436106,
       "learning_rate": 0.00019957564562315583,
-      "loss": 3.7984,
+      "loss": 1.8422,
       "step": 1020
     },
     {
       "epoch": 0.12703724360166077,
-      "grad_norm": 0.26407657263039247,
+      "grad_norm": 0.12915585032705884,
       "learning_rate": 0.0001995555039929935,
-      "loss": 3.7,
+      "loss": 1.9133,
       "step": 1025
     },
     {
       "epoch": 0.1276569374728884,
-      "grad_norm": 0.26992370813940614,
+      "grad_norm": 0.12545620705652993,
       "learning_rate": 0.00019953489644215186,
-      "loss": 3.8031,
+      "loss": 1.9145,
       "step": 1030
     },
     {
       "epoch": 0.12827663134411602,
-      "grad_norm": 0.2619589866347836,
+      "grad_norm": 0.1281887633125648,
       "learning_rate": 0.00019951382306707452,
-      "loss": 3.8203,
+      "loss": 1.8559,
       "step": 1035
     },
     {
       "epoch": 0.1288963252153436,
-      "grad_norm": 0.24801681267457396,
+      "grad_norm": 0.12889758276242114,
       "learning_rate": 0.00019949228396638503,
-      "loss": 3.8016,
+      "loss": 1.8949,
       "step": 1040
     },
     {
       "epoch": 0.12951601908657123,
-      "grad_norm": 0.23959097732969808,
+      "grad_norm": 0.13027383363416042,
       "learning_rate": 0.00019947027924088656,
-      "loss": 3.9289,
+      "loss": 1.875,
       "step": 1045
     },
     {
       "epoch": 0.13013571295779885,
-      "grad_norm": 0.2644886464951842,
+      "grad_norm": 0.1289194496997327,
       "learning_rate": 0.00019944780899356146,
-      "loss": 3.7023,
+      "loss": 1.9328,
       "step": 1050
     },
     {
       "epoch": 0.13075540682902645,
-      "grad_norm": 0.28045523884024126,
+      "grad_norm": 0.13042414632394062,
       "learning_rate": 0.00019942487332957066,
-      "loss": 3.7516,
+      "loss": 1.8754,
       "step": 1055
     },
     {
       "epoch": 0.13137510070025407,
-      "grad_norm": 0.2568392250391569,
+      "grad_norm": 0.1330424701830741,
       "learning_rate": 0.0001994014723562533,
-      "loss": 3.7563,
+      "loss": 1.8813,
       "step": 1060
     },
     {
       "epoch": 0.1319947945714817,
-      "grad_norm": 0.27666747511265155,
+      "grad_norm": 0.13484246451599052,
       "learning_rate": 0.00019937760618312617,
-      "loss": 3.7469,
+      "loss": 1.9234,
       "step": 1065
     },
     {
       "epoch": 0.1326144884427093,
-      "grad_norm": 0.2669419700479829,
+      "grad_norm": 0.14282459314129015,
       "learning_rate": 0.00019935327492188315,
-      "loss": 3.7328,
+      "loss": 1.9066,
       "step": 1070
     },
     {
       "epoch": 0.1332341823139369,
-      "grad_norm": 0.2795594584675045,
+      "grad_norm": 0.1340969880273331,
       "learning_rate": 0.0001993284786863948,
-      "loss": 3.7563,
+      "loss": 1.8687,
       "step": 1075
     },
     {
       "epoch": 0.13385387618516453,
-      "grad_norm": 0.26410153485873783,
+      "grad_norm": 0.13321869788563462,
       "learning_rate": 0.0001993032175927077,
-      "loss": 3.643,
+      "loss": 1.9355,
       "step": 1080
     },
     {
       "epoch": 0.13447357005639216,
-      "grad_norm": 0.28530160898231344,
+      "grad_norm": 0.1312621201180442,
       "learning_rate": 0.00019927749175904403,
-      "loss": 3.6891,
+      "loss": 1.8875,
       "step": 1085
     },
     {
       "epoch": 0.13509326392761975,
-      "grad_norm": 0.2540957405520998,
+      "grad_norm": 0.1336541528152664,
       "learning_rate": 0.0001992513013058009,
-      "loss": 3.743,
+      "loss": 1.8848,
       "step": 1090
     },
     {
       "epoch": 0.13571295779884737,
-      "grad_norm": 0.2490743424896684,
+      "grad_norm": 0.13090078215695478,
       "learning_rate": 0.00019922464635554988,
-      "loss": 3.7648,
+      "loss": 1.8434,
       "step": 1095
     },
     {
       "epoch": 0.136332651670075,
-      "grad_norm": 0.25315072498030244,
+      "grad_norm": 0.12341244251108834,
       "learning_rate": 0.0001991975270330364,
-      "loss": 3.7797,
+      "loss": 1.8719,
       "step": 1100
     },
     {
       "epoch": 0.1369523455413026,
-      "grad_norm": 0.26466186854297175,
+      "grad_norm": 0.13591893507471717,
       "learning_rate": 0.00019916994346517915,
-      "loss": 3.6898,
+      "loss": 1.8703,
       "step": 1105
     },
     {
       "epoch": 0.1375720394125302,
-      "grad_norm": 0.2615425377093863,
+      "grad_norm": 0.1288119831415113,
       "learning_rate": 0.00019914189578106945,
-      "loss": 3.9242,
+      "loss": 1.9027,
       "step": 1110
     },
     {
       "epoch": 0.13819173328375783,
-      "grad_norm": 0.26269293698244284,
+      "grad_norm": 0.1306962730366003,
       "learning_rate": 0.00019911338411197075,
-      "loss": 3.6852,
+      "loss": 1.9148,
       "step": 1115
     },
     {
       "epoch": 0.13881142715498543,
-      "grad_norm": 0.250180142337476,
+      "grad_norm": 0.1293793599928549,
       "learning_rate": 0.0001990844085913179,
-      "loss": 3.8563,
+      "loss": 1.8297,
       "step": 1120
     },
     {
       "epoch": 0.13943112102621305,
-      "grad_norm": 0.2541871059004602,
+      "grad_norm": 0.12111724806408185,
       "learning_rate": 0.00019905496935471658,
-      "loss": 3.775,
+      "loss": 1.8953,
       "step": 1125
     },
     {
       "epoch": 0.14005081489744067,
-      "grad_norm": 0.2679123179600542,
+      "grad_norm": 0.13310075929391021,
       "learning_rate": 0.00019902506653994277,
-      "loss": 3.7586,
+      "loss": 1.8297,
       "step": 1130
     },
     {
       "epoch": 0.14067050876866827,
-      "grad_norm": 0.26249500264698533,
+      "grad_norm": 0.12631873949945813,
       "learning_rate": 0.00019899470028694185,
-      "loss": 3.7977,
+      "loss": 1.8629,
       "step": 1135
     },
     {
       "epoch": 0.1412902026398959,
-      "grad_norm": 0.2617063111629921,
+      "grad_norm": 0.1375008660650019,
       "learning_rate": 0.0001989638707378282,
-      "loss": 3.7687,
+      "loss": 1.9668,
       "step": 1140
     },
     {
       "epoch": 0.1419098965111235,
-      "grad_norm": 0.2458327484673384,
+      "grad_norm": 0.13120298623344726,
       "learning_rate": 0.0001989325780368844,
-      "loss": 3.7109,
+      "loss": 1.8828,
       "step": 1145
     },
     {
       "epoch": 0.1425295903823511,
-      "grad_norm": 0.2522314523920583,
+      "grad_norm": 0.1417726162795728,
       "learning_rate": 0.00019890082233056054,
-      "loss": 3.7414,
+      "loss": 1.9273,
       "step": 1150
     },
     {
       "epoch": 0.14314928425357873,
-      "grad_norm": 0.24581360314066542,
+      "grad_norm": 0.13606374663242826,
       "learning_rate": 0.00019886860376747362,
-      "loss": 3.8203,
+      "loss": 1.8746,
       "step": 1155
     },
     {
       "epoch": 0.14376897812480635,
-      "grad_norm": 0.25279607405695015,
+      "grad_norm": 0.12914991669138237,
       "learning_rate": 0.0001988359224984069,
-      "loss": 3.8406,
+      "loss": 1.9184,
       "step": 1160
     },
     {
       "epoch": 0.14438867199603395,
-      "grad_norm": 0.2567164750916268,
+      "grad_norm": 0.13097438822182214,
       "learning_rate": 0.0001988027786763089,
-      "loss": 3.732,
+      "loss": 1.8188,
       "step": 1165
     },
     {
       "epoch": 0.14500836586726157,
-      "grad_norm": 0.23997160090056055,
+      "grad_norm": 0.12995850441011195,
       "learning_rate": 0.0001987691724562931,
-      "loss": 3.8172,
+      "loss": 1.8746,
       "step": 1170
     },
     {
       "epoch": 0.1456280597384892,
-      "grad_norm": 0.2542739318371719,
+      "grad_norm": 0.13478414995635102,
       "learning_rate": 0.00019873510399563688,
-      "loss": 3.6367,
+      "loss": 1.9023,
       "step": 1175
     },
     {
       "epoch": 0.1462477536097168,
-      "grad_norm": 0.2531159941787502,
+      "grad_norm": 0.1280137392385234,
       "learning_rate": 0.00019870057345378097,
-      "loss": 3.8031,
+      "loss": 1.8648,
       "step": 1180
     },
     {
       "epoch": 0.1468674474809444,
-      "grad_norm": 0.2788637520889176,
+      "grad_norm": 0.12710511011056266,
       "learning_rate": 0.00019866558099232862,
-      "loss": 3.8148,
+      "loss": 1.8637,
       "step": 1185
     },
     {
       "epoch": 0.14748714135217203,
-      "grad_norm": 0.29860991131828724,
+      "grad_norm": 0.13534609625301908,
       "learning_rate": 0.00019863012677504485,
-      "loss": 3.8219,
+      "loss": 1.8734,
       "step": 1190
     },
     {
       "epoch": 0.14810683522339965,
-      "grad_norm": 0.2569903027660448,
+      "grad_norm": 0.13068732229954763,
       "learning_rate": 0.00019859421096785575,
-      "loss": 3.8258,
+      "loss": 1.8891,
       "step": 1195
     },
     {
       "epoch": 0.14872652909462725,
-      "grad_norm": 0.25592105637718865,
+      "grad_norm": 0.13802118838387212,
       "learning_rate": 0.00019855783373884763,
-      "loss": 3.8234,
+      "loss": 1.8449,
       "step": 1200
     },
     {
       "epoch": 0.14934622296585487,
-      "grad_norm": 0.28504647584615034,
+      "grad_norm": 0.13152850461786877,
       "learning_rate": 0.00019852099525826628,
-      "loss": 3.7945,
+      "loss": 1.9012,
       "step": 1205
     },
     {
       "epoch": 0.1499659168370825,
-      "grad_norm": 0.2865619721254542,
+      "grad_norm": 0.13162793815607204,
       "learning_rate": 0.00019848369569851608,
-      "loss": 3.8242,
+      "loss": 1.8387,
       "step": 1210
     },
     {
       "epoch": 0.15058561070831009,
-      "grad_norm": 0.2551591601723071,
+      "grad_norm": 0.12825559680706589,
       "learning_rate": 0.00019844593523415935,
-      "loss": 3.8141,
+      "loss": 1.9254,
       "step": 1215
     },
     {
       "epoch": 0.1512053045795377,
-      "grad_norm": 0.25507234107508003,
+      "grad_norm": 0.14271049535774913,
       "learning_rate": 0.00019840771404191538,
-      "loss": 3.6891,
+      "loss": 1.8801,
       "step": 1220
     },
     {
       "epoch": 0.15182499845076533,
-      "grad_norm": 0.24464355974846816,
+      "grad_norm": 0.12772835947385766,
       "learning_rate": 0.00019836903230065973,
-      "loss": 3.8875,
+      "loss": 1.9152,
       "step": 1225
     },
     {
       "epoch": 0.15244469232199293,
-      "grad_norm": 0.2628854899768638,
+      "grad_norm": 0.13346214143251017,
       "learning_rate": 0.0001983298901914233,
-      "loss": 3.8266,
+      "loss": 1.8984,
       "step": 1230
     },
     {
       "epoch": 0.15306438619322055,
-      "grad_norm": 0.2470165170093542,
+      "grad_norm": 0.13348957014420162,
       "learning_rate": 0.00019829028789739156,
-      "loss": 3.8023,
+      "loss": 1.884,
       "step": 1235
     },
     {
       "epoch": 0.15368408006444817,
-      "grad_norm": 0.2641621943386744,
+      "grad_norm": 0.13030548172513623,
       "learning_rate": 0.00019825022560390353,
-      "loss": 3.8789,
+      "loss": 1.8844,
       "step": 1240
     },
     {
       "epoch": 0.15430377393567576,
-      "grad_norm": 0.2730690120133877,
+      "grad_norm": 0.12325474630414128,
       "learning_rate": 0.00019820970349845117,
-      "loss": 3.7766,
+      "loss": 1.916,
       "step": 1245
     },
     {
       "epoch": 0.1549234678069034,
-      "grad_norm": 0.2590003477418892,
+      "grad_norm": 0.12371037361630992,
       "learning_rate": 0.0001981687217706783,
-      "loss": 3.6641,
+      "loss": 1.8988,
       "step": 1250
     },
     {
       "epoch": 0.155543161678131,
-      "grad_norm": 0.2616580067830838,
+      "grad_norm": 0.12514967560905205,
       "learning_rate": 0.0001981272806123798,
-      "loss": 3.7219,
+      "loss": 1.8445,
       "step": 1255
     },
     {
       "epoch": 0.1561628555493586,
-      "grad_norm": 0.2714185703992873,
+      "grad_norm": 0.13355835058145732,
       "learning_rate": 0.00019808538021750063,
-      "loss": 3.8219,
+      "loss": 1.8598,
       "step": 1260
     },
     {
       "epoch": 0.15678254942058623,
-      "grad_norm": 0.2502930025265547,
+      "grad_norm": 0.1302860636545541,
       "learning_rate": 0.00019804302078213506,
-      "loss": 3.7719,
+      "loss": 1.9062,
       "step": 1265
     },
     {
       "epoch": 0.15740224329181385,
-      "grad_norm": 0.26422875872349105,
+      "grad_norm": 0.12797901973525885,
       "learning_rate": 0.00019800020250452564,
-      "loss": 3.7344,
+      "loss": 1.9098,
       "step": 1270
     },
     {
       "epoch": 0.15802193716304147,
-      "grad_norm": 0.2692299665019243,
+      "grad_norm": 0.1308306448611844,
       "learning_rate": 0.00019795692558506232,
-      "loss": 3.6547,
+      "loss": 1.8914,
       "step": 1275
     },
     {
       "epoch": 0.15864163103426906,
-      "grad_norm": 0.2639854392728509,
+      "grad_norm": 0.12802955593426202,
       "learning_rate": 0.0001979131902262814,
-      "loss": 3.6492,
+      "loss": 1.8906,
       "step": 1280
     },
     {
       "epoch": 0.1592613249054967,
-      "grad_norm": 0.27254984637905877,
+      "grad_norm": 0.12518672264585973,
       "learning_rate": 0.00019786899663286486,
-      "loss": 3.6633,
+      "loss": 1.9082,
       "step": 1285
     },
     {
       "epoch": 0.1598810187767243,
-      "grad_norm": 0.2513431426248905,
+      "grad_norm": 0.13434808762447917,
       "learning_rate": 0.00019782434501163909,
-      "loss": 3.7039,
+      "loss": 1.934,
       "step": 1290
     },
     {
       "epoch": 0.1605007126479519,
-      "grad_norm": 0.25083823338612543,
+      "grad_norm": 0.13038607351214834,
       "learning_rate": 0.0001977792355715741,
-      "loss": 3.875,
+      "loss": 1.9078,
       "step": 1295
     },
     {
       "epoch": 0.16112040651917953,
-      "grad_norm": 0.2384372233764081,
+      "grad_norm": 0.1276058968842769,
       "learning_rate": 0.00019773366852378246,
-      "loss": 3.7117,
+      "loss": 1.8352,
       "step": 1300
     },
     {
       "epoch": 0.16174010039040715,
-      "grad_norm": 0.2691834755767079,
+      "grad_norm": 0.141321358957705,
       "learning_rate": 0.0001976876440815184,
-      "loss": 3.693,
+      "loss": 1.8395,
       "step": 1305
     },
     {
       "epoch": 0.16235979426163474,
-      "grad_norm": 0.2644517005629403,
+      "grad_norm": 0.12259766843977997,
       "learning_rate": 0.0001976411624601767,
-      "loss": 3.8352,
+      "loss": 1.8359,
       "step": 1310
     },
     {
       "epoch": 0.16297948813286237,
-      "grad_norm": 0.2629094670093715,
+      "grad_norm": 0.12782259758660008,
       "learning_rate": 0.00019759422387729183,
-      "loss": 3.832,
+      "loss": 1.8418,
       "step": 1315
     },
     {
       "epoch": 0.16359918200409,
-      "grad_norm": 0.25827593256160586,
+      "grad_norm": 0.12829083288045331,
       "learning_rate": 0.00019754682855253674,
-      "loss": 3.7898,
+      "loss": 1.8805,
       "step": 1320
     },
     {
       "epoch": 0.16421887587531758,
-      "grad_norm": 0.258617601577871,
+      "grad_norm": 0.12446530921607402,
       "learning_rate": 0.00019749897670772205,
-      "loss": 3.7484,
+      "loss": 1.8352,
       "step": 1325
     },
     {
       "epoch": 0.1648385697465452,
-      "grad_norm": 0.2541932905242892,
+      "grad_norm": 0.12468008193096836,
       "learning_rate": 0.00019745066856679478,
-      "loss": 3.7734,
+      "loss": 1.8758,
       "step": 1330
     },
     {
       "epoch": 0.16545826361777283,
-      "grad_norm": 0.2513321092655162,
+      "grad_norm": 0.13195885649345943,
       "learning_rate": 0.0001974019043558375,
-      "loss": 3.6531,
+      "loss": 1.8508,
       "step": 1335
     },
     {
       "epoch": 0.16607795748900042,
-      "grad_norm": 0.28546456311685575,
+      "grad_norm": 0.12664766926925763,
       "learning_rate": 0.00019735268430306718,
-      "loss": 3.6289,
+      "loss": 1.8418,
       "step": 1340
     },
     {
       "epoch": 0.16669765136022804,
-      "grad_norm": 0.25516373996645875,
+      "grad_norm": 0.1332680932050387,
       "learning_rate": 0.00019730300863883405,
-      "loss": 3.8406,
+      "loss": 1.8695,
       "step": 1345
     },
     {
       "epoch": 0.16731734523145567,
-      "grad_norm": 0.23390699759641462,
+      "grad_norm": 0.12608647770348977,
       "learning_rate": 0.0001972528775956208,
-      "loss": 3.8266,
+      "loss": 1.8973,
       "step": 1350
     },
     {
       "epoch": 0.16793703910268326,
-      "grad_norm": 0.2557994609470224,
+      "grad_norm": 0.1316408940066455,
       "learning_rate": 0.0001972022914080411,
-      "loss": 3.8008,
+      "loss": 1.8328,
       "step": 1355
     },
     {
       "epoch": 0.16855673297391088,
-      "grad_norm": 0.25385825342982604,
+      "grad_norm": 0.12013595809919007,
       "learning_rate": 0.00019715125031283877,
-      "loss": 3.7539,
+      "loss": 1.8621,
       "step": 1360
     },
     {
       "epoch": 0.1691764268451385,
-      "grad_norm": 0.2644407002872055,
+      "grad_norm": 0.13491567947036534,
       "learning_rate": 0.00019709975454888662,
-      "loss": 3.7758,
+      "loss": 1.841,
       "step": 1365
     },
     {
       "epoch": 0.16979612071636613,
-      "grad_norm": 0.2460714641028012,
+      "grad_norm": 0.13054236047944698,
       "learning_rate": 0.00019704780435718532,
-      "loss": 3.7328,
+      "loss": 1.8957,
       "step": 1370
     },
     {
       "epoch": 0.17041581458759372,
-      "grad_norm": 0.2533400380158052,
+      "grad_norm": 0.13274138101276614,
       "learning_rate": 0.00019699539998086223,
-      "loss": 3.8414,
+      "loss": 1.8496,
       "step": 1375
     },
     {
       "epoch": 0.17103550845882134,
-      "grad_norm": 0.27880303783152216,
+      "grad_norm": 0.12178156573489422,
       "learning_rate": 0.00019694254166517032,
-      "loss": 3.7539,
+      "loss": 1.8836,
       "step": 1380
     },
     {
       "epoch": 0.17165520233004897,
-      "grad_norm": 0.25638461089130843,
+      "grad_norm": 0.13438311435319733,
       "learning_rate": 0.00019688922965748696,
-      "loss": 3.75,
+      "loss": 1.898,
       "step": 1385
     },
     {
       "epoch": 0.17227489620127656,
-      "grad_norm": 0.2702151632755114,
+      "grad_norm": 0.13378924908569717,
       "learning_rate": 0.0001968354642073129,
-      "loss": 3.6875,
+      "loss": 1.8383,
       "step": 1390
     },
     {
       "epoch": 0.17289459007250418,
-      "grad_norm": 0.24814604099538068,
+      "grad_norm": 0.1307401060574418,
       "learning_rate": 0.00019678124556627094,
-      "loss": 3.6586,
+      "loss": 1.8445,
       "step": 1395
     },
     {
       "epoch": 0.1735142839437318,
-      "grad_norm": 0.2535122565803045,
+      "grad_norm": 0.13793109916885396,
       "learning_rate": 0.00019672657398810478,
-      "loss": 3.6727,
+      "loss": 1.8559,
       "step": 1400
     },
     {
       "epoch": 0.1741339778149594,
-      "grad_norm": 0.25247054392702717,
+      "grad_norm": 0.13413192708976615,
       "learning_rate": 0.00019667144972867795,
-      "loss": 3.7258,
+      "loss": 1.8781,
       "step": 1405
     },
     {
       "epoch": 0.17475367168618702,
-      "grad_norm": 0.266294228324591,
+      "grad_norm": 0.13537690686486675,
       "learning_rate": 0.00019661587304597243,
-      "loss": 3.7164,
+      "loss": 1.8824,
       "step": 1410
     },
     {
       "epoch": 0.17537336555741465,
-      "grad_norm": 0.25031838105186166,
+      "grad_norm": 0.13031507569408332,
       "learning_rate": 0.0001965598442000877,
-      "loss": 3.7711,
+      "loss": 1.9043,
       "step": 1415
     },
     {
       "epoch": 0.17599305942864224,
-      "grad_norm": 0.2504169373702073,
+      "grad_norm": 0.13469238598709168,
       "learning_rate": 0.0001965033634532392,
-      "loss": 3.8008,
+      "loss": 1.8605,
       "step": 1420
     },
     {
       "epoch": 0.17661275329986986,
-      "grad_norm": 0.26471381309079295,
+      "grad_norm": 0.12447049113871386,
       "learning_rate": 0.00019644643106975739,
-      "loss": 3.657,
+      "loss": 1.8254,
       "step": 1425
     },
     {
       "epoch": 0.17723244717109748,
-      "grad_norm": 0.26233759886744334,
+      "grad_norm": 0.13493098167599366,
       "learning_rate": 0.00019638904731608637,
-      "loss": 3.7328,
+      "loss": 1.9117,
       "step": 1430
     },
     {
       "epoch": 0.17785214104232508,
-      "grad_norm": 0.25917034112988885,
+      "grad_norm": 0.12556872950040296,
       "learning_rate": 0.00019633121246078256,
-      "loss": 3.8891,
+      "loss": 1.8539,
       "step": 1435
     },
     {
       "epoch": 0.1784718349135527,
-      "grad_norm": 0.25024559107597927,
+      "grad_norm": 0.12732770515159572,
       "learning_rate": 0.00019627292677451368,
-      "loss": 3.8781,
+      "loss": 1.8891,
       "step": 1440
     },
     {
       "epoch": 0.17909152878478032,
-      "grad_norm": 0.2657054391608054,
+      "grad_norm": 0.13636272478996864,
       "learning_rate": 0.00019621419053005726,
-      "loss": 3.7422,
+      "loss": 1.882,
       "step": 1445
     },
     {
       "epoch": 0.17971122265600792,
-      "grad_norm": 0.26565528290139706,
+      "grad_norm": 0.126334909805653,
       "learning_rate": 0.00019615500400229946,
-      "loss": 3.7602,
+      "loss": 1.8887,
       "step": 1450
     },
     {
       "epoch": 0.18033091652723554,
-      "grad_norm": 0.25397414953791075,
+      "grad_norm": 0.13476880789699827,
       "learning_rate": 0.0001960953674682338,
-      "loss": 3.8078,
+      "loss": 1.8449,
       "step": 1455
     },
     {
       "epoch": 0.18095061039846316,
-      "grad_norm": 0.2586188121736365,
+      "grad_norm": 0.12610772664657266,
       "learning_rate": 0.00019603528120695982,
-      "loss": 3.9047,
+      "loss": 1.9023,
       "step": 1460
     },
     {
       "epoch": 0.18157030426969079,
-      "grad_norm": 0.28307612786781294,
+      "grad_norm": 0.12956314166827756,
       "learning_rate": 0.00019597474549968173,
-      "loss": 3.7742,
+      "loss": 1.8555,
       "step": 1465
     },
     {
       "epoch": 0.18218999814091838,
-      "grad_norm": 0.25496358354507603,
+      "grad_norm": 0.11672761587595178,
       "learning_rate": 0.00019591376062970728,
-      "loss": 3.7016,
+      "loss": 1.952,
       "step": 1470
     },
     {
       "epoch": 0.182809692012146,
-      "grad_norm": 0.2531796270841028,
+      "grad_norm": 0.1292512982323371,
       "learning_rate": 0.00019585232688244613,
-      "loss": 3.6437,
+      "loss": 1.8727,
       "step": 1475
     },
     {
       "epoch": 0.18342938588337362,
-      "grad_norm": 0.25281430577583736,
+      "grad_norm": 0.1309312239173881,
       "learning_rate": 0.00019579044454540883,
-      "loss": 3.7797,
+      "loss": 1.8344,
       "step": 1480
     },
     {
       "epoch": 0.18404907975460122,
-      "grad_norm": 0.24827004047230228,
+      "grad_norm": 0.13366728300584968,
       "learning_rate": 0.0001957281139082053,
-      "loss": 3.8031,
+      "loss": 1.9125,
       "step": 1485
     },
     {
       "epoch": 0.18466877362582884,
-      "grad_norm": 0.2533494089035691,
+      "grad_norm": 0.12537903174842133,
       "learning_rate": 0.00019566533526254348,
-      "loss": 3.6648,
+      "loss": 1.8684,
       "step": 1490
     },
     {
       "epoch": 0.18528846749705646,
-      "grad_norm": 0.25713999090986667,
+      "grad_norm": 0.12392425386470547,
       "learning_rate": 0.00019560210890222802,
-      "loss": 3.6875,
+      "loss": 1.8906,
       "step": 1495
     },
     {
       "epoch": 0.18590816136828406,
-      "grad_norm": 0.26243921456103164,
+      "grad_norm": 0.12391896609740323,
       "learning_rate": 0.00019553843512315887,
-      "loss": 3.7484,
+      "loss": 1.9105,
       "step": 1500
     },
     {
       "epoch": 0.18652785523951168,
-      "grad_norm": 0.26616483181088857,
+      "grad_norm": 0.12818849494366744,
       "learning_rate": 0.00019547431422332992,
-      "loss": 3.6391,
+      "loss": 1.85,
       "step": 1505
     },
     {
       "epoch": 0.1871475491107393,
-      "grad_norm": 0.23959324829479736,
+      "grad_norm": 0.12671159771811846,
       "learning_rate": 0.00019540974650282756,
-      "loss": 3.7461,
+      "loss": 1.916,
       "step": 1510
     },
     {
       "epoch": 0.1877672429819669,
-      "grad_norm": 0.2470546763118829,
+      "grad_norm": 0.13333844500063538,
       "learning_rate": 0.0001953447322638293,
-      "loss": 3.725,
+      "loss": 1.8625,
       "step": 1515
     },
     {
       "epoch": 0.18838693685319452,
-      "grad_norm": 0.256423918618775,
+      "grad_norm": 0.13840145694731687,
       "learning_rate": 0.0001952792718106024,
-      "loss": 3.7383,
+      "loss": 1.8809,
       "step": 1520
     },
     {
       "epoch": 0.18900663072442214,
-      "grad_norm": 0.2654600928240678,
+      "grad_norm": 0.12778029272918287,
       "learning_rate": 0.00019521336544950238,
-      "loss": 3.8961,
+      "loss": 1.8668,
       "step": 1525
     },
     {
       "epoch": 0.18962632459564974,
-      "grad_norm": 0.26718843333620523,
+      "grad_norm": 0.12913540845829236,
       "learning_rate": 0.00019514701348897164,
-      "loss": 3.743,
+      "loss": 1.8062,
       "step": 1530
     },
     {
       "epoch": 0.19024601846687736,
-      "grad_norm": 0.25878122211171267,
+      "grad_norm": 0.12826357981537254,
       "learning_rate": 0.00019508021623953795,
-      "loss": 3.7523,
+      "loss": 1.918,
       "step": 1535
     },
     {
       "epoch": 0.19086571233810498,
-      "grad_norm": 0.2598862844221595,
+      "grad_norm": 0.1342517702736094,
       "learning_rate": 0.00019501297401381304,
-      "loss": 3.7656,
+      "loss": 1.8668,
       "step": 1540
     },
     {
       "epoch": 0.19148540620933258,
-      "grad_norm": 0.24434694229901105,
+      "grad_norm": 0.138832925243521,
       "learning_rate": 0.00019494528712649117,
-      "loss": 3.7805,
+      "loss": 1.9117,
       "step": 1545
     },
     {
       "epoch": 0.1921051000805602,
-      "grad_norm": 0.25796351085501806,
+      "grad_norm": 0.12937438208518143,
       "learning_rate": 0.0001948771558943476,
-      "loss": 3.7406,
+      "loss": 1.907,
       "step": 1550
     },
     {
       "epoch": 0.19272479395178782,
-      "grad_norm": 0.24856327668868097,
+      "grad_norm": 0.1222226036118704,
       "learning_rate": 0.00019480858063623715,
-      "loss": 3.8391,
+      "loss": 1.8738,
       "step": 1555
     },
     {
       "epoch": 0.19334448782301544,
-      "grad_norm": 0.2692860471247067,
+      "grad_norm": 0.13553806750268793,
       "learning_rate": 0.0001947395616730926,
-      "loss": 3.7516,
+      "loss": 1.9059,
       "step": 1560
     },
     {
       "epoch": 0.19396418169424304,
-      "grad_norm": 0.24956285830888045,
+      "grad_norm": 0.13213241501616083,
       "learning_rate": 0.00019467009932792336,
-      "loss": 3.725,
+      "loss": 1.8684,
       "step": 1565
     },
     {
       "epoch": 0.19458387556547066,
-      "grad_norm": 0.24405925694161748,
+      "grad_norm": 0.1277167096208847,
       "learning_rate": 0.00019460019392581387,
-      "loss": 3.8078,
+      "loss": 1.9062,
       "step": 1570
     },
     {
       "epoch": 0.19520356943669828,
-      "grad_norm": 0.24476026905778137,
+      "grad_norm": 0.1269571278814333,
       "learning_rate": 0.00019452984579392205,
-      "loss": 3.7578,
+      "loss": 1.8527,
       "step": 1575
     },
     {
       "epoch": 0.19582326330792588,
-      "grad_norm": 0.25941791216039106,
+      "grad_norm": 0.12728857278045366,
       "learning_rate": 0.0001944590552614778,
-      "loss": 3.8273,
+      "loss": 1.8555,
       "step": 1580
     },
     {
       "epoch": 0.1964429571791535,
-      "grad_norm": 0.2605888382676801,
+      "grad_norm": 0.1290005338315187,
       "learning_rate": 0.0001943878226597815,
-      "loss": 3.6281,
+      "loss": 1.8637,
       "step": 1585
     },
     {
       "epoch": 0.19706265105038112,
-      "grad_norm": 0.25389288675364113,
+      "grad_norm": 0.1338620592085664,
       "learning_rate": 0.0001943161483222023,
-      "loss": 3.6195,
+      "loss": 1.85,
       "step": 1590
     },
     {
       "epoch": 0.19768234492160872,
-      "grad_norm": 0.25255096270670657,
+      "grad_norm": 0.12707291639830157,
       "learning_rate": 0.00019424403258417683,
-      "loss": 3.8656,
+      "loss": 1.8746,
       "step": 1595
     },
     {
       "epoch": 0.19830203879283634,
-      "grad_norm": 0.2546545729930429,
+      "grad_norm": 0.13144019328164225,
       "learning_rate": 0.00019417147578320744,
-      "loss": 3.7852,
+      "loss": 1.8664,
       "step": 1600
     },
     {
       "epoch": 0.19892173266406396,
-      "grad_norm": 0.25448534902199255,
+      "grad_norm": 0.12826047192933435,
       "learning_rate": 0.00019409847825886054,
-      "loss": 3.6578,
+      "loss": 1.9234,
       "step": 1605
     },
     {
       "epoch": 0.19954142653529156,
-      "grad_norm": 0.2531732716191023,
+      "grad_norm": 0.1397378039359775,
       "learning_rate": 0.00019402504035276525,
-      "loss": 3.7539,
+      "loss": 1.8086,
       "step": 1610
     },
     {
       "epoch": 0.20016112040651918,
-      "grad_norm": 0.2629784531419189,
+      "grad_norm": 0.12570861948831283,
       "learning_rate": 0.00019395116240861172,
-      "loss": 3.7195,
+      "loss": 1.8324,
       "step": 1615
     },
     {
       "epoch": 0.2007808142777468,
-      "grad_norm": 0.25369503528632603,
+      "grad_norm": 0.12079023395005281,
       "learning_rate": 0.0001938768447721493,
-      "loss": 3.7469,
+      "loss": 1.8758,
       "step": 1620
     },
     {
       "epoch": 0.2014005081489744,
-      "grad_norm": 0.2645391294956752,
+      "grad_norm": 0.13715689189513838,
       "learning_rate": 0.00019380208779118532,
-      "loss": 3.6891,
+      "loss": 1.8438,
       "step": 1625
     },
     {
       "epoch": 0.20202020202020202,
-      "grad_norm": 0.249604741933017,
+      "grad_norm": 0.12784219253316514,
       "learning_rate": 0.00019372689181558307,
-      "loss": 3.6945,
+      "loss": 1.8254,
       "step": 1630
     },
     {
       "epoch": 0.20263989589142964,
-      "grad_norm": 0.2612227448118402,
+      "grad_norm": 0.13127637020426203,
       "learning_rate": 0.00019365125719726046,
-      "loss": 3.6328,
+      "loss": 1.8363,
       "step": 1635
     },
     {
       "epoch": 0.20325958976265726,
-      "grad_norm": 0.25850843906032656,
+      "grad_norm": 0.12390585547014671,
       "learning_rate": 0.00019357518429018815,
-      "loss": 3.8016,
+      "loss": 1.8984,
       "step": 1640
     },
     {
       "epoch": 0.20387928363388486,
-      "grad_norm": 0.2473382579306127,
+      "grad_norm": 0.1296901168065532,
       "learning_rate": 0.00019349867345038808,
-      "loss": 3.8117,
+      "loss": 1.8891,
       "step": 1645
     },
     {
       "epoch": 0.20449897750511248,
-      "grad_norm": 0.24637915773390445,
+      "grad_norm": 0.13009453466712564,
       "learning_rate": 0.0001934217250359317,
-      "loss": 3.7609,
+      "loss": 1.8418,
       "step": 1650
     },
     {
       "epoch": 0.2051186713763401,
-      "grad_norm": 0.3072935679669988,
+      "grad_norm": 0.12805361264829143,
       "learning_rate": 0.0001933443394069383,
-      "loss": 3.6883,
+      "loss": 1.8605,
       "step": 1655
     },
     {
       "epoch": 0.2057383652475677,
-      "grad_norm": 0.250656654674928,
+      "grad_norm": 0.12965700777684402,
       "learning_rate": 0.0001932665169255733,
-      "loss": 3.8211,
+      "loss": 1.8945,
       "step": 1660
     },
     {
       "epoch": 0.20635805911879532,
-      "grad_norm": 0.25258094915573104,
+      "grad_norm": 0.14176151018901473,
       "learning_rate": 0.00019318825795604667,
-      "loss": 3.6961,
+      "loss": 1.868,
       "step": 1665
     },
     {
       "epoch": 0.20697775299002294,
-      "grad_norm": 0.2593606925045606,
+      "grad_norm": 0.12783314302778317,
       "learning_rate": 0.00019310956286461108,
-      "loss": 3.8031,
+      "loss": 1.8461,
       "step": 1670
     },
     {
       "epoch": 0.20759744686125053,
-      "grad_norm": 0.25793981295340673,
+      "grad_norm": 0.12476969665972379,
       "learning_rate": 0.00019303043201956033,
-      "loss": 3.8773,
+      "loss": 1.8465,
       "step": 1675
     },
     {
       "epoch": 0.20821714073247816,
-      "grad_norm": 0.2475268229332171,
+      "grad_norm": 0.13447320778956978,
       "learning_rate": 0.00019295086579122748,
-      "loss": 3.6773,
+      "loss": 1.807,
       "step": 1680
     },
     {
       "epoch": 0.20883683460370578,
-      "grad_norm": 0.2576357399823311,
+      "grad_norm": 0.12224616952170654,
       "learning_rate": 0.0001928708645519832,
-      "loss": 3.7305,
+      "loss": 1.8797,
       "step": 1685
     },
     {
       "epoch": 0.20945652847493337,
-      "grad_norm": 0.27035332613564755,
+      "grad_norm": 0.14073036790617033,
       "learning_rate": 0.00019279042867623405,
-      "loss": 3.7625,
+      "loss": 1.8734,
       "step": 1690
     },
     {
       "epoch": 0.210076222346161,
-      "grad_norm": 0.25029387687847293,
+      "grad_norm": 0.12419689358583988,
       "learning_rate": 0.00019270955854042065,
-      "loss": 3.7711,
+      "loss": 1.8969,
       "step": 1695
     },
     {
       "epoch": 0.21069591621738862,
-      "grad_norm": 0.26762236137322387,
+      "grad_norm": 0.1384451198274959,
       "learning_rate": 0.00019262825452301603,
-      "loss": 3.8469,
+      "loss": 1.8703,
       "step": 1700
     },
     {
       "epoch": 0.2113156100886162,
-      "grad_norm": 0.24285364151377078,
+      "grad_norm": 0.13754390968700744,
       "learning_rate": 0.0001925465170045237,
-      "loss": 3.7234,
+      "loss": 1.8277,
       "step": 1705
     },
     {
       "epoch": 0.21193530395984383,
-      "grad_norm": 0.2506313769424627,
+      "grad_norm": 0.12544899126243475,
       "learning_rate": 0.00019246434636747603,
-      "loss": 3.8797,
+      "loss": 1.8746,
       "step": 1710
     },
     {
       "epoch": 0.21255499783107146,
-      "grad_norm": 0.2623791318366264,
+      "grad_norm": 0.1316314602696802,
       "learning_rate": 0.00019238174299643235,
-      "loss": 3.6961,
+      "loss": 1.8535,
       "step": 1715
     },
     {
       "epoch": 0.21317469170229905,
-      "grad_norm": 0.24286466702261822,
+      "grad_norm": 0.12691928860824186,
       "learning_rate": 0.00019229870727797716,
-      "loss": 3.7289,
+      "loss": 1.9141,
       "step": 1720
     },
     {
       "epoch": 0.21379438557352667,
-      "grad_norm": 0.269295012443464,
+      "grad_norm": 0.13203627480238206,
       "learning_rate": 0.00019221523960071847,
-      "loss": 3.7172,
+      "loss": 1.8492,
       "step": 1725
     },
     {
       "epoch": 0.2144140794447543,
-      "grad_norm": 0.24914306646362705,
+      "grad_norm": 0.12505426294779426,
       "learning_rate": 0.00019213134035528574,
-      "loss": 3.8211,
+      "loss": 1.8984,
       "step": 1730
     },
     {
       "epoch": 0.21503377331598192,
-      "grad_norm": 0.261309598468901,
+      "grad_norm": 0.12969984327318165,
       "learning_rate": 0.0001920470099343282,
-      "loss": 3.7805,
+      "loss": 1.8727,
       "step": 1735
     },
     {
       "epoch": 0.2156534671872095,
-      "grad_norm": 0.2633828071876572,
+      "grad_norm": 0.12717555043591958,
       "learning_rate": 0.000191962248732513,
-      "loss": 3.857,
+      "loss": 1.9051,
       "step": 1740
     },
     {
       "epoch": 0.21627316105843714,
-      "grad_norm": 0.25552537780686957,
+      "grad_norm": 0.12782790245258435,
       "learning_rate": 0.00019187705714652337,
-      "loss": 3.7883,
+      "loss": 1.8527,
       "step": 1745
     },
     {
       "epoch": 0.21689285492966476,
-      "grad_norm": 0.25873934840784607,
+      "grad_norm": 0.12718190088753764,
       "learning_rate": 0.00019179143557505676,
-      "loss": 3.8133,
+      "loss": 1.8527,
       "step": 1750
     },
     {
       "epoch": 0.21751254880089235,
-      "grad_norm": 0.25040643355771275,
+      "grad_norm": 0.12935442052963278,
       "learning_rate": 0.0001917053844188228,
-      "loss": 3.7898,
+      "loss": 1.8668,
       "step": 1755
     },
     {
       "epoch": 0.21813224267211997,
-      "grad_norm": 0.2385870198485536,
+      "grad_norm": 0.13483832081677283,
       "learning_rate": 0.0001916189040805418,
-      "loss": 3.6844,
+      "loss": 1.8332,
       "step": 1760
     },
     {
       "epoch": 0.2187519365433476,
-      "grad_norm": 0.24428707193063878,
+      "grad_norm": 0.1231867163781486,
       "learning_rate": 0.0001915319949649425,
-      "loss": 3.6703,
+      "loss": 1.8461,
       "step": 1765
     },
     {
       "epoch": 0.2193716304145752,
-      "grad_norm": 0.25981823425530814,
+      "grad_norm": 0.1296012114936124,
       "learning_rate": 0.00019144465747876038,
-      "loss": 3.7188,
+      "loss": 1.8633,
       "step": 1770
     },
     {
       "epoch": 0.21999132428580281,
-      "grad_norm": 0.24977675057584312,
+      "grad_norm": 0.1358765864573398,
       "learning_rate": 0.00019135689203073563,
-      "loss": 3.6984,
+      "loss": 1.8586,
       "step": 1775
     },
     {
       "epoch": 0.22061101815703044,
-      "grad_norm": 0.2595528653750561,
+      "grad_norm": 0.12635502496539475,
       "learning_rate": 0.00019126869903161146,
-      "loss": 3.7547,
+      "loss": 1.9148,
       "step": 1780
     },
     {
       "epoch": 0.22123071202825803,
-      "grad_norm": 0.26071155800209916,
+      "grad_norm": 0.1382566677032265,
       "learning_rate": 0.00019118007889413186,
-      "loss": 3.7578,
+      "loss": 1.8082,
       "step": 1785
     },
     {
       "epoch": 0.22185040589948565,
-      "grad_norm": 0.257088768747436,
+      "grad_norm": 0.13033747033534399,
       "learning_rate": 0.00019109103203303988,
-      "loss": 3.8156,
+      "loss": 1.8367,
       "step": 1790
     },
     {
       "epoch": 0.22247009977071328,
-      "grad_norm": 0.25375238741582434,
+      "grad_norm": 0.12701483391149915,
       "learning_rate": 0.00019100155886507566,
-      "loss": 3.6352,
+      "loss": 1.9246,
       "step": 1795
     },
     {
       "epoch": 0.22308979364194087,
-      "grad_norm": 0.26492351236454703,
+      "grad_norm": 0.1364067044564757,
       "learning_rate": 0.0001909116598089745,
-      "loss": 3.6828,
+      "loss": 1.8582,
       "step": 1800
     },
     {
       "epoch": 0.2237094875131685,
-      "grad_norm": 0.24876024736288194,
+      "grad_norm": 0.12282125656108547,
       "learning_rate": 0.00019082133528546476,
-      "loss": 3.5773,
+      "loss": 1.8738,
       "step": 1805
     },
     {
       "epoch": 0.22432918138439611,
-      "grad_norm": 0.2576625159577276,
+      "grad_norm": 0.1272071393926594,
       "learning_rate": 0.0001907305857172661,
-      "loss": 3.7227,
+      "loss": 1.884,
       "step": 1810
     },
     {
       "epoch": 0.2249488752556237,
-      "grad_norm": 0.287526982161955,
+      "grad_norm": 0.12322060913981069,
       "learning_rate": 0.00019063941152908727,
-      "loss": 3.7297,
+      "loss": 1.8641,
       "step": 1815
     },
     {
       "epoch": 0.22556856912685133,
-      "grad_norm": 0.26483091023621297,
+      "grad_norm": 0.1269763797987862,
       "learning_rate": 0.00019054781314762433,
-      "loss": 3.7305,
+      "loss": 1.8559,
       "step": 1820
     },
     {
       "epoch": 0.22618826299807895,
-      "grad_norm": 0.2595249577164607,
+      "grad_norm": 0.13038449210432226,
       "learning_rate": 0.0001904557910015586,
-      "loss": 3.7523,
+      "loss": 1.8762,
       "step": 1825
     },
     {
       "epoch": 0.22680795686930658,
-      "grad_norm": 0.2505670142517455,
+      "grad_norm": 0.12529384355483927,
       "learning_rate": 0.00019036334552155452,
-      "loss": 3.6516,
+      "loss": 1.8332,
       "step": 1830
     },
     {
       "epoch": 0.22742765074053417,
-      "grad_norm": 0.27127556440251427,
+      "grad_norm": 0.1262927530224639,
       "learning_rate": 0.00019027047714025784,
-      "loss": 3.7352,
+      "loss": 1.8523,
       "step": 1835
     },
     {
       "epoch": 0.2280473446117618,
-      "grad_norm": 0.25134431087469267,
+      "grad_norm": 0.13984844751042014,
       "learning_rate": 0.0001901771862922934,
-      "loss": 3.7844,
+      "loss": 1.8566,
       "step": 1840
     },
     {
       "epoch": 0.22866703848298942,
-      "grad_norm": 0.2460567514125104,
+      "grad_norm": 0.14759646104938404,
       "learning_rate": 0.00019008347341426324,
-      "loss": 3.6508,
+      "loss": 1.8492,
       "step": 1845
     },
     {
       "epoch": 0.229286732354217,
-      "grad_norm": 0.2629987393065304,
+      "grad_norm": 0.12762255487022825,
       "learning_rate": 0.0001899893389447445,
-      "loss": 3.7219,
+      "loss": 1.8805,
       "step": 1850
     },
     {
       "epoch": 0.22990642622544463,
-      "grad_norm": 0.27575106848753533,
+      "grad_norm": 0.13415887670392074,
       "learning_rate": 0.0001898947833242874,
-      "loss": 3.7836,
+      "loss": 1.8461,
       "step": 1855
     },
     {
       "epoch": 0.23052612009667225,
-      "grad_norm": 0.25662445782308285,
+      "grad_norm": 0.13815909675804716,
       "learning_rate": 0.00018979980699541308,
-      "loss": 3.6367,
+      "loss": 1.8609,
       "step": 1860
     },
     {
       "epoch": 0.23114581396789985,
-      "grad_norm": 0.2507993462493921,
+      "grad_norm": 0.1261638415507045,
       "learning_rate": 0.00018970441040261165,
-      "loss": 3.6445,
+      "loss": 1.8043,
       "step": 1865
     },
     {
       "epoch": 0.23176550783912747,
-      "grad_norm": 0.2676156725812598,
+      "grad_norm": 0.132226368455845,
       "learning_rate": 0.00018960859399234006,
-      "loss": 3.7938,
+      "loss": 1.8152,
       "step": 1870
     },
     {
       "epoch": 0.2323852017103551,
-      "grad_norm": 0.2577237299425109,
+      "grad_norm": 0.12702999428539313,
       "learning_rate": 0.00018951235821301995,
-      "loss": 3.7422,
+      "loss": 1.8914,
       "step": 1875
     },
     {
       "epoch": 0.2330048955815827,
-      "grad_norm": 0.26252136602344933,
+      "grad_norm": 0.13100989208545588,
       "learning_rate": 0.0001894157035150357,
-      "loss": 3.6305,
+      "loss": 1.8391,
       "step": 1880
     },
     {
       "epoch": 0.2336245894528103,
-      "grad_norm": 0.26133595577475316,
+      "grad_norm": 0.12474165165129887,
       "learning_rate": 0.00018931863035073217,
-      "loss": 3.6617,
+      "loss": 1.9211,
       "step": 1885
     },
     {
       "epoch": 0.23424428332403793,
-      "grad_norm": 0.262362699420388,
+      "grad_norm": 0.12400290821379777,
       "learning_rate": 0.00018922113917441269,
-      "loss": 3.7039,
+      "loss": 1.8531,
       "step": 1890
     },
     {
       "epoch": 0.23486397719526553,
-      "grad_norm": 0.2657160759884634,
+      "grad_norm": 0.1293696900950748,
       "learning_rate": 0.00018912323044233684,
-      "loss": 3.7758,
+      "loss": 1.8555,
       "step": 1895
     },
     {
       "epoch": 0.23548367106649315,
-      "grad_norm": 0.2524857167548615,
+      "grad_norm": 0.13775339083965105,
       "learning_rate": 0.00018902490461271843,
-      "loss": 3.8492,
+      "loss": 1.877,
       "step": 1900
     },
     {
       "epoch": 0.23610336493772077,
-      "grad_norm": 0.26495908971079557,
+      "grad_norm": 0.12991294611318904,
       "learning_rate": 0.00018892616214572319,
-      "loss": 3.6422,
+      "loss": 1.8926,
       "step": 1905
     },
     {
       "epoch": 0.23672305880894837,
-      "grad_norm": 0.2684330797115416,
+      "grad_norm": 0.1306998350118421,
       "learning_rate": 0.0001888270035034668,
-      "loss": 3.7172,
+      "loss": 1.8473,
       "step": 1910
     },
     {
       "epoch": 0.237342752680176,
-      "grad_norm": 0.2683759303138262,
+      "grad_norm": 0.12748036755001935,
       "learning_rate": 0.00018872742915001267,
-      "loss": 3.6922,
+      "loss": 1.8184,
       "step": 1915
     },
     {
       "epoch": 0.2379624465514036,
-      "grad_norm": 0.26240769269212194,
+      "grad_norm": 0.13129037118419162,
       "learning_rate": 0.00018862743955136966,
-      "loss": 3.7898,
+      "loss": 1.8344,
       "step": 1920
     },
     {
       "epoch": 0.23858214042263123,
-      "grad_norm": 0.26589301276971006,
+      "grad_norm": 0.13675780584711228,
       "learning_rate": 0.00018852703517549,
-      "loss": 3.6891,
+      "loss": 1.8785,
       "step": 1925
     },
     {
       "epoch": 0.23920183429385883,
-      "grad_norm": 0.2542964819622325,
+      "grad_norm": 0.13672370672678413,
       "learning_rate": 0.00018842621649226712,
-      "loss": 3.6898,
+      "loss": 1.8977,
       "step": 1930
     },
     {
       "epoch": 0.23982152816508645,
-      "grad_norm": 0.2526849705264363,
+      "grad_norm": 0.136994334104686,
       "learning_rate": 0.00018832498397353337,
-      "loss": 3.7172,
+      "loss": 1.9043,
       "step": 1935
     },
     {
       "epoch": 0.24044122203631407,
-      "grad_norm": 0.25106965146419435,
+      "grad_norm": 0.12404741020853537,
       "learning_rate": 0.0001882233380930579,
-      "loss": 3.8023,
+      "loss": 1.9023,
       "step": 1940
     },
     {
       "epoch": 0.24106091590754167,
-      "grad_norm": 0.2651428117203693,
+      "grad_norm": 0.14219684197325005,
       "learning_rate": 0.00018812127932654437,
-      "loss": 3.7484,
+      "loss": 1.8391,
       "step": 1945
     },
     {
       "epoch": 0.2416806097787693,
-      "grad_norm": 0.2760638561741223,
+      "grad_norm": 0.14076460578083125,
       "learning_rate": 0.00018801880815162873,
-      "loss": 3.6172,
+      "loss": 1.8676,
       "step": 1950
     },
     {
       "epoch": 0.2423003036499969,
-      "grad_norm": 0.2456224462085717,
+      "grad_norm": 0.12874788431477818,
       "learning_rate": 0.00018791592504787704,
-      "loss": 3.7297,
+      "loss": 1.8375,
       "step": 1955
     },
     {
       "epoch": 0.2429199975212245,
-      "grad_norm": 0.2494028881960688,
+      "grad_norm": 0.12553931062925705,
       "learning_rate": 0.00018781263049678318,
-      "loss": 3.6156,
+      "loss": 1.8508,
       "step": 1960
     },
     {
       "epoch": 0.24353969139245213,
-      "grad_norm": 0.26095765336800963,
+      "grad_norm": 0.12091941894161848,
       "learning_rate": 0.00018770892498176658,
-      "loss": 3.807,
+      "loss": 1.9434,
       "step": 1965
     },
     {
       "epoch": 0.24415938526367975,
-      "grad_norm": 0.2704204880074346,
+      "grad_norm": 0.1339190592545441,
       "learning_rate": 0.00018760480898817003,
-      "loss": 3.6828,
+      "loss": 1.8926,
       "step": 1970
     },
     {
       "epoch": 0.24477907913490735,
-      "grad_norm": 0.2801106765194449,
+      "grad_norm": 0.12907299618607065,
       "learning_rate": 0.00018750028300325733,
-      "loss": 3.6758,
+      "loss": 1.7914,
       "step": 1975
     },
     {
       "epoch": 0.24539877300613497,
-      "grad_norm": 0.2492976531159216,
+      "grad_norm": 0.1400210529224134,
       "learning_rate": 0.0001873953475162111,
-      "loss": 3.7031,
+      "loss": 1.8559,
       "step": 1980
     },
     {
       "epoch": 0.2460184668773626,
-      "grad_norm": 0.24369383072754122,
+      "grad_norm": 0.1313714701900259,
       "learning_rate": 0.00018729000301813032,
-      "loss": 3.7703,
+      "loss": 1.902,
       "step": 1985
     },
     {
       "epoch": 0.24663816074859019,
-      "grad_norm": 0.2634660714556242,
+      "grad_norm": 0.1374832083968447,
       "learning_rate": 0.00018718425000202826,
-      "loss": 3.6648,
+      "loss": 1.8305,
       "step": 1990
     },
     {
       "epoch": 0.2472578546198178,
-      "grad_norm": 0.2506439015505356,
+      "grad_norm": 0.1305538539148369,
       "learning_rate": 0.00018707808896283,
-      "loss": 3.6344,
+      "loss": 1.8934,
       "step": 1995
     },
     {
       "epoch": 0.24787754849104543,
-      "grad_norm": 0.26475470061946976,
+      "grad_norm": 0.12798042833527448,
       "learning_rate": 0.00018697152039737018,
-      "loss": 3.8516,
+      "loss": 1.798,
       "step": 2000
     },
     {
       "epoch": 0.24849724236227302,
-      "grad_norm": 0.2519925926674702,
+      "grad_norm": 0.1224972820994739,
       "learning_rate": 0.0001868645448043907,
-      "loss": 3.8375,
+      "loss": 1.8176,
       "step": 2005
     },
     {
       "epoch": 0.24911693623350065,
-      "grad_norm": 0.2525671283371183,
+      "grad_norm": 0.13156854661928893,
       "learning_rate": 0.00018675716268453827,
-      "loss": 3.7492,
+      "loss": 1.8309,
       "step": 2010
     },
     {
       "epoch": 0.24973663010472827,
-      "grad_norm": 0.2717739214225155,
+      "grad_norm": 0.1326350510944194,
       "learning_rate": 0.00018664937454036226,
-      "loss": 3.6961,
+      "loss": 1.8426,
       "step": 2015
     },
     {
       "epoch": 0.25035632397595586,
-      "grad_norm": 0.2652751934837347,
+      "grad_norm": 0.12724824879009583,
       "learning_rate": 0.0001865411808763122,
-      "loss": 3.7086,
+      "loss": 1.8633,
       "step": 2020
     },
     {
       "epoch": 0.2509760178471835,
-      "grad_norm": 0.2648807010430378,
+      "grad_norm": 0.13143276628129275,
       "learning_rate": 0.00018643258219873534,
-      "loss": 3.7617,
+      "loss": 1.8672,
       "step": 2025
     },
     {
       "epoch": 0.2515957117184111,
-      "grad_norm": 0.24785920340040016,
+      "grad_norm": 0.12055786233157745,
       "learning_rate": 0.00018632357901587456,
-      "loss": 3.7375,
+      "loss": 1.8992,
       "step": 2030
     },
     {
       "epoch": 0.25221540558963873,
-      "grad_norm": 0.25091594030972114,
+      "grad_norm": 0.1283496889470771,
       "learning_rate": 0.00018621417183786577,
-      "loss": 3.7328,
+      "loss": 1.8969,
       "step": 2035
     },
     {
       "epoch": 0.25283509946086635,
-      "grad_norm": 0.24568898688178742,
+      "grad_norm": 0.12479918550598595,
       "learning_rate": 0.00018610436117673555,
-      "loss": 3.7461,
+      "loss": 1.8383,
       "step": 2040
     },
     {
       "epoch": 0.2534547933320939,
-      "grad_norm": 0.2631910743510119,
+      "grad_norm": 0.12527718321183984,
       "learning_rate": 0.00018599414754639883,
-      "loss": 3.6859,
+      "loss": 1.8961,
       "step": 2045
     },
     {
       "epoch": 0.25407448720332154,
-      "grad_norm": 0.2579231850021295,
+      "grad_norm": 0.13056814750458098,
       "learning_rate": 0.00018588353146265643,
-      "loss": 3.7602,
+      "loss": 1.8562,
       "step": 2050
     },
     {
       "epoch": 0.25469418107454916,
-      "grad_norm": 0.2504974253689712,
+      "grad_norm": 0.12452625637363586,
       "learning_rate": 0.0001857725134431926,
-      "loss": 3.7172,
+      "loss": 1.8633,
       "step": 2055
     },
     {
       "epoch": 0.2553138749457768,
-      "grad_norm": 0.2560287024318882,
+      "grad_norm": 0.12684769404009308,
       "learning_rate": 0.00018566109400757272,
-      "loss": 3.7547,
+      "loss": 1.8727,
       "step": 2060
     },
     {
       "epoch": 0.2559335688170044,
-      "grad_norm": 0.2607957181258484,
+      "grad_norm": 0.1313216307636194,
       "learning_rate": 0.0001855492736772408,
-      "loss": 3.8188,
+      "loss": 1.8906,
       "step": 2065
     },
     {
       "epoch": 0.25655326268823203,
-      "grad_norm": 0.25419874060956393,
+      "grad_norm": 0.13639962397305716,
       "learning_rate": 0.00018543705297551698,
-      "loss": 3.7648,
+      "loss": 1.9059,
       "step": 2070
     },
     {
       "epoch": 0.25717295655945965,
-      "grad_norm": 0.2647249642923883,
+      "grad_norm": 0.13156534728467967,
       "learning_rate": 0.0001853244324275952,
-      "loss": 3.6305,
+      "loss": 1.8484,
       "step": 2075
     },
     {
       "epoch": 0.2577926504306872,
-      "grad_norm": 0.2560264250612117,
+      "grad_norm": 0.13081763644721794,
       "learning_rate": 0.00018521141256054067,
-      "loss": 3.7625,
+      "loss": 1.8695,
       "step": 2080
     },
     {
       "epoch": 0.25841234430191484,
-      "grad_norm": 0.2719021879374889,
+      "grad_norm": 0.12594234500358284,
       "learning_rate": 0.00018509799390328742,
-      "loss": 3.7797,
+      "loss": 1.8414,
       "step": 2085
     },
     {
       "epoch": 0.25903203817314246,
-      "grad_norm": 0.26428917216504066,
+      "grad_norm": 0.13042394051093623,
       "learning_rate": 0.00018498417698663584,
-      "loss": 3.7148,
+      "loss": 1.8094,
       "step": 2090
     },
     {
       "epoch": 0.2596517320443701,
-      "grad_norm": 0.2513232763080223,
+      "grad_norm": 0.1281810590087273,
       "learning_rate": 0.00018486996234325009,
-      "loss": 3.8078,
+      "loss": 1.8336,
       "step": 2095
     },
     {
       "epoch": 0.2602714259155977,
-      "grad_norm": 0.2614505704904821,
+      "grad_norm": 0.12569535944479185,
       "learning_rate": 0.00018475535050765577,
-      "loss": 3.6078,
+      "loss": 1.8434,
       "step": 2100
     },
     {
       "epoch": 0.26089111978682533,
-      "grad_norm": 0.27228462221200644,
+      "grad_norm": 0.12417048765178107,
       "learning_rate": 0.00018464034201623737,
-      "loss": 3.7547,
+      "loss": 1.8211,
       "step": 2105
     },
     {
       "epoch": 0.2615108136580529,
-      "grad_norm": 0.2505037237378973,
+      "grad_norm": 0.12883572085670414,
       "learning_rate": 0.00018452493740723567,
-      "loss": 3.7414,
+      "loss": 1.8805,
       "step": 2110
     },
     {
       "epoch": 0.2621305075292805,
-      "grad_norm": 0.270150970994609,
+      "grad_norm": 0.12862790707551727,
       "learning_rate": 0.0001844091372207453,
-      "loss": 3.7188,
+      "loss": 1.7957,
       "step": 2115
     },
     {
       "epoch": 0.26275020140050814,
-      "grad_norm": 0.2594189582815906,
+      "grad_norm": 0.12715104158430907,
       "learning_rate": 0.00018429294199871218,
-      "loss": 3.7172,
+      "loss": 1.8562,
       "step": 2120
     },
     {
       "epoch": 0.26336989527173577,
-      "grad_norm": 0.25810176583742167,
+      "grad_norm": 0.12629988077958154,
       "learning_rate": 0.0001841763522849311,
-      "loss": 3.7359,
+      "loss": 1.825,
       "step": 2125
     },
     {
       "epoch": 0.2639895891429634,
-      "grad_norm": 0.2650696314876113,
+      "grad_norm": 0.13297933463768885,
       "learning_rate": 0.00018405936862504293,
-      "loss": 3.693,
+      "loss": 1.8133,
       "step": 2130
     },
     {
       "epoch": 0.264609283014191,
-      "grad_norm": 0.25786214468577856,
+      "grad_norm": 0.12653209526655865,
       "learning_rate": 0.00018394199156653233,
-      "loss": 3.7633,
+      "loss": 1.8574,
       "step": 2135
     },
     {
       "epoch": 0.2652289768854186,
-      "grad_norm": 0.26529708459406043,
+      "grad_norm": 0.12202510651987598,
       "learning_rate": 0.00018382422165872498,
-      "loss": 3.8055,
+      "loss": 1.8496,
       "step": 2140
     },
     {
       "epoch": 0.2658486707566462,
-      "grad_norm": 0.25696309007901624,
+      "grad_norm": 0.1353666668217046,
       "learning_rate": 0.00018370605945278512,
-      "loss": 3.6859,
+      "loss": 1.8637,
       "step": 2145
     },
     {
       "epoch": 0.2664683646278738,
-      "grad_norm": 0.2503534337459447,
+      "grad_norm": 0.13805247188442749,
       "learning_rate": 0.00018358750550171303,
-      "loss": 3.7547,
+      "loss": 1.9406,
       "step": 2150
     },
     {
       "epoch": 0.26708805849910144,
-      "grad_norm": 0.24875491233752253,
+      "grad_norm": 0.1250178668969665,
       "learning_rate": 0.00018346856036034225,
-      "loss": 3.7773,
+      "loss": 1.8527,
       "step": 2155
     },
     {
       "epoch": 0.26770775237032907,
-      "grad_norm": 0.2533128310142299,
+      "grad_norm": 0.12679232416551073,
       "learning_rate": 0.0001833492245853371,
-      "loss": 3.7398,
+      "loss": 1.8559,
       "step": 2160
     },
     {
       "epoch": 0.2683274462415567,
-      "grad_norm": 0.2539163968924575,
+      "grad_norm": 0.1363953224589558,
       "learning_rate": 0.00018322949873519028,
-      "loss": 3.7344,
+      "loss": 1.8895,
       "step": 2165
     },
     {
       "epoch": 0.2689471401127843,
-      "grad_norm": 0.2596999484239305,
+      "grad_norm": 0.1316621505083288,
       "learning_rate": 0.00018310938337021967,
-      "loss": 3.5758,
+      "loss": 1.8863,
       "step": 2170
     },
     {
       "epoch": 0.2695668339840119,
-      "grad_norm": 0.2689073376598725,
+      "grad_norm": 0.13658584100967,
       "learning_rate": 0.00018298887905256642,
-      "loss": 3.6281,
+      "loss": 1.8426,
       "step": 2175
     },
     {
       "epoch": 0.2701865278552395,
-      "grad_norm": 0.2674798310731981,
+      "grad_norm": 0.1399720872648625,
       "learning_rate": 0.00018286798634619178,
-      "loss": 3.7422,
+      "loss": 1.8887,
       "step": 2180
     },
     {
       "epoch": 0.2708062217264671,
-      "grad_norm": 0.24878965043899778,
+      "grad_norm": 0.13009508757933128,
       "learning_rate": 0.0001827467058168748,
-      "loss": 3.6938,
+      "loss": 1.8621,
       "step": 2185
     },
     {
       "epoch": 0.27142591559769474,
-      "grad_norm": 0.2564005390096162,
+      "grad_norm": 0.12921741777040827,
       "learning_rate": 0.00018262503803220941,
-      "loss": 3.6469,
+      "loss": 1.9246,
       "step": 2190
     },
     {
       "epoch": 0.27204560946892237,
-      "grad_norm": 0.26875929187484876,
+      "grad_norm": 0.12712470891174327,
       "learning_rate": 0.00018250298356160203,
-      "loss": 3.7062,
+      "loss": 1.8152,
       "step": 2195
     },
     {
       "epoch": 0.27266530334015,
-      "grad_norm": 0.25702595886613516,
+      "grad_norm": 0.13590278878639858,
       "learning_rate": 0.00018238054297626868,
-      "loss": 3.7711,
+      "loss": 1.8336,
       "step": 2200
     },
     {
       "epoch": 0.27328499721137756,
-      "grad_norm": 0.26204309861599234,
+      "grad_norm": 0.1238425521745261,
       "learning_rate": 0.0001822577168492324,
-      "loss": 3.6945,
+      "loss": 1.8859,
       "step": 2205
     },
     {
       "epoch": 0.2739046910826052,
-      "grad_norm": 0.25048034699771604,
+      "grad_norm": 0.11850858633920164,
       "learning_rate": 0.00018213450575532068,
-      "loss": 3.7687,
+      "loss": 1.8977,
       "step": 2210
     },
     {
       "epoch": 0.2745243849538328,
-      "grad_norm": 0.25907241148512433,
+      "grad_norm": 0.12913707015807815,
       "learning_rate": 0.0001820109102711625,
-      "loss": 3.7227,
+      "loss": 1.8637,
       "step": 2215
     },
     {
       "epoch": 0.2751440788250604,
-      "grad_norm": 0.2542048392477533,
+      "grad_norm": 0.1275936025502557,
       "learning_rate": 0.00018188693097518589,
-      "loss": 3.7156,
+      "loss": 1.8895,
       "step": 2220
     },
     {
       "epoch": 0.27576377269628805,
-      "grad_norm": 0.25291052345796766,
+      "grad_norm": 0.1289979754489145,
       "learning_rate": 0.00018176256844761515,
-      "loss": 3.6727,
+      "loss": 1.857,
       "step": 2225
     },
     {
       "epoch": 0.27638346656751567,
-      "grad_norm": 0.26113718400524455,
+      "grad_norm": 0.12857925042981372,
       "learning_rate": 0.000181637823270468,
-      "loss": 3.693,
+      "loss": 1.909,
       "step": 2230
     },
     {
       "epoch": 0.27700316043874323,
-      "grad_norm": 0.2630654914651346,
+      "grad_norm": 0.13116917461488,
       "learning_rate": 0.00018151269602755305,
-      "loss": 3.832,
+      "loss": 1.8363,
       "step": 2235
     },
     {
       "epoch": 0.27762285430997086,
-      "grad_norm": 0.27022263122244966,
+      "grad_norm": 0.12114760763849447,
       "learning_rate": 0.00018138718730446694,
-      "loss": 3.7406,
+      "loss": 1.8156,
       "step": 2240
     },
     {
       "epoch": 0.2782425481811985,
-      "grad_norm": 0.27804099995875925,
+      "grad_norm": 0.12970449110445717,
       "learning_rate": 0.00018126129768859166,
-      "loss": 3.6273,
+      "loss": 1.8266,
       "step": 2245
     },
     {
       "epoch": 0.2788622420524261,
-      "grad_norm": 0.26155058708223855,
+      "grad_norm": 0.13107665022911177,
       "learning_rate": 0.0001811350277690918,
-      "loss": 3.8539,
+      "loss": 1.8668,
       "step": 2250
     },
     {
       "epoch": 0.2794819359236537,
-      "grad_norm": 0.2571035053961106,
+      "grad_norm": 0.13037179328112342,
       "learning_rate": 0.00018100837813691173,
-      "loss": 3.8227,
+      "loss": 1.8992,
       "step": 2255
     },
     {
       "epoch": 0.28010162979488135,
-      "grad_norm": 0.2532323009912883,
+      "grad_norm": 0.13261626647936872,
       "learning_rate": 0.00018088134938477285,
-      "loss": 3.6812,
+      "loss": 1.9145,
       "step": 2260
     },
     {
       "epoch": 0.28072132366610897,
-      "grad_norm": 0.2653512113079037,
+      "grad_norm": 0.12467462496231817,
       "learning_rate": 0.00018075394210717097,
-      "loss": 3.8602,
+      "loss": 1.9527,
       "step": 2265
     },
     {
       "epoch": 0.28134101753733654,
-      "grad_norm": 0.2572939728938908,
+      "grad_norm": 0.1302006394858805,
       "learning_rate": 0.0001806261569003733,
-      "loss": 3.7461,
+      "loss": 1.8801,
       "step": 2270
     },
     {
       "epoch": 0.28196071140856416,
-      "grad_norm": 0.25590770385874284,
+      "grad_norm": 0.12656138072020032,
       "learning_rate": 0.00018049799436241584,
-      "loss": 3.6844,
+      "loss": 1.8168,
       "step": 2275
     },
     {
       "epoch": 0.2825804052797918,
-      "grad_norm": 0.25039800307535715,
+      "grad_norm": 0.13340657198127945,
       "learning_rate": 0.00018036945509310035,
-      "loss": 3.7859,
+      "loss": 1.8332,
       "step": 2280
     },
     {
       "epoch": 0.2832000991510194,
-      "grad_norm": 0.2588818018107696,
+      "grad_norm": 0.14913861128980632,
       "learning_rate": 0.00018024053969399186,
-      "loss": 3.7945,
+      "loss": 1.8199,
       "step": 2285
     },
     {
       "epoch": 0.283819793022247,
-      "grad_norm": 0.2560361453547467,
+      "grad_norm": 0.1306108935364312,
       "learning_rate": 0.00018011124876841564,
-      "loss": 3.657,
+      "loss": 1.8574,
       "step": 2290
     },
     {
       "epoch": 0.28443948689347465,
-      "grad_norm": 0.25369507057586327,
+      "grad_norm": 0.13401325283879312,
       "learning_rate": 0.0001799815829214544,
-      "loss": 3.6859,
+      "loss": 1.8879,
       "step": 2295
     },
     {
       "epoch": 0.2850591807647022,
-      "grad_norm": 0.26821575452634666,
+      "grad_norm": 0.12714018883217135,
       "learning_rate": 0.00017985154275994546,
-      "loss": 3.9195,
+      "loss": 1.843,
       "step": 2300
     },
     {
       "epoch": 0.28567887463592984,
-      "grad_norm": 0.25981458312044664,
+      "grad_norm": 0.12920439535349587,
       "learning_rate": 0.00017972112889247808,
-      "loss": 3.7891,
+      "loss": 1.8574,
       "step": 2305
     },
     {
       "epoch": 0.28629856850715746,
-      "grad_norm": 0.26331362477647047,
+      "grad_norm": 0.13423426078686024,
       "learning_rate": 0.00017959034192939027,
-      "loss": 3.7188,
+      "loss": 1.7715,
       "step": 2310
     },
     {
       "epoch": 0.2869182623783851,
-      "grad_norm": 0.2722924028576296,
+      "grad_norm": 0.13865809582419814,
       "learning_rate": 0.0001794591824827663,
-      "loss": 3.6688,
+      "loss": 1.8223,
       "step": 2315
     },
     {
       "epoch": 0.2875379562496127,
-      "grad_norm": 0.2514795296292142,
+      "grad_norm": 0.13168701447577058,
       "learning_rate": 0.0001793276511664335,
-      "loss": 3.6875,
+      "loss": 1.8875,
       "step": 2320
     },
     {
       "epoch": 0.2881576501208403,
-      "grad_norm": 0.2484928821399613,
+      "grad_norm": 0.12997910837997945,
       "learning_rate": 0.00017919574859595977,
-      "loss": 3.6039,
+      "loss": 1.8531,
       "step": 2325
     },
     {
       "epoch": 0.2887773439920679,
-      "grad_norm": 0.2627066856031102,
+      "grad_norm": 0.1223672846252809,
       "learning_rate": 0.00017906347538865021,
-      "loss": 3.6539,
+      "loss": 1.8539,
       "step": 2330
     },
     {
       "epoch": 0.2893970378632955,
-      "grad_norm": 0.25896716308452516,
+      "grad_norm": 0.13376807374029479,
       "learning_rate": 0.00017893083216354477,
-      "loss": 3.8016,
+      "loss": 1.882,
       "step": 2335
     },
     {
       "epoch": 0.29001673173452314,
-      "grad_norm": 0.260088068114753,
+      "grad_norm": 0.12472196578748385,
       "learning_rate": 0.00017879781954141497,
-      "loss": 3.6742,
+      "loss": 1.823,
       "step": 2340
     },
     {
       "epoch": 0.29063642560575076,
-      "grad_norm": 0.254152692296712,
+      "grad_norm": 0.13189694188445558,
       "learning_rate": 0.00017866443814476107,
-      "loss": 3.6477,
+      "loss": 1.8527,
       "step": 2345
     },
     {
       "epoch": 0.2912561194769784,
-      "grad_norm": 0.24886258494332086,
+      "grad_norm": 0.13537569866831578,
       "learning_rate": 0.0001785306885978092,
-      "loss": 3.8352,
+      "loss": 1.8832,
       "step": 2350
     },
     {
       "epoch": 0.291875813348206,
-      "grad_norm": 0.24656789699910267,
+      "grad_norm": 0.13295767983771653,
       "learning_rate": 0.00017839657152650856,
-      "loss": 3.7727,
+      "loss": 1.8457,
       "step": 2355
     },
     {
       "epoch": 0.2924955072194336,
-      "grad_norm": 0.25800157226082343,
+      "grad_norm": 0.1282927097281891,
       "learning_rate": 0.00017826208755852827,
-      "loss": 3.7109,
+      "loss": 1.8816,
       "step": 2360
     },
     {
       "epoch": 0.2931152010906612,
-      "grad_norm": 0.2690360891332733,
+      "grad_norm": 0.13049913149183853,
       "learning_rate": 0.00017812723732325446,
-      "loss": 3.7773,
+      "loss": 1.8641,
       "step": 2365
     },
     {
       "epoch": 0.2937348949618888,
-      "grad_norm": 0.2463411380990516,
+      "grad_norm": 0.13433617093166675,
       "learning_rate": 0.00017799202145178758,
-      "loss": 3.8328,
+      "loss": 1.8414,
       "step": 2370
     },
     {
       "epoch": 0.29435458883311644,
-      "grad_norm": 0.27495830500412854,
+      "grad_norm": 0.12882036108794354,
       "learning_rate": 0.00017785644057693913,
-      "loss": 3.9031,
+      "loss": 1.8551,
       "step": 2375
     },
     {
       "epoch": 0.29497428270434406,
-      "grad_norm": 0.2558003349123726,
+      "grad_norm": 0.12432550105269342,
       "learning_rate": 0.0001777204953332288,
-      "loss": 3.6437,
+      "loss": 1.8871,
       "step": 2380
     },
     {
       "epoch": 0.2955939765755717,
-      "grad_norm": 0.2765345649233851,
+      "grad_norm": 0.13582464892038923,
       "learning_rate": 0.00017758418635688167,
-      "loss": 3.7281,
+      "loss": 1.9059,
       "step": 2385
     },
     {
       "epoch": 0.2962136704467993,
-      "grad_norm": 0.24566185406030525,
+      "grad_norm": 0.1272933073062253,
       "learning_rate": 0.00017744751428582496,
-      "loss": 3.6414,
+      "loss": 1.8855,
       "step": 2390
     },
     {
       "epoch": 0.29683336431802687,
-      "grad_norm": 0.2720845656807831,
+      "grad_norm": 0.13122540435689012,
       "learning_rate": 0.00017731047975968523,
-      "loss": 3.6953,
+      "loss": 1.8566,
       "step": 2395
     },
     {
       "epoch": 0.2974530581892545,
-      "grad_norm": 0.26681875281668055,
+      "grad_norm": 0.1303624882568514,
       "learning_rate": 0.00017717308341978538,
-      "loss": 3.718,
+      "loss": 1.8562,
       "step": 2400
     },
     {
       "epoch": 0.2980727520604821,
-      "grad_norm": 0.2730768460301598,
+      "grad_norm": 0.12353259441658039,
       "learning_rate": 0.00017703532590914147,
-      "loss": 3.6094,
+      "loss": 1.8684,
       "step": 2405
     },
     {
       "epoch": 0.29869244593170974,
-      "grad_norm": 0.25665099624288207,
+      "grad_norm": 0.13170341964131901,
       "learning_rate": 0.00017689720787245997,
-      "loss": 3.6328,
+      "loss": 1.923,
       "step": 2410
     },
     {
       "epoch": 0.29931213980293736,
-      "grad_norm": 0.2671341617363374,
+      "grad_norm": 0.12876096829831757,
       "learning_rate": 0.00017675872995613458,
-      "loss": 3.6383,
+      "loss": 1.8668,
       "step": 2415
     },
     {
       "epoch": 0.299931833674165,
-      "grad_norm": 0.2548902040563108,
+      "grad_norm": 0.12369005168320343,
       "learning_rate": 0.0001766198928082432,
-      "loss": 3.6313,
+      "loss": 1.8613,
       "step": 2420
     },
     {
       "epoch": 0.30055152754539255,
-      "grad_norm": 0.25534063192461004,
+      "grad_norm": 0.13131764221149608,
       "learning_rate": 0.00017648069707854497,
-      "loss": 3.6695,
+      "loss": 1.9008,
       "step": 2425
     },
     {
       "epoch": 0.30117122141662017,
-      "grad_norm": 0.262691153256819,
+      "grad_norm": 0.12451191294123531,
       "learning_rate": 0.0001763411434184772,
-      "loss": 3.7055,
+      "loss": 1.8531,
       "step": 2430
     },
     {
       "epoch": 0.3017909152878478,
-      "grad_norm": 0.27308998925615785,
+      "grad_norm": 0.13236533602259273,
       "learning_rate": 0.00017620123248115235,
-      "loss": 3.6695,
+      "loss": 1.8691,
       "step": 2435
     },
     {
       "epoch": 0.3024106091590754,
-      "grad_norm": 0.25813554283407286,
+      "grad_norm": 0.12496123040012845,
       "learning_rate": 0.0001760609649213548,
-      "loss": 3.7078,
+      "loss": 1.8895,
       "step": 2440
     },
     {
       "epoch": 0.30303030303030304,
-      "grad_norm": 0.26071668852667024,
+      "grad_norm": 0.12136460576867035,
       "learning_rate": 0.00017592034139553812,
-      "loss": 3.6914,
+      "loss": 1.9016,
       "step": 2445
     },
     {
       "epoch": 0.30364999690153066,
-      "grad_norm": 0.25305199056761896,
+      "grad_norm": 0.12521529870266043,
       "learning_rate": 0.00017577936256182167,
-      "loss": 3.7555,
+      "loss": 1.8586,
       "step": 2450
     },
     {
       "epoch": 0.3042696907727583,
-      "grad_norm": 0.26384333576048713,
+      "grad_norm": 0.13202218968445123,
       "learning_rate": 0.00017563802907998773,
-      "loss": 3.7148,
+      "loss": 1.8457,
       "step": 2455
     },
     {
       "epoch": 0.30488938464398585,
-      "grad_norm": 0.25628783851634435,
+      "grad_norm": 0.12735675597883855,
       "learning_rate": 0.00017549634161147823,
-      "loss": 3.6867,
+      "loss": 1.8461,
       "step": 2460
     },
     {
       "epoch": 0.3055090785152135,
-      "grad_norm": 0.2532902126189516,
+      "grad_norm": 0.12834901220735337,
       "learning_rate": 0.0001753543008193919,
-      "loss": 3.6672,
+      "loss": 1.8824,
       "step": 2465
     },
     {
       "epoch": 0.3061287723864411,
-      "grad_norm": 0.25920690025946386,
+      "grad_norm": 0.13076216762403509,
       "learning_rate": 0.00017521190736848096,
-      "loss": 3.7102,
+      "loss": 1.8941,
       "step": 2470
     },
     {
       "epoch": 0.3067484662576687,
-      "grad_norm": 0.26672975988767644,
+      "grad_norm": 0.12527567116147303,
       "learning_rate": 0.00017506916192514801,
-      "loss": 3.7539,
+      "loss": 1.8156,
       "step": 2475
     },
     {
       "epoch": 0.30736816012889634,
-      "grad_norm": 0.26966415309842773,
+      "grad_norm": 0.12718249741367704,
       "learning_rate": 0.0001749260651574431,
-      "loss": 3.725,
+      "loss": 1.8629,
       "step": 2480
     },
     {
       "epoch": 0.30798785400012396,
-      "grad_norm": 0.25823480092565315,
+      "grad_norm": 0.13306132546258587,
       "learning_rate": 0.00017478261773506043,
-      "loss": 3.7008,
+      "loss": 1.8305,
       "step": 2485
     },
     {
       "epoch": 0.30860754787135153,
-      "grad_norm": 0.26871657326044296,
+      "grad_norm": 0.1307987467060954,
       "learning_rate": 0.00017463882032933524,
-      "loss": 3.643,
+      "loss": 1.8238,
       "step": 2490
     },
     {
       "epoch": 0.30922724174257915,
-      "grad_norm": 0.2676289335607888,
+      "grad_norm": 0.1351453589760986,
       "learning_rate": 0.00017449467361324076,
-      "loss": 3.5414,
+      "loss": 1.8543,
       "step": 2495
     },
     {
       "epoch": 0.3098469356138068,
-      "grad_norm": 0.27029407180457116,
+      "grad_norm": 0.12307357041545904,
       "learning_rate": 0.0001743501782613849,
-      "loss": 3.6156,
+      "loss": 1.9164,
       "step": 2500
     },
     {
       "epoch": 0.3104666294850344,
-      "grad_norm": 0.25225214661766965,
+      "grad_norm": 0.13149377717893693,
       "learning_rate": 0.00017420533495000727,
-      "loss": 3.6461,
+      "loss": 1.8473,
       "step": 2505
     },
     {
       "epoch": 0.311086323356262,
-      "grad_norm": 0.2519061042368278,
+      "grad_norm": 0.12779775983346198,
       "learning_rate": 0.0001740601443569759,
-      "loss": 3.7109,
+      "loss": 1.8031,
       "step": 2510
     },
     {
       "epoch": 0.31170601722748964,
-      "grad_norm": 0.26008459941014955,
+      "grad_norm": 0.13210455876193677,
       "learning_rate": 0.0001739146071617841,
-      "loss": 3.7523,
+      "loss": 1.8637,
       "step": 2515
     },
     {
       "epoch": 0.3123257110987172,
-      "grad_norm": 0.25855244878903594,
+      "grad_norm": 0.12810367969961917,
       "learning_rate": 0.0001737687240455473,
-      "loss": 3.8008,
+      "loss": 1.8316,
       "step": 2520
     },
     {
       "epoch": 0.31294540496994483,
-      "grad_norm": 0.25573633147599,
+      "grad_norm": 0.12704595997205193,
       "learning_rate": 0.00017362249569099982,
-      "loss": 3.6852,
+      "loss": 1.8445,
       "step": 2525
     },
     {
       "epoch": 0.31356509884117245,
-      "grad_norm": 0.2613543224212082,
+      "grad_norm": 0.12919348193178157,
       "learning_rate": 0.00017347592278249175,
-      "loss": 3.7219,
+      "loss": 1.8586,
       "step": 2530
     },
     {
       "epoch": 0.3141847927124001,
-      "grad_norm": 0.26233818596949066,
+      "grad_norm": 0.13802571807791786,
       "learning_rate": 0.00017332900600598562,
-      "loss": 3.7789,
+      "loss": 1.8703,
       "step": 2535
     },
     {
       "epoch": 0.3148044865836277,
-      "grad_norm": 0.25769955374979814,
+      "grad_norm": 0.1353603976296967,
       "learning_rate": 0.00017318174604905327,
-      "loss": 3.5703,
+      "loss": 1.8824,
       "step": 2540
     },
     {
       "epoch": 0.3154241804548553,
-      "grad_norm": 0.2501072256047701,
+      "grad_norm": 0.12737035386801074,
       "learning_rate": 0.00017303414360087278,
-      "loss": 3.7758,
+      "loss": 1.8246,
       "step": 2545
     },
     {
       "epoch": 0.31604387432608294,
-      "grad_norm": 0.26789431822777315,
+      "grad_norm": 0.13290623068299737,
       "learning_rate": 0.00017288619935222486,
-      "loss": 3.8359,
+      "loss": 1.8527,
       "step": 2550
     },
     {
       "epoch": 0.3166635681973105,
-      "grad_norm": 0.2523306821617771,
+      "grad_norm": 0.1207102015079047,
       "learning_rate": 0.00017273791399548998,
-      "loss": 3.793,
+      "loss": 1.8414,
       "step": 2555
     },
     {
       "epoch": 0.31728326206853813,
-      "grad_norm": 0.29899788219343393,
+      "grad_norm": 0.13358773743845612,
       "learning_rate": 0.000172589288224645,
-      "loss": 3.7398,
+      "loss": 1.8574,
       "step": 2560
     },
     {
       "epoch": 0.31790295593976575,
-      "grad_norm": 0.27736503749696745,
+      "grad_norm": 0.1279169729154266,
       "learning_rate": 0.00017244032273525995,
-      "loss": 3.743,
+      "loss": 1.8414,
       "step": 2565
     },
     {
       "epoch": 0.3185226498109934,
-      "grad_norm": 0.2624525004569622,
+      "grad_norm": 0.12718429357614378,
       "learning_rate": 0.0001722910182244946,
-      "loss": 3.668,
+      "loss": 1.8523,
       "step": 2570
     },
     {
       "epoch": 0.319142343682221,
-      "grad_norm": 0.2754468952649189,
+      "grad_norm": 0.1288546174064208,
       "learning_rate": 0.00017214137539109552,
-      "loss": 3.6688,
+      "loss": 1.8805,
       "step": 2575
     },
     {
       "epoch": 0.3197620375534486,
-      "grad_norm": 0.27290988521036397,
+      "grad_norm": 0.1339746303542721,
       "learning_rate": 0.0001719913949353925,
-      "loss": 3.7594,
+      "loss": 1.8687,
       "step": 2580
     },
     {
       "epoch": 0.3203817314246762,
-      "grad_norm": 0.2663978562264804,
+      "grad_norm": 0.12550192433671525,
       "learning_rate": 0.0001718410775592955,
-      "loss": 3.6953,
+      "loss": 1.8957,
       "step": 2585
     },
     {
       "epoch": 0.3210014252959038,
-      "grad_norm": 0.251095453478846,
+      "grad_norm": 0.13266972541549396,
       "learning_rate": 0.00017169042396629117,
-      "loss": 3.7336,
+      "loss": 1.8492,
       "step": 2590
     },
     {
       "epoch": 0.32162111916713143,
-      "grad_norm": 0.2669971493186749,
+      "grad_norm": 0.14169225665911975,
       "learning_rate": 0.00017153943486143978,
-      "loss": 3.6742,
+      "loss": 1.8582,
       "step": 2595
     },
     {
       "epoch": 0.32224081303835905,
-      "grad_norm": 0.24850318565449725,
+      "grad_norm": 0.12830541584647112,
       "learning_rate": 0.00017138811095137175,
-      "loss": 3.7211,
+      "loss": 1.8621,
       "step": 2600
     },
     {
       "epoch": 0.3228605069095867,
-      "grad_norm": 0.2895715920171382,
+      "grad_norm": 0.13328089401058096,
       "learning_rate": 0.0001712364529442843,
-      "loss": 3.7852,
+      "loss": 1.8047,
       "step": 2605
     },
     {
       "epoch": 0.3234802007808143,
-      "grad_norm": 0.27180495114459347,
+      "grad_norm": 0.12824865870602767,
       "learning_rate": 0.00017108446154993838,
-      "loss": 3.6875,
+      "loss": 1.8648,
       "step": 2610
     },
     {
       "epoch": 0.32409989465204186,
-      "grad_norm": 0.2548966676504203,
+      "grad_norm": 0.13486957428477936,
       "learning_rate": 0.0001709321374796551,
-      "loss": 3.7875,
+      "loss": 1.8773,
       "step": 2615
     },
     {
       "epoch": 0.3247195885232695,
-      "grad_norm": 0.26288507772094294,
+      "grad_norm": 0.14478150049090685,
       "learning_rate": 0.00017077948144631248,
-      "loss": 3.6688,
+      "loss": 1.8574,
       "step": 2620
     },
     {
       "epoch": 0.3253392823944971,
-      "grad_norm": 0.26277070377953077,
+      "grad_norm": 0.13019704674061816,
       "learning_rate": 0.00017062649416434223,
-      "loss": 3.6469,
+      "loss": 1.8359,
       "step": 2625
     },
     {
       "epoch": 0.32595897626572473,
-      "grad_norm": 0.26365748542058276,
+      "grad_norm": 0.12921336608460762,
       "learning_rate": 0.00017047317634972617,
-      "loss": 3.6617,
+      "loss": 1.8598,
       "step": 2630
     },
     {
       "epoch": 0.32657867013695235,
-      "grad_norm": 0.2458080656469107,
+      "grad_norm": 0.12820404459063692,
       "learning_rate": 0.00017031952871999315,
-      "loss": 3.7359,
+      "loss": 1.925,
       "step": 2635
     },
     {
       "epoch": 0.32719836400818,
-      "grad_norm": 0.27576035893728557,
+      "grad_norm": 0.1333050138527027,
       "learning_rate": 0.0001701655519942155,
-      "loss": 3.7281,
+      "loss": 1.8805,
       "step": 2640
     },
     {
       "epoch": 0.3278180578794076,
-      "grad_norm": 0.2671236481432457,
+      "grad_norm": 0.12777262066895273,
       "learning_rate": 0.00017001124689300568,
-      "loss": 3.7727,
+      "loss": 1.8207,
       "step": 2645
     },
     {
       "epoch": 0.32843775175063517,
-      "grad_norm": 0.25646009969395855,
+      "grad_norm": 0.13548725964588937,
       "learning_rate": 0.00016985661413851304,
-      "loss": 3.6984,
+      "loss": 1.8512,
       "step": 2650
     },
     {
       "epoch": 0.3290574456218628,
-      "grad_norm": 0.24855468305555162,
+      "grad_norm": 0.1291350186134146,
       "learning_rate": 0.00016970165445442023,
-      "loss": 3.5922,
+      "loss": 1.8574,
       "step": 2655
     },
     {
       "epoch": 0.3296771394930904,
-      "grad_norm": 0.2488779303959064,
+      "grad_norm": 0.13286049954171614,
       "learning_rate": 0.00016954636856594005,
-      "loss": 3.7273,
+      "loss": 1.8398,
       "step": 2660
     },
     {
       "epoch": 0.33029683336431803,
-      "grad_norm": 0.24991520834091216,
+      "grad_norm": 0.12250586047506855,
       "learning_rate": 0.00016939075719981194,
-      "loss": 3.6844,
+      "loss": 1.8598,
       "step": 2665
     },
     {
       "epoch": 0.33091652723554565,
-      "grad_norm": 0.27341253703778134,
+      "grad_norm": 0.12758848458031308,
       "learning_rate": 0.00016923482108429844,
-      "loss": 3.6289,
+      "loss": 1.8445,
       "step": 2670
     },
     {
       "epoch": 0.3315362211067733,
-      "grad_norm": 0.25347885179835555,
+      "grad_norm": 0.13664664659730347,
       "learning_rate": 0.00016907856094918207,
-      "loss": 3.7313,
+      "loss": 1.8059,
       "step": 2675
     },
     {
       "epoch": 0.33215591497800084,
-      "grad_norm": 0.2664388675272623,
+      "grad_norm": 0.12746464026705376,
       "learning_rate": 0.0001689219775257617,
-      "loss": 3.7398,
+      "loss": 1.8188,
       "step": 2680
     },
     {
       "epoch": 0.33277560884922847,
-      "grad_norm": 0.26965614612569255,
+      "grad_norm": 0.12755329667300946,
       "learning_rate": 0.00016876507154684918,
-      "loss": 3.6805,
+      "loss": 1.8035,
       "step": 2685
     },
     {
       "epoch": 0.3333953027204561,
-      "grad_norm": 0.25477336727357236,
+      "grad_norm": 0.1285133337473147,
       "learning_rate": 0.00016860784374676593,
-      "loss": 3.7359,
+      "loss": 1.8609,
       "step": 2690
     },
     {
       "epoch": 0.3340149965916837,
-      "grad_norm": 0.26828594917552684,
+      "grad_norm": 0.1313491570500189,
       "learning_rate": 0.00016845029486133956,
-      "loss": 3.6789,
+      "loss": 1.8512,
       "step": 2695
     },
     {
       "epoch": 0.33463469046291133,
-      "grad_norm": 0.28043441092000476,
+      "grad_norm": 0.1258582951786215,
       "learning_rate": 0.00016829242562790026,
-      "loss": 3.7039,
+      "loss": 1.8379,
       "step": 2700
     },
     {
       "epoch": 0.33525438433413896,
-      "grad_norm": 0.2719579591511516,
+      "grad_norm": 0.12654293645403783,
       "learning_rate": 0.00016813423678527754,
-      "loss": 3.7852,
+      "loss": 1.9211,
       "step": 2705
     },
     {
       "epoch": 0.3358740782053665,
-      "grad_norm": 0.25604408210266827,
+      "grad_norm": 0.1430960515630514,
       "learning_rate": 0.00016797572907379667,
-      "loss": 3.6742,
+      "loss": 1.848,
       "step": 2710
     },
     {
       "epoch": 0.33649377207659414,
-      "grad_norm": 0.26206590705100935,
+      "grad_norm": 0.13294562149445674,
       "learning_rate": 0.00016781690323527511,
-      "loss": 3.8953,
+      "loss": 1.8676,
       "step": 2715
     },
     {
       "epoch": 0.33711346594782177,
-      "grad_norm": 0.26396741494967757,
+      "grad_norm": 0.12823331423568302,
       "learning_rate": 0.00016765776001301933,
-      "loss": 3.7289,
+      "loss": 1.8555,
       "step": 2720
     },
     {
       "epoch": 0.3377331598190494,
-      "grad_norm": 0.2547598811513136,
+      "grad_norm": 0.13133147708443205,
       "learning_rate": 0.00016749830015182107,
-      "loss": 3.6258,
+      "loss": 1.9402,
       "step": 2725
     },
     {
       "epoch": 0.338352853690277,
-      "grad_norm": 0.2741006690732575,
+      "grad_norm": 0.1307417640679184,
       "learning_rate": 0.00016733852439795394,
-      "loss": 3.7898,
+      "loss": 1.8723,
       "step": 2730
     },
     {
       "epoch": 0.33897254756150463,
-      "grad_norm": 0.2641777459426088,
+      "grad_norm": 0.14036159913262772,
       "learning_rate": 0.00016717843349916994,
-      "loss": 3.7031,
+      "loss": 1.8359,
       "step": 2735
     },
     {
       "epoch": 0.33959224143273226,
-      "grad_norm": 0.2629606866295394,
+      "grad_norm": 0.1275526648144705,
       "learning_rate": 0.000167018028204696,
-      "loss": 3.8133,
+      "loss": 1.8734,
       "step": 2740
     },
     {
       "epoch": 0.3402119353039598,
-      "grad_norm": 0.24477609948639015,
+      "grad_norm": 0.12574598292314898,
       "learning_rate": 0.0001668573092652303,
-      "loss": 3.7102,
+      "loss": 1.8637,
       "step": 2745
     },
     {
       "epoch": 0.34083162917518744,
-      "grad_norm": 0.27068594003654395,
+      "grad_norm": 0.13622787470190775,
       "learning_rate": 0.00016669627743293907,
-      "loss": 3.682,
+      "loss": 1.9211,
       "step": 2750
     },
     {
       "epoch": 0.34145132304641507,
-      "grad_norm": 0.26929411586592,
+      "grad_norm": 0.13437389620413182,
       "learning_rate": 0.00016653493346145267,
-      "loss": 3.6445,
+      "loss": 1.8262,
       "step": 2755
     },
     {
       "epoch": 0.3420710169176427,
-      "grad_norm": 0.24939469621006297,
+      "grad_norm": 0.1344605856943175,
       "learning_rate": 0.00016637327810586246,
-      "loss": 3.7125,
+      "loss": 1.8367,
       "step": 2760
     },
     {
       "epoch": 0.3426907107888703,
-      "grad_norm": 0.2530396996711045,
+      "grad_norm": 0.1291930553452629,
       "learning_rate": 0.00016621131212271695,
-      "loss": 3.6828,
+      "loss": 1.8316,
       "step": 2765
     },
     {
       "epoch": 0.34331040466009793,
-      "grad_norm": 0.25826979485096646,
+      "grad_norm": 0.1343538598231125,
       "learning_rate": 0.00016604903627001844,
-      "loss": 3.718,
+      "loss": 1.8586,
       "step": 2770
     },
     {
       "epoch": 0.3439300985313255,
-      "grad_norm": 0.27768697587316615,
+      "grad_norm": 0.13323136800247856,
       "learning_rate": 0.00016588645130721948,
-      "loss": 3.618,
+      "loss": 1.9086,
       "step": 2775
     },
     {
       "epoch": 0.3445497924025531,
-      "grad_norm": 0.2650900726296455,
+      "grad_norm": 0.12809181295437982,
       "learning_rate": 0.00016572355799521912,
-      "loss": 3.7219,
+      "loss": 1.8621,
       "step": 2780
     },
     {
       "epoch": 0.34516948627378075,
-      "grad_norm": 0.2633177048203003,
+      "grad_norm": 0.13373391040989493,
       "learning_rate": 0.0001655603570963596,
-      "loss": 3.718,
+      "loss": 1.8402,
       "step": 2785
     },
     {
       "epoch": 0.34578918014500837,
-      "grad_norm": 0.26296181136909086,
+      "grad_norm": 0.12823175261000916,
       "learning_rate": 0.00016539684937442263,
-      "loss": 3.7039,
+      "loss": 1.8855,
       "step": 2790
     },
     {
       "epoch": 0.346408874016236,
-      "grad_norm": 0.2783942226825099,
+      "grad_norm": 0.12427006053911512,
       "learning_rate": 0.0001652330355946259,
-      "loss": 3.6758,
+      "loss": 1.8035,
       "step": 2795
     },
     {
       "epoch": 0.3470285678874636,
-      "grad_norm": 0.2600566703174294,
+      "grad_norm": 0.14309427079315348,
       "learning_rate": 0.00016506891652361933,
-      "loss": 3.6969,
+      "loss": 1.8367,
       "step": 2800
     },
     {
       "epoch": 0.3476482617586912,
-      "grad_norm": 0.2542741242950709,
+      "grad_norm": 0.1346672703530673,
       "learning_rate": 0.0001649044929294818,
-      "loss": 3.6078,
+      "loss": 1.8531,
       "step": 2805
     },
     {
       "epoch": 0.3482679556299188,
-      "grad_norm": 0.25822378785350575,
+      "grad_norm": 0.1348324178068194,
       "learning_rate": 0.00016473976558171714,
-      "loss": 3.7125,
+      "loss": 1.9059,
       "step": 2810
     },
     {
       "epoch": 0.3488876495011464,
-      "grad_norm": 0.2641370960337805,
+      "grad_norm": 0.13127286752655276,
       "learning_rate": 0.00016457473525125093,
-      "loss": 3.7141,
+      "loss": 1.8086,
       "step": 2815
     },
     {
       "epoch": 0.34950734337237405,
-      "grad_norm": 0.25569765031719666,
+      "grad_norm": 0.13288592496132326,
       "learning_rate": 0.00016440940271042663,
-      "loss": 3.643,
+      "loss": 1.8898,
       "step": 2820
     },
     {
       "epoch": 0.35012703724360167,
-      "grad_norm": 0.26808702545434854,
+      "grad_norm": 0.13569741783533862,
       "learning_rate": 0.00016424376873300207,
-      "loss": 3.6211,
+      "loss": 1.8484,
       "step": 2825
     },
     {
       "epoch": 0.3507467311148293,
-      "grad_norm": 0.2608273876865326,
+      "grad_norm": 0.13416342536609563,
       "learning_rate": 0.00016407783409414577,
-      "loss": 3.6508,
+      "loss": 1.8609,
       "step": 2830
     },
     {
       "epoch": 0.3513664249860569,
-      "grad_norm": 0.2479224348257157,
+      "grad_norm": 0.1169790399721275,
       "learning_rate": 0.00016391159957043335,
-      "loss": 3.6773,
+      "loss": 1.8531,
       "step": 2835
     },
     {
       "epoch": 0.3519861188572845,
-      "grad_norm": 0.27099013945725076,
+      "grad_norm": 0.13616470490760044,
       "learning_rate": 0.000163745065939844,
-      "loss": 3.775,
+      "loss": 1.8703,
       "step": 2840
     },
     {
       "epoch": 0.3526058127285121,
-      "grad_norm": 0.2678681599061846,
+      "grad_norm": 0.1253790357970052,
       "learning_rate": 0.0001635782339817566,
-      "loss": 3.7352,
+      "loss": 1.9043,
       "step": 2845
     },
     {
       "epoch": 0.3532255065997397,
-      "grad_norm": 0.25434553862487636,
+      "grad_norm": 0.12874313573044202,
       "learning_rate": 0.00016341110447694624,
-      "loss": 3.7531,
+      "loss": 1.8266,
       "step": 2850
     },
     {
       "epoch": 0.35384520047096735,
-      "grad_norm": 0.2600099088957585,
+      "grad_norm": 0.1424792287256125,
       "learning_rate": 0.00016324367820758057,
-      "loss": 3.7352,
+      "loss": 1.8273,
       "step": 2855
     },
     {
       "epoch": 0.35446489434219497,
-      "grad_norm": 0.2576820816746782,
+      "grad_norm": 0.1274395857428731,
       "learning_rate": 0.0001630759559572161,
-      "loss": 3.6844,
+      "loss": 1.8473,
       "step": 2860
     },
     {
       "epoch": 0.3550845882134226,
-      "grad_norm": 0.26429572988622685,
+      "grad_norm": 0.12980584268562773,
       "learning_rate": 0.00016290793851079447,
-      "loss": 3.6648,
+      "loss": 1.8859,
       "step": 2865
     },
     {
       "epoch": 0.35570428208465016,
-      "grad_norm": 0.25628056770143104,
+      "grad_norm": 0.1284857003389399,
       "learning_rate": 0.00016273962665463892,
-      "loss": 3.768,
+      "loss": 1.8883,
       "step": 2870
     },
     {
       "epoch": 0.3563239759558778,
-      "grad_norm": 0.26314053347734023,
+      "grad_norm": 0.13062335696884655,
       "learning_rate": 0.00016257102117645048,
-      "loss": 3.7867,
+      "loss": 1.8113,
       "step": 2875
     },
     {
       "epoch": 0.3569436698271054,
-      "grad_norm": 0.26827011010620455,
+      "grad_norm": 0.13020290307748666,
       "learning_rate": 0.00016240212286530432,
-      "loss": 3.5789,
+      "loss": 1.8633,
       "step": 2880
     },
     {
       "epoch": 0.357563363698333,
-      "grad_norm": 0.2717920774102298,
+      "grad_norm": 0.12282658467967292,
       "learning_rate": 0.00016223293251164616,
-      "loss": 3.7508,
+      "loss": 1.8703,
       "step": 2885
     },
     {
       "epoch": 0.35818305756956065,
-      "grad_norm": 0.2553465718103751,
+      "grad_norm": 0.12853916417645042,
       "learning_rate": 0.00016206345090728834,
-      "loss": 3.7437,
+      "loss": 1.8379,
       "step": 2890
     },
     {
       "epoch": 0.35880275144078827,
-      "grad_norm": 0.24526796804421827,
+      "grad_norm": 0.12994523896895013,
       "learning_rate": 0.00016189367884540638,
-      "loss": 3.6945,
+      "loss": 1.8469,
       "step": 2895
     },
     {
       "epoch": 0.35942244531201584,
-      "grad_norm": 0.2602330023370647,
+      "grad_norm": 0.13578425779672484,
       "learning_rate": 0.00016172361712053513,
-      "loss": 3.5922,
+      "loss": 1.8594,
       "step": 2900
     },
     {
       "epoch": 0.36004213918324346,
-      "grad_norm": 0.25584605115473547,
+      "grad_norm": 0.1264516379346521,
       "learning_rate": 0.00016155326652856497,
-      "loss": 3.775,
+      "loss": 1.8816,
       "step": 2905
     },
     {
       "epoch": 0.3606618330544711,
-      "grad_norm": 0.2679851558322471,
+      "grad_norm": 0.1285061053282456,
       "learning_rate": 0.0001613826278667383,
-      "loss": 3.7484,
+      "loss": 1.7895,
       "step": 2910
     },
     {
       "epoch": 0.3612815269256987,
-      "grad_norm": 0.2721594838633521,
+      "grad_norm": 0.13094212766467833,
       "learning_rate": 0.00016121170193364557,
-      "loss": 3.632,
+      "loss": 1.8172,
       "step": 2915
     },
     {
       "epoch": 0.3619012207969263,
-      "grad_norm": 0.365423422711038,
+      "grad_norm": 0.12276398604923604,
       "learning_rate": 0.0001610404895292218,
-      "loss": 3.7281,
+      "loss": 1.9117,
       "step": 2920
     },
     {
       "epoch": 0.36252091466815395,
-      "grad_norm": 0.2633704122091394,
+      "grad_norm": 0.12739854449016166,
       "learning_rate": 0.0001608689914547426,
-      "loss": 3.6812,
+      "loss": 1.8629,
       "step": 2925
     },
     {
       "epoch": 0.36314060853938157,
-      "grad_norm": 0.24858833944835435,
+      "grad_norm": 0.1292586487909269,
       "learning_rate": 0.00016069720851282052,
-      "loss": 3.6594,
+      "loss": 1.8477,
       "step": 2930
     },
     {
       "epoch": 0.36376030241060914,
-      "grad_norm": 0.2615686620373773,
+      "grad_norm": 0.12593917875384036,
       "learning_rate": 0.00016052514150740135,
-      "loss": 3.7031,
+      "loss": 1.8938,
       "step": 2935
     },
     {
       "epoch": 0.36437999628183676,
-      "grad_norm": 0.2608420079149271,
+      "grad_norm": 0.12784560560432576,
       "learning_rate": 0.00016035279124376026,
-      "loss": 3.7375,
+      "loss": 1.8219,
       "step": 2940
     },
     {
       "epoch": 0.3649996901530644,
-      "grad_norm": 0.2669515382231752,
+      "grad_norm": 0.1257259960150117,
       "learning_rate": 0.00016018015852849806,
-      "loss": 3.6461,
+      "loss": 1.8211,
       "step": 2945
     },
     {
       "epoch": 0.365619384024292,
-      "grad_norm": 0.25530803310408307,
+      "grad_norm": 0.12929114347135717,
       "learning_rate": 0.00016000724416953744,
-      "loss": 3.7367,
+      "loss": 1.8547,
       "step": 2950
     },
     {
       "epoch": 0.3662390778955196,
-      "grad_norm": 0.2627912871812224,
+      "grad_norm": 0.13234287937382339,
       "learning_rate": 0.00015983404897611928,
-      "loss": 3.8031,
+      "loss": 1.8484,
       "step": 2955
     },
     {
       "epoch": 0.36685877176674725,
-      "grad_norm": 0.25961614988600124,
+      "grad_norm": 0.1340348171981682,
       "learning_rate": 0.00015966057375879858,
-      "loss": 3.7047,
+      "loss": 1.8441,
       "step": 2960
     },
     {
       "epoch": 0.3674784656379748,
-      "grad_norm": 0.26065112696557874,
+      "grad_norm": 0.1331258514431446,
       "learning_rate": 0.00015948681932944104,
-      "loss": 3.8273,
+      "loss": 1.8586,
       "step": 2965
     },
     {
       "epoch": 0.36809815950920244,
-      "grad_norm": 0.24901377524126064,
+      "grad_norm": 0.13335124378868662,
       "learning_rate": 0.00015931278650121897,
-      "loss": 3.9438,
+      "loss": 1.866,
       "step": 2970
     },
     {
       "epoch": 0.36871785338043006,
-      "grad_norm": 0.2632224432964343,
+      "grad_norm": 0.13250687545966788,
       "learning_rate": 0.0001591384760886076,
-      "loss": 3.6617,
+      "loss": 1.8316,
       "step": 2975
     },
     {
       "epoch": 0.3693375472516577,
-      "grad_norm": 0.2589867585144316,
+      "grad_norm": 0.13274506427928803,
       "learning_rate": 0.00015896388890738127,
-      "loss": 3.6992,
+      "loss": 1.8535,
       "step": 2980
     },
     {
       "epoch": 0.3699572411228853,
-      "grad_norm": 0.25293946941175477,
+      "grad_norm": 0.13699657491823822,
       "learning_rate": 0.00015878902577460963,
-      "loss": 3.7984,
+      "loss": 1.8355,
       "step": 2985
     },
     {
       "epoch": 0.3705769349941129,
-      "grad_norm": 0.2551526711728292,
+      "grad_norm": 0.12986974166640744,
       "learning_rate": 0.00015861388750865375,
-      "loss": 3.6258,
+      "loss": 1.8656,
       "step": 2990
     },
     {
       "epoch": 0.3711966288653405,
-      "grad_norm": 0.2694067525937876,
+      "grad_norm": 0.13547504112097924,
       "learning_rate": 0.00015843847492916235,
-      "loss": 3.6578,
+      "loss": 1.859,
       "step": 2995
     },
     {
       "epoch": 0.3718163227365681,
-      "grad_norm": 0.2799079437013687,
+      "grad_norm": 0.14199787848571688,
       "learning_rate": 0.00015826278885706788,
-      "loss": 3.6336,
+      "loss": 1.8289,
       "step": 3000
     },
     {
       "epoch": 0.37243601660779574,
-      "grad_norm": 0.26696040065372506,
+      "grad_norm": 0.1304953733283427,
       "learning_rate": 0.0001580868301145828,
-      "loss": 3.8367,
+      "loss": 1.857,
       "step": 3005
     },
     {
       "epoch": 0.37305571047902336,
-      "grad_norm": 0.2748637513729295,
+      "grad_norm": 0.1448823429952658,
       "learning_rate": 0.00015791059952519567,
-      "loss": 3.6117,
+      "loss": 1.8668,
       "step": 3010
     },
     {
       "epoch": 0.373675404350251,
-      "grad_norm": 0.2678057810299332,
+      "grad_norm": 0.1257480778182083,
       "learning_rate": 0.00015773409791366728,
-      "loss": 3.7703,
+      "loss": 1.8605,
       "step": 3015
     },
     {
       "epoch": 0.3742950982214786,
-      "grad_norm": 0.26522017217912247,
+      "grad_norm": 0.13028189442894114,
       "learning_rate": 0.00015755732610602677,
-      "loss": 3.6922,
+      "loss": 1.8863,
       "step": 3020
     },
     {
       "epoch": 0.37491479209270623,
-      "grad_norm": 0.2663224361660835,
+      "grad_norm": 0.1265454045809503,
       "learning_rate": 0.00015738028492956786,
-      "loss": 3.6187,
+      "loss": 1.8395,
       "step": 3025
     },
     {
       "epoch": 0.3755344859639338,
-      "grad_norm": 0.27035194720649564,
+      "grad_norm": 0.13032347867143443,
       "learning_rate": 0.0001572029752128449,
-      "loss": 3.8258,
+      "loss": 1.8648,
       "step": 3030
     },
     {
       "epoch": 0.3761541798351614,
-      "grad_norm": 0.2838577149104673,
+      "grad_norm": 0.13110543872248587,
       "learning_rate": 0.00015702539778566897,
-      "loss": 3.7875,
+      "loss": 1.8715,
       "step": 3035
     },
     {
       "epoch": 0.37677387370638904,
-      "grad_norm": 0.2703312743767914,
+      "grad_norm": 0.1274284160679105,
       "learning_rate": 0.0001568475534791041,
-      "loss": 3.757,
+      "loss": 1.8086,
       "step": 3040
     },
     {
       "epoch": 0.37739356757761666,
-      "grad_norm": 0.2560504441017639,
+      "grad_norm": 0.1398602369199391,
       "learning_rate": 0.00015666944312546328,
-      "loss": 3.7008,
+      "loss": 1.8254,
       "step": 3045
     },
     {
       "epoch": 0.3780132614488443,
-      "grad_norm": 0.2698560889509554,
+      "grad_norm": 0.12891041218992944,
       "learning_rate": 0.0001564910675583046,
-      "loss": 3.6461,
+      "loss": 1.8781,
       "step": 3050
     },
     {
       "epoch": 0.3786329553200719,
-      "grad_norm": 0.28083643078936255,
+      "grad_norm": 0.1282542155930589,
       "learning_rate": 0.0001563124276124274,
-      "loss": 3.6453,
+      "loss": 1.884,
       "step": 3055
     },
     {
       "epoch": 0.3792526491912995,
-      "grad_norm": 0.271443414081348,
+      "grad_norm": 0.1356534667765166,
       "learning_rate": 0.00015613352412386825,
-      "loss": 3.6398,
+      "loss": 1.8012,
       "step": 3060
     },
     {
       "epoch": 0.3798723430625271,
-      "grad_norm": 0.2838039929744833,
+      "grad_norm": 0.1305522702140196,
       "learning_rate": 0.00015595435792989718,
-      "loss": 3.7586,
+      "loss": 1.8656,
       "step": 3065
     },
     {
       "epoch": 0.3804920369337547,
-      "grad_norm": 0.2647322485900605,
+      "grad_norm": 0.1337186431902072,
       "learning_rate": 0.0001557749298690135,
-      "loss": 3.6781,
+      "loss": 1.8883,
       "step": 3070
     },
     {
       "epoch": 0.38111173080498234,
-      "grad_norm": 0.2626946352782939,
+      "grad_norm": 0.1267318699700854,
       "learning_rate": 0.00015559524078094235,
-      "loss": 3.5805,
+      "loss": 1.8426,
       "step": 3075
     },
     {
       "epoch": 0.38173142467620996,
-      "grad_norm": 0.25933872796885626,
+      "grad_norm": 0.135229302814105,
       "learning_rate": 0.00015541529150663022,
-      "loss": 3.6148,
+      "loss": 1.8293,
       "step": 3080
     },
     {
       "epoch": 0.3823511185474376,
-      "grad_norm": 0.2821869561149129,
+      "grad_norm": 0.1325469745229473,
       "learning_rate": 0.00015523508288824145,
-      "loss": 3.6742,
+      "loss": 1.8598,
       "step": 3085
     },
     {
       "epoch": 0.38297081241866515,
-      "grad_norm": 0.25882550675825317,
+      "grad_norm": 0.13377461767007065,
       "learning_rate": 0.00015505461576915402,
-      "loss": 3.6023,
+      "loss": 1.8438,
       "step": 3090
     },
     {
       "epoch": 0.3835905062898928,
-      "grad_norm": 0.2589531444373828,
+      "grad_norm": 0.12407978728389873,
       "learning_rate": 0.00015487389099395565,
-      "loss": 3.6742,
+      "loss": 1.8715,
       "step": 3095
     },
     {
       "epoch": 0.3842102001611204,
-      "grad_norm": 0.26451321091244295,
+      "grad_norm": 0.1359925759348009,
       "learning_rate": 0.00015469290940844005,
-      "loss": 3.5445,
+      "loss": 1.8883,
       "step": 3100
     },
     {
       "epoch": 0.384829894032348,
-      "grad_norm": 0.2546537891704748,
+      "grad_norm": 0.13659768219370053,
       "learning_rate": 0.00015451167185960267,
-      "loss": 3.6609,
+      "loss": 1.8687,
       "step": 3105
     },
     {
       "epoch": 0.38544958790357564,
-      "grad_norm": 0.25027138697419005,
+      "grad_norm": 0.13102087732269108,
       "learning_rate": 0.00015433017919563692,
-      "loss": 3.5992,
+      "loss": 1.8902,
       "step": 3110
     },
     {
       "epoch": 0.38606928177480326,
-      "grad_norm": 0.27380537472291283,
+      "grad_norm": 0.13153742455134876,
       "learning_rate": 0.00015414843226593016,
-      "loss": 3.8633,
+      "loss": 1.8695,
       "step": 3115
     },
     {
       "epoch": 0.3866889756460309,
-      "grad_norm": 0.26921886223630787,
+      "grad_norm": 0.12492614271915935,
       "learning_rate": 0.0001539664319210597,
-      "loss": 3.7367,
+      "loss": 1.8867,
       "step": 3120
     },
     {
       "epoch": 0.38730866951725845,
-      "grad_norm": 0.25751518580294624,
+      "grad_norm": 0.1270230322153982,
       "learning_rate": 0.0001537841790127888,
-      "loss": 3.6586,
+      "loss": 1.8914,
       "step": 3125
     },
     {
       "epoch": 0.3879283633884861,
-      "grad_norm": 0.25605562468731,
+      "grad_norm": 0.13908099972235655,
       "learning_rate": 0.00015360167439406274,
-      "loss": 3.6141,
+      "loss": 1.8918,
       "step": 3130
     },
     {
       "epoch": 0.3885480572597137,
-      "grad_norm": 0.26611424342018436,
+      "grad_norm": 0.13260524566329848,
       "learning_rate": 0.00015341891891900494,
-      "loss": 3.6664,
+      "loss": 1.8469,
       "step": 3135
     },
     {
       "epoch": 0.3891677511309413,
-      "grad_norm": 0.2550066843855869,
+      "grad_norm": 0.1324975813379851,
       "learning_rate": 0.00015323591344291258,
-      "loss": 3.6484,
+      "loss": 1.8289,
       "step": 3140
     },
     {
       "epoch": 0.38978744500216894,
-      "grad_norm": 0.2821207360840596,
+      "grad_norm": 0.13120666575694298,
       "learning_rate": 0.00015305265882225303,
-      "loss": 3.6,
+      "loss": 1.8547,
       "step": 3145
     },
     {
       "epoch": 0.39040713887339656,
-      "grad_norm": 0.266631454756881,
+      "grad_norm": 0.12685427898285384,
       "learning_rate": 0.00015286915591465969,
-      "loss": 3.6953,
+      "loss": 1.8461,
       "step": 3150
     },
     {
       "epoch": 0.39102683274462413,
-      "grad_norm": 0.2686000416360816,
+      "grad_norm": 0.13572934465730757,
       "learning_rate": 0.00015268540557892773,
-      "loss": 3.6992,
+      "loss": 1.8562,
       "step": 3155
     },
     {
       "epoch": 0.39164652661585175,
-      "grad_norm": 0.25655840028409566,
+      "grad_norm": 0.12897435466478618,
       "learning_rate": 0.0001525014086750105,
-      "loss": 3.6492,
+      "loss": 1.8648,
       "step": 3160
     },
     {
       "epoch": 0.3922662204870794,
-      "grad_norm": 0.2567247759313816,
+      "grad_norm": 0.12721217574888402,
       "learning_rate": 0.0001523171660640152,
-      "loss": 3.6383,
+      "loss": 1.85,
       "step": 3165
     },
     {
       "epoch": 0.392885914358307,
-      "grad_norm": 0.2614801283302848,
+      "grad_norm": 0.1315866272081269,
       "learning_rate": 0.00015213267860819896,
-      "loss": 3.718,
+      "loss": 1.834,
       "step": 3170
     },
     {
       "epoch": 0.3935056082295346,
-      "grad_norm": 0.2538151707407498,
+      "grad_norm": 0.1302044760111287,
       "learning_rate": 0.00015194794717096475,
-      "loss": 3.8602,
+      "loss": 1.8512,
       "step": 3175
     },
     {
       "epoch": 0.39412530210076224,
-      "grad_norm": 0.2791424481656367,
+      "grad_norm": 0.13001468678731848,
       "learning_rate": 0.00015176297261685742,
-      "loss": 3.6617,
+      "loss": 1.8543,
       "step": 3180
     },
     {
       "epoch": 0.3947449959719898,
-      "grad_norm": 0.25798245204294173,
+      "grad_norm": 0.12952145972803736,
       "learning_rate": 0.00015157775581155957,
-      "loss": 3.8,
+      "loss": 1.8969,
       "step": 3185
     },
     {
       "epoch": 0.39536468984321743,
-      "grad_norm": 0.2614655828962737,
+      "grad_norm": 0.1267527829101824,
       "learning_rate": 0.00015139229762188761,
-      "loss": 3.6961,
+      "loss": 1.8113,
       "step": 3190
     },
     {
       "epoch": 0.39598438371444505,
-      "grad_norm": 0.25516127522728255,
+      "grad_norm": 0.1300639208984554,
       "learning_rate": 0.00015120659891578754,
-      "loss": 3.6828,
+      "loss": 1.7961,
       "step": 3195
     },
     {
       "epoch": 0.3966040775856727,
-      "grad_norm": 0.2585870302228633,
+      "grad_norm": 0.13154316703911634,
       "learning_rate": 0.00015102066056233104,
-      "loss": 3.5375,
+      "loss": 1.809,
       "step": 3200
     },
     {
       "epoch": 0.3972237714569003,
-      "grad_norm": 0.26672211172503196,
+      "grad_norm": 0.12568749277869354,
       "learning_rate": 0.00015083448343171138,
-      "loss": 3.632,
+      "loss": 1.8641,
       "step": 3205
     },
     {
       "epoch": 0.3978434653281279,
-      "grad_norm": 0.2641159714007487,
+      "grad_norm": 0.1319736744786111,
       "learning_rate": 0.00015064806839523915,
-      "loss": 3.7586,
+      "loss": 1.8223,
       "step": 3210
     },
     {
       "epoch": 0.39846315919935554,
-      "grad_norm": 0.2546667091800772,
+      "grad_norm": 0.13834323428246187,
       "learning_rate": 0.00015046141632533844,
-      "loss": 3.6641,
+      "loss": 1.798,
       "step": 3215
     },
     {
       "epoch": 0.3990828530705831,
-      "grad_norm": 0.25959333523005085,
+      "grad_norm": 0.1383968813858178,
       "learning_rate": 0.0001502745280955428,
-      "loss": 3.6898,
+      "loss": 1.8828,
       "step": 3220
     },
     {
       "epoch": 0.39970254694181073,
-      "grad_norm": 0.26928119386383115,
+      "grad_norm": 0.1350264311456221,
       "learning_rate": 0.00015008740458049075,
-      "loss": 3.7297,
+      "loss": 1.85,
       "step": 3225
     },
     {
       "epoch": 0.40032224081303835,
-      "grad_norm": 0.26421342010073934,
+      "grad_norm": 0.12512719310704826,
       "learning_rate": 0.0001499000466559221,
-      "loss": 3.7062,
+      "loss": 1.8148,
       "step": 3230
     },
     {
       "epoch": 0.400941934684266,
-      "grad_norm": 0.2643356968703811,
+      "grad_norm": 0.12311809256785136,
       "learning_rate": 0.0001497124551986737,
-      "loss": 3.5867,
+      "loss": 1.8617,
       "step": 3235
     },
     {
       "epoch": 0.4015616285554936,
-      "grad_norm": 0.26207830122927056,
+      "grad_norm": 0.13652089005902895,
       "learning_rate": 0.00014952463108667527,
-      "loss": 3.6039,
+      "loss": 1.8809,
       "step": 3240
     },
     {
       "epoch": 0.4021813224267212,
-      "grad_norm": 0.24755063014948148,
+      "grad_norm": 0.13149944904761232,
       "learning_rate": 0.0001493365751989454,
-      "loss": 3.6055,
+      "loss": 1.8309,
       "step": 3245
     },
     {
       "epoch": 0.4028010162979488,
-      "grad_norm": 0.2549356478882979,
+      "grad_norm": 0.1345131553547909,
       "learning_rate": 0.0001491482884155874,
-      "loss": 3.7414,
+      "loss": 1.8598,
       "step": 3250
     },
     {
       "epoch": 0.4034207101691764,
-      "grad_norm": 0.25938021841255027,
+      "grad_norm": 0.1316820704074098,
       "learning_rate": 0.00014895977161778515,
-      "loss": 3.6727,
+      "loss": 1.8113,
       "step": 3255
     },
     {
       "epoch": 0.40404040404040403,
-      "grad_norm": 0.2749384696162015,
+      "grad_norm": 0.13378258777396831,
       "learning_rate": 0.00014877102568779896,
-      "loss": 3.7242,
+      "loss": 1.8176,
       "step": 3260
     },
     {
       "epoch": 0.40466009791163166,
-      "grad_norm": 0.2648786444292922,
+      "grad_norm": 0.12936930942353125,
       "learning_rate": 0.00014858205150896161,
-      "loss": 3.5367,
+      "loss": 1.8996,
       "step": 3265
     },
     {
       "epoch": 0.4052797917828593,
-      "grad_norm": 0.26384530229777414,
+      "grad_norm": 0.12946325447363718,
       "learning_rate": 0.00014839284996567392,
-      "loss": 3.7203,
+      "loss": 1.8574,
       "step": 3270
     },
     {
       "epoch": 0.4058994856540869,
-      "grad_norm": 0.25117656502349667,
+      "grad_norm": 0.12501290179364566,
       "learning_rate": 0.00014820342194340097,
-      "loss": 3.7609,
+      "loss": 1.8613,
       "step": 3275
     },
     {
       "epoch": 0.4065191795253145,
-      "grad_norm": 0.2729406020541514,
+      "grad_norm": 0.13006013274328393,
       "learning_rate": 0.00014801376832866754,
-      "loss": 3.7437,
+      "loss": 1.7965,
       "step": 3280
     },
     {
       "epoch": 0.4071388733965421,
-      "grad_norm": 0.2863416580211302,
+      "grad_norm": 0.1388419140952983,
       "learning_rate": 0.00014782389000905435,
-      "loss": 3.6664,
+      "loss": 1.8965,
       "step": 3285
     },
     {
       "epoch": 0.4077585672677697,
-      "grad_norm": 0.2698057546507351,
+      "grad_norm": 0.1324648077294965,
       "learning_rate": 0.00014763378787319373,
-      "loss": 3.7109,
+      "loss": 1.8023,
       "step": 3290
     },
     {
       "epoch": 0.40837826113899733,
-      "grad_norm": 0.2697403365035814,
+      "grad_norm": 0.12507324333204756,
       "learning_rate": 0.00014744346281076536,
-      "loss": 3.7273,
+      "loss": 1.8473,
       "step": 3295
     },
     {
       "epoch": 0.40899795501022496,
-      "grad_norm": 0.28347942725540815,
+      "grad_norm": 0.13248330177911272,
       "learning_rate": 0.00014725291571249236,
-      "loss": 3.5773,
+      "loss": 1.798,
       "step": 3300
     },
     {
       "epoch": 0.4096176488814526,
-      "grad_norm": 0.26301691355137335,
+      "grad_norm": 0.13183523671686795,
       "learning_rate": 0.00014706214747013685,
-      "loss": 3.7578,
+      "loss": 1.9359,
       "step": 3305
     },
     {
       "epoch": 0.4102373427526802,
-      "grad_norm": 0.26809494981415455,
+      "grad_norm": 0.1338057384961836,
       "learning_rate": 0.00014687115897649603,
-      "loss": 3.6211,
+      "loss": 1.8891,
       "step": 3310
     },
     {
       "epoch": 0.41085703662390777,
-      "grad_norm": 0.2723319521596946,
+      "grad_norm": 0.15985021119124002,
       "learning_rate": 0.00014667995112539774,
-      "loss": 3.6117,
+      "loss": 1.8215,
       "step": 3315
     },
     {
       "epoch": 0.4114767304951354,
-      "grad_norm": 0.26087202121957026,
+      "grad_norm": 0.1293260479521013,
       "learning_rate": 0.00014648852481169658,
-      "loss": 3.6602,
+      "loss": 1.8918,
       "step": 3320
     },
     {
       "epoch": 0.412096424366363,
-      "grad_norm": 0.26358093924993475,
+      "grad_norm": 0.13146519522588096,
       "learning_rate": 0.0001462968809312694,
-      "loss": 3.7313,
+      "loss": 1.875,
       "step": 3325
     },
     {
       "epoch": 0.41271611823759063,
-      "grad_norm": 0.2712089109808533,
+      "grad_norm": 0.1315902361033057,
       "learning_rate": 0.00014610502038101138,
-      "loss": 3.6992,
+      "loss": 1.8781,
       "step": 3330
     },
     {
       "epoch": 0.41333581210881826,
-      "grad_norm": 0.2590088092278918,
+      "grad_norm": 0.1333738909869963,
       "learning_rate": 0.00014591294405883162,
-      "loss": 3.6703,
+      "loss": 1.8441,
       "step": 3335
     },
     {
       "epoch": 0.4139555059800459,
-      "grad_norm": 0.241021196941427,
+      "grad_norm": 0.12218014895336611,
       "learning_rate": 0.00014572065286364908,
-      "loss": 3.8258,
+      "loss": 1.884,
       "step": 3340
     },
     {
       "epoch": 0.41457519985127345,
-      "grad_norm": 0.2702742950321888,
+      "grad_norm": 0.12242530399637636,
       "learning_rate": 0.00014552814769538844,
-      "loss": 3.7008,
+      "loss": 1.8645,
       "step": 3345
     },
     {
       "epoch": 0.41519489372250107,
-      "grad_norm": 0.27282696995789457,
+      "grad_norm": 0.13310842863340136,
       "learning_rate": 0.00014533542945497553,
-      "loss": 3.5375,
+      "loss": 1.8629,
       "step": 3350
     },
     {
       "epoch": 0.4158145875937287,
-      "grad_norm": 0.2573140589414904,
+      "grad_norm": 0.1284963494667993,
       "learning_rate": 0.00014514249904433362,
-      "loss": 3.6344,
+      "loss": 1.8809,
       "step": 3355
     },
     {
       "epoch": 0.4164342814649563,
-      "grad_norm": 0.26868082368329343,
+      "grad_norm": 0.1319653346345284,
       "learning_rate": 0.0001449493573663787,
-      "loss": 3.6586,
+      "loss": 1.9082,
       "step": 3360
     },
     {
       "epoch": 0.41705397533618394,
-      "grad_norm": 0.26050328505952564,
+      "grad_norm": 0.14072551811843476,
       "learning_rate": 0.00014475600532501566,
-      "loss": 3.7141,
+      "loss": 1.8258,
       "step": 3365
     },
     {
       "epoch": 0.41767366920741156,
-      "grad_norm": 0.260282390367127,
+      "grad_norm": 0.1274232137809557,
       "learning_rate": 0.00014456244382513386,
-      "loss": 3.7531,
+      "loss": 1.8305,
       "step": 3370
     },
     {
       "epoch": 0.4182933630786392,
-      "grad_norm": 0.2813266769594806,
+      "grad_norm": 0.12333734269464691,
       "learning_rate": 0.0001443686737726029,
-      "loss": 3.6906,
+      "loss": 1.8648,
       "step": 3375
     },
     {
       "epoch": 0.41891305694986675,
-      "grad_norm": 0.2739367336724833,
+      "grad_norm": 0.13623034425385716,
       "learning_rate": 0.00014417469607426838,
-      "loss": 3.6008,
+      "loss": 1.8617,
       "step": 3380
     },
     {
       "epoch": 0.41953275082109437,
-      "grad_norm": 0.280672946378796,
+      "grad_norm": 0.13153662546427136,
       "learning_rate": 0.00014398051163794776,
-      "loss": 3.6461,
+      "loss": 1.8262,
       "step": 3385
     },
     {
       "epoch": 0.420152444692322,
-      "grad_norm": 0.2743306758287759,
+      "grad_norm": 0.13256507723918945,
       "learning_rate": 0.000143786121372426,
-      "loss": 3.7211,
+      "loss": 1.8586,
       "step": 3390
     },
     {
       "epoch": 0.4207721385635496,
-      "grad_norm": 0.261594675478806,
+      "grad_norm": 0.12990196578968077,
       "learning_rate": 0.00014359152618745132,
-      "loss": 3.6203,
+      "loss": 1.8621,
       "step": 3395
     },
     {
       "epoch": 0.42139183243477724,
-      "grad_norm": 0.2575401667881192,
+      "grad_norm": 0.13238313238253183,
       "learning_rate": 0.00014339672699373104,
-      "loss": 3.7273,
+      "loss": 1.8941,
       "step": 3400
     },
     {
       "epoch": 0.42201152630600486,
-      "grad_norm": 0.2850665124359037,
+      "grad_norm": 0.14263827897990308,
       "learning_rate": 0.0001432017247029271,
-      "loss": 3.9281,
+      "loss": 1.8219,
       "step": 3405
     },
     {
       "epoch": 0.4226312201772324,
-      "grad_norm": 0.25780272018370204,
+      "grad_norm": 0.15066475507049107,
       "learning_rate": 0.00014300652022765207,
-      "loss": 3.6789,
+      "loss": 1.8301,
       "step": 3410
     },
     {
       "epoch": 0.42325091404846005,
-      "grad_norm": 0.2680886719065801,
+      "grad_norm": 0.13504045941780513,
       "learning_rate": 0.00014281111448146468,
-      "loss": 3.6797,
+      "loss": 1.7824,
       "step": 3415
     },
     {
       "epoch": 0.42387060791968767,
-      "grad_norm": 0.29366146711228375,
+      "grad_norm": 0.12845118948413084,
       "learning_rate": 0.00014261550837886566,
-      "loss": 3.5805,
+      "loss": 1.8328,
       "step": 3420
     },
     {
       "epoch": 0.4244903017909153,
-      "grad_norm": 0.27265368036979465,
+      "grad_norm": 0.13416853262567752,
       "learning_rate": 0.00014241970283529338,
-      "loss": 3.718,
+      "loss": 1.8379,
       "step": 3425
     },
     {
       "epoch": 0.4251099956621429,
-      "grad_norm": 0.2462250838243364,
+      "grad_norm": 0.1275150961925631,
       "learning_rate": 0.00014222369876711955,
-      "loss": 3.7469,
+      "loss": 1.8461,
       "step": 3430
     },
     {
       "epoch": 0.42572968953337054,
-      "grad_norm": 0.2693760214974965,
+      "grad_norm": 0.129005270669972,
       "learning_rate": 0.00014202749709164506,
-      "loss": 3.6523,
+      "loss": 1.873,
       "step": 3435
     },
     {
       "epoch": 0.4263493834045981,
-      "grad_norm": 0.26792153518324946,
+      "grad_norm": 0.1366842217344907,
       "learning_rate": 0.00014183109872709557,
-      "loss": 3.7125,
+      "loss": 1.8344,
       "step": 3440
     },
     {
       "epoch": 0.4269690772758257,
-      "grad_norm": 0.26064403191013946,
+      "grad_norm": 0.1302801253762385,
       "learning_rate": 0.0001416345045926172,
-      "loss": 3.6281,
+      "loss": 1.8301,
       "step": 3445
     },
     {
       "epoch": 0.42758877114705335,
-      "grad_norm": 0.26654172823695954,
+      "grad_norm": 0.12938457791256947,
       "learning_rate": 0.00014143771560827238,
-      "loss": 3.6727,
+      "loss": 1.8762,
       "step": 3450
     },
     {
       "epoch": 0.42820846501828097,
-      "grad_norm": 0.260911428860639,
+      "grad_norm": 0.1324358723606549,
       "learning_rate": 0.00014124073269503534,
-      "loss": 3.7625,
+      "loss": 1.8738,
       "step": 3455
     },
     {
       "epoch": 0.4288281588895086,
-      "grad_norm": 0.2651966413063993,
+      "grad_norm": 0.12690470679362756,
       "learning_rate": 0.0001410435567747879,
-      "loss": 3.6906,
+      "loss": 1.8445,
       "step": 3460
     },
     {
       "epoch": 0.4294478527607362,
-      "grad_norm": 0.25866632724130795,
+      "grad_norm": 0.12607166701292938,
       "learning_rate": 0.00014084618877031524,
-      "loss": 3.7539,
+      "loss": 1.8586,
       "step": 3465
     },
     {
       "epoch": 0.43006754663196384,
-      "grad_norm": 0.2633955145782413,
+      "grad_norm": 0.1295605508529987,
       "learning_rate": 0.00014064862960530143,
-      "loss": 3.6023,
+      "loss": 1.8449,
       "step": 3470
     },
     {
       "epoch": 0.4306872405031914,
-      "grad_norm": 0.26034897512497684,
+      "grad_norm": 0.12894088957046304,
       "learning_rate": 0.0001404508802043252,
-      "loss": 3.6977,
+      "loss": 1.8484,
       "step": 3475
     },
     {
       "epoch": 0.431306934374419,
-      "grad_norm": 0.2613913190175063,
+      "grad_norm": 0.1310486419482613,
       "learning_rate": 0.0001402529414928556,
-      "loss": 3.7391,
+      "loss": 1.8367,
       "step": 3480
     },
     {
       "epoch": 0.43192662824564665,
-      "grad_norm": 0.28135113964539377,
+      "grad_norm": 0.13439055155902957,
       "learning_rate": 0.00014005481439724753,
-      "loss": 3.6219,
+      "loss": 1.8523,
       "step": 3485
     },
     {
       "epoch": 0.43254632211687427,
-      "grad_norm": 0.2611757651696741,
+      "grad_norm": 0.14979693773268604,
       "learning_rate": 0.00013985649984473773,
-      "loss": 3.7156,
+      "loss": 1.8684,
       "step": 3490
     },
     {
       "epoch": 0.4331660159881019,
-      "grad_norm": 0.24919761728858297,
+      "grad_norm": 0.12645292055136997,
       "learning_rate": 0.0001396579987634401,
-      "loss": 3.6852,
+      "loss": 1.7848,
       "step": 3495
     },
     {
       "epoch": 0.4337857098593295,
-      "grad_norm": 0.26626990952610513,
+      "grad_norm": 0.13147498487604903,
       "learning_rate": 0.00013945931208234156,
-      "loss": 3.6867,
+      "loss": 1.8895,
       "step": 3500
     },
     {
       "epoch": 0.4344054037305571,
-      "grad_norm": 0.26859713621539,
+      "grad_norm": 0.13150080296671127,
       "learning_rate": 0.00013926044073129759,
-      "loss": 3.7789,
+      "loss": 1.8254,
       "step": 3505
     },
     {
       "epoch": 0.4350250976017847,
-      "grad_norm": 0.26988300474797133,
+      "grad_norm": 0.13683895326823053,
       "learning_rate": 0.00013906138564102793,
-      "loss": 3.7164,
+      "loss": 1.8613,
       "step": 3510
     },
     {
       "epoch": 0.4356447914730123,
-      "grad_norm": 0.261113153649504,
+      "grad_norm": 0.12805602955656897,
       "learning_rate": 0.0001388621477431123,
-      "loss": 3.6758,
+      "loss": 1.8676,
       "step": 3515
     },
     {
       "epoch": 0.43626448534423995,
-      "grad_norm": 0.2552976713312743,
+      "grad_norm": 0.12813673559403296,
       "learning_rate": 0.00013866272796998587,
-      "loss": 3.7383,
+      "loss": 1.8883,
       "step": 3520
     },
     {
       "epoch": 0.43688417921546757,
-      "grad_norm": 0.30498123644083946,
+      "grad_norm": 0.1370489786872922,
       "learning_rate": 0.00013846312725493504,
-      "loss": 3.7609,
+      "loss": 1.9152,
       "step": 3525
     },
     {
       "epoch": 0.4375038730866952,
-      "grad_norm": 0.2570153268183945,
+      "grad_norm": 0.13373578100982492,
       "learning_rate": 0.00013826334653209297,
-      "loss": 3.6797,
+      "loss": 1.8449,
       "step": 3530
     },
     {
       "epoch": 0.43812356695792276,
-      "grad_norm": 0.25538549220186507,
+      "grad_norm": 0.14141792280370644,
       "learning_rate": 0.00013806338673643534,
-      "loss": 3.6398,
+      "loss": 1.8258,
       "step": 3535
     },
     {
       "epoch": 0.4387432608291504,
-      "grad_norm": 0.28356474464611625,
+      "grad_norm": 0.12964895565829695,
       "learning_rate": 0.00013786324880377576,
-      "loss": 3.7352,
+      "loss": 1.8156,
       "step": 3540
     },
     {
       "epoch": 0.439362954700378,
-      "grad_norm": 0.26457024903444254,
+      "grad_norm": 0.13419405554541766,
       "learning_rate": 0.0001376629336707617,
-      "loss": 3.6758,
+      "loss": 1.8219,
       "step": 3545
     },
     {
       "epoch": 0.43998264857160563,
-      "grad_norm": 0.25432212592677816,
+      "grad_norm": 0.1328791545044252,
       "learning_rate": 0.0001374624422748698,
-      "loss": 3.6469,
+      "loss": 1.873,
       "step": 3550
     },
     {
       "epoch": 0.44060234244283325,
-      "grad_norm": 0.2734955603605783,
+      "grad_norm": 0.136160988058161,
       "learning_rate": 0.00013726177555440164,
-      "loss": 3.7242,
+      "loss": 1.8789,
       "step": 3555
     },
     {
       "epoch": 0.4412220363140609,
-      "grad_norm": 0.27154068833373923,
+      "grad_norm": 0.1335013252887933,
       "learning_rate": 0.0001370609344484793,
-      "loss": 3.8172,
+      "loss": 1.8137,
       "step": 3560
     },
     {
       "epoch": 0.4418417301852885,
-      "grad_norm": 0.2715147456646156,
+      "grad_norm": 0.13713205846127619,
       "learning_rate": 0.00013685991989704105,
-      "loss": 3.7508,
+      "loss": 1.8207,
       "step": 3565
     },
     {
       "epoch": 0.44246142405651606,
-      "grad_norm": 0.26561938249699224,
+      "grad_norm": 0.12998754814726307,
       "learning_rate": 0.00013665873284083685,
-      "loss": 3.6898,
+      "loss": 1.8371,
       "step": 3570
     },
     {
       "epoch": 0.4430811179277437,
-      "grad_norm": 0.25783961144820916,
+      "grad_norm": 0.13470803808424642,
       "learning_rate": 0.000136457374221424,
-      "loss": 3.5945,
+      "loss": 1.8477,
       "step": 3575
     },
     {
       "epoch": 0.4437008117989713,
-      "grad_norm": 0.27738329704213377,
+      "grad_norm": 0.13606447312464934,
       "learning_rate": 0.00013625584498116262,
-      "loss": 3.8328,
+      "loss": 1.8625,
       "step": 3580
     },
     {
       "epoch": 0.44432050567019893,
-      "grad_norm": 0.2647663920533736,
+      "grad_norm": 0.1274570535945815,
       "learning_rate": 0.00013605414606321148,
-      "loss": 3.693,
+      "loss": 1.8641,
       "step": 3585
     },
     {
       "epoch": 0.44494019954142655,
-      "grad_norm": 0.2592544235227693,
+      "grad_norm": 0.1380534637059273,
       "learning_rate": 0.0001358522784115234,
-      "loss": 3.7898,
+      "loss": 1.8723,
       "step": 3590
     },
     {
       "epoch": 0.4455598934126542,
-      "grad_norm": 0.2614819616971976,
+      "grad_norm": 0.13189168748093322,
       "learning_rate": 0.00013565024297084084,
-      "loss": 3.6461,
+      "loss": 1.8891,
       "step": 3595
     },
     {
       "epoch": 0.44617958728388174,
-      "grad_norm": 0.26575079677829216,
+      "grad_norm": 0.1343904076364209,
       "learning_rate": 0.0001354480406866915,
-      "loss": 3.6117,
+      "loss": 1.802,
       "step": 3600
     },
     {
       "epoch": 0.44679928115510936,
-      "grad_norm": 0.2856386573982759,
+      "grad_norm": 0.13145575069138363,
       "learning_rate": 0.00013524567250538396,
-      "loss": 3.6984,
+      "loss": 1.8598,
       "step": 3605
     },
     {
       "epoch": 0.447418975026337,
-      "grad_norm": 0.26706444441282795,
+      "grad_norm": 0.1253081356231435,
       "learning_rate": 0.00013504313937400317,
-      "loss": 3.7031,
+      "loss": 1.8656,
       "step": 3610
     },
     {
       "epoch": 0.4480386688975646,
-      "grad_norm": 0.2611623259231443,
+      "grad_norm": 0.1359581987770173,
       "learning_rate": 0.00013484044224040606,
-      "loss": 3.6125,
+      "loss": 1.8406,
       "step": 3615
     },
     {
       "epoch": 0.44865836276879223,
-      "grad_norm": 0.27732328827061903,
+      "grad_norm": 0.13280063216880456,
       "learning_rate": 0.00013463758205321715,
-      "loss": 3.7352,
+      "loss": 1.8676,
       "step": 3620
     },
     {
       "epoch": 0.44927805664001985,
-      "grad_norm": 0.26044312902407674,
+      "grad_norm": 0.13852509864663975,
       "learning_rate": 0.0001344345597618239,
-      "loss": 3.6938,
+      "loss": 1.8273,
       "step": 3625
     },
     {
       "epoch": 0.4498977505112474,
-      "grad_norm": 0.2912247141373232,
+      "grad_norm": 0.13283988137837371,
       "learning_rate": 0.00013423137631637258,
-      "loss": 3.6617,
+      "loss": 1.8582,
       "step": 3630
     },
     {
       "epoch": 0.45051744438247504,
-      "grad_norm": 0.2571060992437349,
+      "grad_norm": 0.13375149735488798,
       "learning_rate": 0.00013402803266776353,
-      "loss": 3.7484,
+      "loss": 1.8668,
       "step": 3635
     },
     {
       "epoch": 0.45113713825370266,
-      "grad_norm": 0.25618111573779184,
+      "grad_norm": 0.1385343237508899,
       "learning_rate": 0.00013382452976764693,
-      "loss": 3.7219,
+      "loss": 1.8301,
       "step": 3640
     },
     {
       "epoch": 0.4517568321249303,
-      "grad_norm": 0.27582839571574014,
+      "grad_norm": 0.1336021388356578,
       "learning_rate": 0.00013362086856841826,
-      "loss": 3.6859,
+      "loss": 1.868,
       "step": 3645
     },
     {
       "epoch": 0.4523765259961579,
-      "grad_norm": 0.27749394162596547,
+      "grad_norm": 0.1328856221983309,
       "learning_rate": 0.0001334170500232138,
-      "loss": 3.7211,
+      "loss": 1.8555,
       "step": 3650
     },
     {
       "epoch": 0.45299621986738553,
-      "grad_norm": 0.2584168295006351,
+      "grad_norm": 0.1341766922517231,
       "learning_rate": 0.00013321307508590624,
-      "loss": 3.6625,
+      "loss": 1.8523,
       "step": 3655
     },
     {
       "epoch": 0.45361591373861315,
-      "grad_norm": 0.27693424489312013,
+      "grad_norm": 0.1306615008434302,
       "learning_rate": 0.00013300894471110014,
-      "loss": 3.6625,
+      "loss": 1.8617,
       "step": 3660
     },
     {
       "epoch": 0.4542356076098407,
-      "grad_norm": 0.25983758091273174,
+      "grad_norm": 0.13912806028945696,
       "learning_rate": 0.00013280465985412757,
-      "loss": 3.6641,
+      "loss": 1.8266,
       "step": 3665
     },
     {
       "epoch": 0.45485530148106834,
-      "grad_norm": 0.26415706763108404,
+      "grad_norm": 0.12851937389904652,
       "learning_rate": 0.00013260022147104354,
-      "loss": 3.6781,
+      "loss": 1.8516,
       "step": 3670
     },
     {
       "epoch": 0.45547499535229596,
-      "grad_norm": 0.2719567568985379,
+      "grad_norm": 0.13943073131832875,
       "learning_rate": 0.00013239563051862158,
-      "loss": 3.5859,
+      "loss": 1.825,
       "step": 3675
     },
     {
       "epoch": 0.4560946892235236,
-      "grad_norm": 0.26545474762934684,
+      "grad_norm": 0.13670686199925514,
       "learning_rate": 0.00013219088795434923,
-      "loss": 3.8219,
+      "loss": 1.8387,
       "step": 3680
     },
     {
       "epoch": 0.4567143830947512,
-      "grad_norm": 0.2708748781556674,
+      "grad_norm": 0.1356087167140527,
       "learning_rate": 0.00013198599473642354,
-      "loss": 3.7672,
+      "loss": 1.8902,
       "step": 3685
     },
     {
       "epoch": 0.45733407696597883,
-      "grad_norm": 0.2715721444457544,
+      "grad_norm": 0.13676670096618956,
       "learning_rate": 0.00013178095182374676,
-      "loss": 3.6945,
+      "loss": 1.7934,
       "step": 3690
     },
     {
       "epoch": 0.4579537708372064,
-      "grad_norm": 0.2716587540915426,
+      "grad_norm": 0.13391467393093984,
       "learning_rate": 0.00013157576017592157,
-      "loss": 3.682,
+      "loss": 1.8254,
       "step": 3695
     },
     {
       "epoch": 0.458573464708434,
-      "grad_norm": 0.2582125776966492,
+      "grad_norm": 0.13764440897029354,
       "learning_rate": 0.0001313704207532468,
-      "loss": 3.6469,
+      "loss": 1.8453,
       "step": 3700
     },
     {
       "epoch": 0.45919315857966164,
-      "grad_norm": 0.2589681636647802,
+      "grad_norm": 0.13160749306039446,
       "learning_rate": 0.00013116493451671279,
-      "loss": 3.7172,
+      "loss": 1.8352,
       "step": 3705
     },
     {
       "epoch": 0.45981285245088926,
-      "grad_norm": 0.268075424802474,
+      "grad_norm": 0.12949459213686204,
       "learning_rate": 0.000130959302427997,
-      "loss": 3.6781,
+      "loss": 1.7816,
       "step": 3710
     },
     {
       "epoch": 0.4604325463221169,
-      "grad_norm": 0.2701875386080495,
+      "grad_norm": 0.13738089812902068,
       "learning_rate": 0.00013075352544945966,
-      "loss": 3.6602,
+      "loss": 1.8633,
       "step": 3715
     },
     {
       "epoch": 0.4610522401933445,
-      "grad_norm": 0.2874768630812172,
+      "grad_norm": 0.1338213874344654,
       "learning_rate": 0.00013054760454413882,
-      "loss": 3.632,
+      "loss": 1.8727,
       "step": 3720
     },
     {
       "epoch": 0.4616719340645721,
-      "grad_norm": 0.2912578065906555,
+      "grad_norm": 0.13465341196540553,
       "learning_rate": 0.00013034154067574622,
-      "loss": 3.7148,
+      "loss": 1.8824,
       "step": 3725
     },
     {
       "epoch": 0.4622916279357997,
-      "grad_norm": 0.2710623247745548,
+      "grad_norm": 0.1312583538008533,
       "learning_rate": 0.00013013533480866273,
-      "loss": 3.7938,
+      "loss": 1.8133,
       "step": 3730
     },
     {
       "epoch": 0.4629113218070273,
-      "grad_norm": 0.26344004601414733,
+      "grad_norm": 0.13231035374316044,
       "learning_rate": 0.00012992898790793362,
-      "loss": 3.5797,
+      "loss": 1.8496,
       "step": 3735
     },
     {
       "epoch": 0.46353101567825494,
-      "grad_norm": 0.26207544083031686,
+      "grad_norm": 0.1322425189925765,
       "learning_rate": 0.00012972250093926436,
-      "loss": 3.6602,
+      "loss": 1.8512,
       "step": 3740
     },
     {
       "epoch": 0.46415070954948257,
-      "grad_norm": 0.2601250259564004,
+      "grad_norm": 0.12728720946766525,
       "learning_rate": 0.0001295158748690159,
-      "loss": 3.6859,
+      "loss": 1.8441,
       "step": 3745
     },
     {
       "epoch": 0.4647704034207102,
-      "grad_norm": 0.26401294757239785,
+      "grad_norm": 0.12825916363660014,
       "learning_rate": 0.0001293091106642001,
-      "loss": 3.7633,
+      "loss": 1.8645,
       "step": 3750
     },
     {
       "epoch": 0.4653900972919378,
-      "grad_norm": 0.26767917213945047,
+      "grad_norm": 0.13151182512874157,
       "learning_rate": 0.00012910220929247538,
-      "loss": 3.7188,
+      "loss": 1.8449,
       "step": 3755
     },
     {
       "epoch": 0.4660097911631654,
-      "grad_norm": 0.258331889535157,
+      "grad_norm": 0.12577699319145508,
       "learning_rate": 0.00012889517172214206,
-      "loss": 3.8008,
+      "loss": 1.8254,
       "step": 3760
     },
     {
       "epoch": 0.466629485034393,
-      "grad_norm": 0.2662816231079526,
+      "grad_norm": 0.13662024127159117,
       "learning_rate": 0.0001286879989221379,
-      "loss": 3.5742,
+      "loss": 1.8336,
       "step": 3765
     },
     {
       "epoch": 0.4672491789056206,
-      "grad_norm": 0.2678195712749907,
+      "grad_norm": 0.13740650410044333,
       "learning_rate": 0.0001284806918620335,
-      "loss": 3.518,
+      "loss": 1.8289,
       "step": 3770
     },
     {
       "epoch": 0.46786887277684824,
-      "grad_norm": 0.26956828977591457,
+      "grad_norm": 0.12670365930896482,
       "learning_rate": 0.00012827325151202782,
-      "loss": 3.6875,
+      "loss": 1.8402,
       "step": 3775
     },
     {
       "epoch": 0.46848856664807587,
-      "grad_norm": 0.2632140447372612,
+      "grad_norm": 0.13596871709025227,
       "learning_rate": 0.00012806567884294362,
-      "loss": 3.6727,
+      "loss": 1.8809,
       "step": 3780
     },
     {
       "epoch": 0.4691082605193035,
-      "grad_norm": 0.26622278914922576,
+      "grad_norm": 0.12839958268082616,
       "learning_rate": 0.00012785797482622294,
-      "loss": 3.7008,
+      "loss": 1.8414,
       "step": 3785
     },
     {
       "epoch": 0.46972795439053106,
-      "grad_norm": 0.25787244958734795,
+      "grad_norm": 0.13501786379676173,
       "learning_rate": 0.00012765014043392242,
-      "loss": 3.7172,
+      "loss": 1.8508,
       "step": 3790
     },
     {
       "epoch": 0.4703476482617587,
-      "grad_norm": 0.2628067137621567,
+      "grad_norm": 0.14517241168125267,
       "learning_rate": 0.00012744217663870902,
-      "loss": 3.6273,
+      "loss": 1.8797,
       "step": 3795
     },
     {
       "epoch": 0.4709673421329863,
-      "grad_norm": 0.26372203837092634,
+      "grad_norm": 0.14211490967215193,
       "learning_rate": 0.00012723408441385521,
-      "loss": 3.682,
+      "loss": 1.8613,
       "step": 3800
     },
     {
       "epoch": 0.4715870360042139,
-      "grad_norm": 0.2824894463232923,
+      "grad_norm": 0.12861242954967542,
       "learning_rate": 0.0001270258647332345,
-      "loss": 3.5516,
+      "loss": 1.8535,
       "step": 3805
     },
     {
       "epoch": 0.47220672987544154,
-      "grad_norm": 0.27690777294629687,
+      "grad_norm": 0.12940845574021922,
       "learning_rate": 0.00012681751857131693,
-      "loss": 3.693,
+      "loss": 1.7973,
       "step": 3810
     },
     {
       "epoch": 0.47282642374666917,
-      "grad_norm": 0.2819860304503665,
+      "grad_norm": 0.13214817236795431,
       "learning_rate": 0.00012660904690316445,
-      "loss": 3.7016,
+      "loss": 1.8492,
       "step": 3815
     },
     {
       "epoch": 0.47344611761789673,
-      "grad_norm": 0.26197810073423866,
+      "grad_norm": 0.12562431874962993,
       "learning_rate": 0.00012640045070442643,
-      "loss": 3.7711,
+      "loss": 1.8629,
       "step": 3820
     },
     {
       "epoch": 0.47406581148912436,
-      "grad_norm": 0.27307637087247627,
+      "grad_norm": 0.1373453111960608,
       "learning_rate": 0.000126191730951335,
-      "loss": 3.7219,
+      "loss": 1.8086,
       "step": 3825
     },
     {
       "epoch": 0.474685505360352,
-      "grad_norm": 0.28020444831757324,
+      "grad_norm": 0.12730725941472784,
       "learning_rate": 0.0001259828886207005,
-      "loss": 3.6328,
+      "loss": 1.85,
       "step": 3830
     },
     {
       "epoch": 0.4753051992315796,
-      "grad_norm": 0.27306251075030796,
+      "grad_norm": 0.14568143374443207,
       "learning_rate": 0.00012577392468990695,
-      "loss": 3.6797,
+      "loss": 1.9332,
       "step": 3835
     },
     {
       "epoch": 0.4759248931028072,
-      "grad_norm": 0.25355275857310394,
+      "grad_norm": 0.1294161453337625,
       "learning_rate": 0.00012556484013690763,
-      "loss": 3.6898,
+      "loss": 1.8328,
       "step": 3840
     },
     {
       "epoch": 0.47654458697403485,
-      "grad_norm": 0.2597476632713353,
+      "grad_norm": 0.1360420665603735,
       "learning_rate": 0.00012535563594022,
-      "loss": 3.7625,
+      "loss": 1.8281,
       "step": 3845
     },
     {
       "epoch": 0.47716428084526247,
-      "grad_norm": 0.2658878725226564,
+      "grad_norm": 0.13045423595487943,
       "learning_rate": 0.0001251463130789217,
-      "loss": 3.6867,
+      "loss": 1.8367,
       "step": 3850
     },
     {
       "epoch": 0.47778397471649003,
-      "grad_norm": 0.2605086032950418,
+      "grad_norm": 0.1293506708386949,
       "learning_rate": 0.0001249368725326457,
-      "loss": 3.7094,
+      "loss": 1.8082,
       "step": 3855
     },
     {
       "epoch": 0.47840366858771766,
-      "grad_norm": 0.2616579389175633,
+      "grad_norm": 0.1279784621009452,
       "learning_rate": 0.00012472731528157563,
-      "loss": 3.7898,
+      "loss": 1.843,
       "step": 3860
     },
     {
       "epoch": 0.4790233624589453,
-      "grad_norm": 0.2685603453669324,
+      "grad_norm": 0.13460853167180994,
       "learning_rate": 0.00012451764230644145,
-      "loss": 3.7,
+      "loss": 1.893,
       "step": 3865
     },
     {
       "epoch": 0.4796430563301729,
-      "grad_norm": 0.26551263915399026,
+      "grad_norm": 0.13623636515106394,
       "learning_rate": 0.0001243078545885145,
-      "loss": 3.7445,
+      "loss": 1.8594,
       "step": 3870
     },
     {
       "epoch": 0.4802627502014005,
-      "grad_norm": 0.26462924352305306,
+      "grad_norm": 0.13469126915119556,
       "learning_rate": 0.00012409795310960333,
-      "loss": 3.7461,
+      "loss": 1.8188,
       "step": 3875
     },
     {
       "epoch": 0.48088244407262815,
-      "grad_norm": 0.2885862355706348,
+      "grad_norm": 0.14728913841193927,
       "learning_rate": 0.00012388793885204875,
-      "loss": 3.7258,
+      "loss": 1.8496,
       "step": 3880
     },
     {
       "epoch": 0.4815021379438557,
-      "grad_norm": 0.2572423376845503,
+      "grad_norm": 0.1329784046218279,
       "learning_rate": 0.00012367781279871946,
-      "loss": 3.7477,
+      "loss": 1.7828,
       "step": 3885
     },
     {
       "epoch": 0.48212183181508333,
-      "grad_norm": 0.2575035889902655,
+      "grad_norm": 0.14450915407347387,
       "learning_rate": 0.00012346757593300733,
-      "loss": 3.7156,
+      "loss": 1.8199,
       "step": 3890
     },
     {
       "epoch": 0.48274152568631096,
-      "grad_norm": 0.26737117219004586,
+      "grad_norm": 0.13074669708751388,
       "learning_rate": 0.00012325722923882285,
-      "loss": 3.7125,
+      "loss": 1.8406,
       "step": 3895
     },
     {
       "epoch": 0.4833612195575386,
-      "grad_norm": 0.26778941850758253,
+      "grad_norm": 0.1320766633941541,
       "learning_rate": 0.00012304677370059047,
-      "loss": 3.7703,
+      "loss": 1.825,
       "step": 3900
     },
     {
       "epoch": 0.4839809134287662,
-      "grad_norm": 0.2617718410058238,
+      "grad_norm": 0.13382977679788616,
       "learning_rate": 0.00012283621030324403,
-      "loss": 3.6742,
+      "loss": 1.7426,
       "step": 3905
     },
     {
       "epoch": 0.4846006072999938,
-      "grad_norm": 0.27110705498999677,
+      "grad_norm": 0.1300774595042929,
       "learning_rate": 0.00012262554003222221,
-      "loss": 3.6625,
+      "loss": 1.8277,
       "step": 3910
     },
     {
       "epoch": 0.4852203011712214,
-      "grad_norm": 0.26656137927275286,
+      "grad_norm": 0.13823641067099687,
       "learning_rate": 0.00012241476387346386,
-      "loss": 3.5656,
+      "loss": 1.8344,
       "step": 3915
     },
     {
       "epoch": 0.485839995042449,
-      "grad_norm": 0.27327458650598496,
+      "grad_norm": 0.13309285522719982,
       "learning_rate": 0.00012220388281340328,
-      "loss": 3.6492,
+      "loss": 1.8332,
       "step": 3920
     },
     {
       "epoch": 0.48645968891367664,
-      "grad_norm": 0.26535475968509176,
+      "grad_norm": 0.13989344182618982,
       "learning_rate": 0.00012199289783896582,
-      "loss": 3.6297,
+      "loss": 1.8418,
       "step": 3925
     },
     {
       "epoch": 0.48707938278490426,
-      "grad_norm": 0.27391081465042744,
+      "grad_norm": 0.13908793914835565,
       "learning_rate": 0.0001217818099375631,
-      "loss": 3.75,
+      "loss": 1.7738,
       "step": 3930
     },
     {
       "epoch": 0.4876990766561319,
-      "grad_norm": 0.2706935971949415,
+      "grad_norm": 0.13203859661605985,
       "learning_rate": 0.00012157062009708847,
-      "loss": 3.6688,
+      "loss": 1.8012,
       "step": 3935
     },
     {
       "epoch": 0.4883187705273595,
-      "grad_norm": 0.26673132822964696,
+      "grad_norm": 0.1387991780525731,
       "learning_rate": 0.00012135932930591232,
-      "loss": 3.7383,
+      "loss": 1.8258,
       "step": 3940
     },
     {
       "epoch": 0.4889384643985871,
-      "grad_norm": 0.2587169856596499,
+      "grad_norm": 0.1272927886215849,
       "learning_rate": 0.00012114793855287749,
-      "loss": 3.7258,
+      "loss": 1.8207,
       "step": 3945
     },
     {
       "epoch": 0.4895581582698147,
-      "grad_norm": 0.27051352083466496,
+      "grad_norm": 0.13317685573409976,
       "learning_rate": 0.00012093644882729473,
-      "loss": 3.7766,
+      "loss": 1.8484,
       "step": 3950
     },
     {
       "epoch": 0.4901778521410423,
-      "grad_norm": 0.2846141346585085,
+      "grad_norm": 0.13060406777243355,
       "learning_rate": 0.0001207248611189378,
-      "loss": 3.7437,
+      "loss": 1.8062,
       "step": 3955
     },
     {
       "epoch": 0.49079754601226994,
-      "grad_norm": 0.2901661878259012,
+      "grad_norm": 0.1323653833316828,
       "learning_rate": 0.00012051317641803921,
-      "loss": 3.6555,
+      "loss": 1.868,
       "step": 3960
     },
     {
       "epoch": 0.49141723988349756,
-      "grad_norm": 0.25373316084020386,
+      "grad_norm": 0.13374731314053315,
       "learning_rate": 0.00012030139571528534,
-      "loss": 3.6492,
+      "loss": 1.8027,
       "step": 3965
     },
     {
       "epoch": 0.4920369337547252,
-      "grad_norm": 0.26797411021921486,
+      "grad_norm": 0.13117397580253184,
       "learning_rate": 0.00012008952000181175,
-      "loss": 3.7234,
+      "loss": 1.8105,
       "step": 3970
     },
     {
       "epoch": 0.4926566276259528,
-      "grad_norm": 0.2586687656902996,
+      "grad_norm": 0.13913815292173712,
       "learning_rate": 0.0001198775502691988,
-      "loss": 3.5828,
+      "loss": 1.9066,
       "step": 3975
     },
     {
       "epoch": 0.49327632149718037,
-      "grad_norm": 0.2769207433199797,
+      "grad_norm": 0.13254556995995362,
       "learning_rate": 0.00011966548750946678,
-      "loss": 3.693,
+      "loss": 1.7578,
       "step": 3980
     },
     {
       "epoch": 0.493896015368408,
-      "grad_norm": 0.2660860997146998,
+      "grad_norm": 0.1263645566258469,
       "learning_rate": 0.00011945333271507138,
-      "loss": 3.6289,
+      "loss": 1.8551,
       "step": 3985
     },
     {
       "epoch": 0.4945157092396356,
-      "grad_norm": 0.28692103057195106,
+      "grad_norm": 0.13848000567316143,
       "learning_rate": 0.00011924108687889899,
-      "loss": 3.6469,
+      "loss": 1.8434,
       "step": 3990
     },
     {
       "epoch": 0.49513540311086324,
-      "grad_norm": 0.3003655228076814,
+      "grad_norm": 0.1361787753395237,
       "learning_rate": 0.00011902875099426207,
-      "loss": 3.5742,
+      "loss": 1.8449,
       "step": 3995
     },
     {
       "epoch": 0.49575509698209086,
-      "grad_norm": 0.26679154810260447,
+      "grad_norm": 0.12990782377541138,
       "learning_rate": 0.00011881632605489457,
-      "loss": 3.732,
+      "loss": 1.898,
       "step": 4000
     },
     {
       "epoch": 0.4963747908533185,
-      "grad_norm": 0.2584591602474966,
+      "grad_norm": 0.13345131711305486,
       "learning_rate": 0.0001186038130549471,
-      "loss": 3.6531,
+      "loss": 1.798,
       "step": 4005
     },
     {
       "epoch": 0.49699448472454605,
-      "grad_norm": 0.26241847954669006,
+      "grad_norm": 0.13073403167785114,
       "learning_rate": 0.00011839121298898253,
-      "loss": 3.6906,
+      "loss": 1.9184,
       "step": 4010
     },
     {
       "epoch": 0.49761417859577367,
-      "grad_norm": 0.2698264200108138,
+      "grad_norm": 0.128726619507065,
       "learning_rate": 0.00011817852685197109,
-      "loss": 3.6711,
+      "loss": 1.7777,
       "step": 4015
     },
     {
       "epoch": 0.4982338724670013,
-      "grad_norm": 0.26601297976121263,
+      "grad_norm": 0.1402678002131599,
       "learning_rate": 0.00011796575563928591,
-      "loss": 3.6437,
+      "loss": 1.8352,
       "step": 4020
     },
     {
       "epoch": 0.4988535663382289,
-      "grad_norm": 0.2611501124428443,
+      "grad_norm": 0.1532780370824944,
       "learning_rate": 0.00011775290034669822,
-      "loss": 3.6219,
+      "loss": 1.8375,
       "step": 4025
     },
     {
       "epoch": 0.49947326020945654,
-      "grad_norm": 0.2740304561773694,
+      "grad_norm": 0.1428411670169983,
       "learning_rate": 0.00011753996197037272,
-      "loss": 3.7164,
+      "loss": 1.8383,
       "step": 4030
     },
     {
       "epoch": 0.5000929540806841,
-      "grad_norm": 0.2604912300476886,
+      "grad_norm": 0.12794764388496246,
       "learning_rate": 0.00011732694150686301,
-      "loss": 3.7711,
+      "loss": 1.8629,
       "step": 4035
     },
     {
       "epoch": 0.5007126479519117,
-      "grad_norm": 0.2639969283942867,
+      "grad_norm": 0.12573790381059152,
       "learning_rate": 0.00011711383995310681,
-      "loss": 3.6781,
+      "loss": 1.8379,
       "step": 4040
     },
     {
       "epoch": 0.5013323418231393,
-      "grad_norm": 0.26395125505247186,
+      "grad_norm": 0.13444917133351966,
       "learning_rate": 0.00011690065830642143,
-      "loss": 3.6398,
+      "loss": 1.7645,
       "step": 4045
     },
     {
       "epoch": 0.501952035694367,
-      "grad_norm": 0.26422182168550484,
+      "grad_norm": 0.14311239318297794,
       "learning_rate": 0.00011668739756449885,
-      "loss": 3.7203,
+      "loss": 1.7723,
       "step": 4050
     },
     {
       "epoch": 0.5025717295655946,
-      "grad_norm": 0.2617310020917149,
+      "grad_norm": 0.13642980498274354,
       "learning_rate": 0.00011647405872540138,
-      "loss": 3.6469,
+      "loss": 1.8195,
       "step": 4055
     },
     {
       "epoch": 0.5031914234368222,
-      "grad_norm": 0.27066930312905935,
+      "grad_norm": 0.12807575497597842,
       "learning_rate": 0.00011626064278755673,
-      "loss": 3.8062,
+      "loss": 1.7895,
       "step": 4060
     },
     {
       "epoch": 0.5038111173080498,
-      "grad_norm": 0.27961161598532136,
+      "grad_norm": 0.1357891940700086,
       "learning_rate": 0.00011604715074975347,
-      "loss": 3.6719,
+      "loss": 1.8449,
       "step": 4065
     },
     {
       "epoch": 0.5044308111792775,
-      "grad_norm": 0.2666222034516121,
+      "grad_norm": 0.12954313901516068,
       "learning_rate": 0.00011583358361113632,
-      "loss": 3.7719,
+      "loss": 1.9016,
       "step": 4070
     },
     {
       "epoch": 0.5050505050505051,
-      "grad_norm": 0.2676080885232622,
+      "grad_norm": 0.13833393699221658,
       "learning_rate": 0.00011561994237120148,
-      "loss": 3.7039,
+      "loss": 1.8691,
       "step": 4075
     },
     {
       "epoch": 0.5056701989217327,
-      "grad_norm": 0.26051227668479,
+      "grad_norm": 0.13271819169332488,
       "learning_rate": 0.00011540622802979187,
-      "loss": 3.6602,
+      "loss": 1.8602,
       "step": 4080
     },
     {
       "epoch": 0.5062898927929603,
-      "grad_norm": 0.2620984894937847,
+      "grad_norm": 0.13361629977356454,
       "learning_rate": 0.00011519244158709263,
-      "loss": 3.7961,
+      "loss": 1.8945,
       "step": 4085
     },
     {
       "epoch": 0.5069095866641878,
-      "grad_norm": 0.27682148528587996,
+      "grad_norm": 0.13835947674800525,
       "learning_rate": 0.00011497858404362631,
-      "loss": 3.7305,
+      "loss": 1.8945,
       "step": 4090
     },
     {
       "epoch": 0.5075292805354155,
-      "grad_norm": 0.26332627536351605,
+      "grad_norm": 0.13507741578680432,
       "learning_rate": 0.00011476465640024814,
-      "loss": 3.7531,
+      "loss": 1.8098,
       "step": 4095
     },
     {
       "epoch": 0.5081489744066431,
-      "grad_norm": 0.2733755117038867,
+      "grad_norm": 0.12961267041068253,
       "learning_rate": 0.00011455065965814148,
-      "loss": 3.6711,
+      "loss": 1.8758,
       "step": 4100
     },
     {
       "epoch": 0.5087686682778707,
-      "grad_norm": 0.2547409812758734,
+      "grad_norm": 0.1426487693561521,
       "learning_rate": 0.00011433659481881307,
-      "loss": 3.8086,
+      "loss": 1.8414,
       "step": 4105
     },
     {
       "epoch": 0.5093883621490983,
-      "grad_norm": 0.2571255669278216,
+      "grad_norm": 0.129533245275186,
       "learning_rate": 0.00011412246288408835,
-      "loss": 3.6516,
+      "loss": 1.8836,
       "step": 4110
     },
     {
       "epoch": 0.510008056020326,
-      "grad_norm": 0.2854681210247425,
+      "grad_norm": 0.13192948437358754,
       "learning_rate": 0.00011390826485610675,
-      "loss": 3.6461,
+      "loss": 1.8355,
       "step": 4115
     },
     {
       "epoch": 0.5106277498915536,
-      "grad_norm": 0.2559172024534188,
+      "grad_norm": 0.12792800610505192,
       "learning_rate": 0.000113694001737317,
-      "loss": 3.7508,
+      "loss": 1.884,
       "step": 4120
     },
     {
       "epoch": 0.5112474437627812,
-      "grad_norm": 0.25176439604382256,
+      "grad_norm": 0.12825945154342566,
       "learning_rate": 0.00011347967453047248,
-      "loss": 3.7828,
+      "loss": 1.8324,
       "step": 4125
     },
     {
       "epoch": 0.5118671376340088,
-      "grad_norm": 0.2655443073530328,
+      "grad_norm": 0.12864350361168767,
       "learning_rate": 0.00011326528423862653,
-      "loss": 3.782,
+      "loss": 1.8367,
       "step": 4130
     },
     {
       "epoch": 0.5124868315052364,
-      "grad_norm": 0.2660618678109569,
+      "grad_norm": 0.13080630072709626,
       "learning_rate": 0.00011305083186512765,
-      "loss": 3.6,
+      "loss": 1.8031,
       "step": 4135
     },
     {
       "epoch": 0.5131065253764641,
-      "grad_norm": 0.2566751684212444,
+      "grad_norm": 0.1334944999274583,
       "learning_rate": 0.00011283631841361499,
-      "loss": 3.7523,
+      "loss": 1.8277,
       "step": 4140
     },
     {
       "epoch": 0.5137262192476917,
-      "grad_norm": 0.2798916301465522,
+      "grad_norm": 0.13277800972691645,
       "learning_rate": 0.00011262174488801349,
-      "loss": 3.7398,
+      "loss": 1.868,
       "step": 4145
     },
     {
       "epoch": 0.5143459131189193,
-      "grad_norm": 0.2631319357716128,
+      "grad_norm": 0.13163204057888897,
       "learning_rate": 0.00011240711229252915,
-      "loss": 3.668,
+      "loss": 1.827,
       "step": 4150
     },
     {
       "epoch": 0.5149656069901468,
-      "grad_norm": 0.2573477794528783,
+      "grad_norm": 0.1294009876519144,
       "learning_rate": 0.00011219242163164457,
-      "loss": 3.6672,
+      "loss": 1.927,
       "step": 4155
     },
     {
       "epoch": 0.5155853008613744,
-      "grad_norm": 0.2696847838745472,
+      "grad_norm": 0.13220239163919237,
       "learning_rate": 0.000111977673910114,
-      "loss": 3.625,
+      "loss": 1.9047,
       "step": 4160
     },
     {
       "epoch": 0.5162049947326021,
-      "grad_norm": 0.25202944696465945,
+      "grad_norm": 0.132410040296461,
       "learning_rate": 0.00011176287013295879,
-      "loss": 3.7016,
+      "loss": 1.8996,
       "step": 4165
     },
     {
       "epoch": 0.5168246886038297,
-      "grad_norm": 0.2833311975297903,
+      "grad_norm": 0.13967410850834808,
       "learning_rate": 0.0001115480113054626,
-      "loss": 3.6531,
+      "loss": 1.8652,
       "step": 4170
     },
     {
       "epoch": 0.5174443824750573,
-      "grad_norm": 0.25698009648870623,
+      "grad_norm": 0.13812711555448426,
       "learning_rate": 0.00011133309843316669,
-      "loss": 3.5906,
+      "loss": 1.8859,
       "step": 4175
     },
     {
       "epoch": 0.5180640763462849,
-      "grad_norm": 0.2628386812811596,
+      "grad_norm": 0.1370329050816476,
       "learning_rate": 0.0001111181325218653,
-      "loss": 3.5805,
+      "loss": 1.7641,
       "step": 4180
     },
     {
       "epoch": 0.5186837702175126,
-      "grad_norm": 0.27066938247651806,
+      "grad_norm": 0.13608145182115125,
       "learning_rate": 0.00011090311457760094,
-      "loss": 3.6664,
+      "loss": 1.8043,
       "step": 4185
     },
     {
       "epoch": 0.5193034640887402,
-      "grad_norm": 0.27158751271221065,
+      "grad_norm": 0.134999506527859,
       "learning_rate": 0.0001106880456066595,
-      "loss": 3.5961,
+      "loss": 1.7937,
       "step": 4190
     },
     {
       "epoch": 0.5199231579599678,
-      "grad_norm": 0.2752422036574746,
+      "grad_norm": 0.12986685167360099,
       "learning_rate": 0.00011047292661556581,
-      "loss": 3.7078,
+      "loss": 1.8008,
       "step": 4195
     },
     {
       "epoch": 0.5205428518311954,
-      "grad_norm": 0.2641601154800112,
+      "grad_norm": 0.1344755084959134,
       "learning_rate": 0.00011025775861107874,
-      "loss": 3.7266,
+      "loss": 1.8262,
       "step": 4200
     },
     {
       "epoch": 0.521162545702423,
-      "grad_norm": 0.2665093713894038,
+      "grad_norm": 0.13044935784007144,
       "learning_rate": 0.00011004254260018648,
-      "loss": 3.7969,
+      "loss": 1.8434,
       "step": 4205
     },
     {
       "epoch": 0.5217822395736507,
-      "grad_norm": 0.27287397250323636,
+      "grad_norm": 0.13877368447780922,
       "learning_rate": 0.00010982727959010201,
-      "loss": 3.7023,
+      "loss": 1.8852,
       "step": 4210
     },
     {
       "epoch": 0.5224019334448782,
-      "grad_norm": 0.2550715980357899,
+      "grad_norm": 0.12973163353638237,
       "learning_rate": 0.00010961197058825817,
-      "loss": 3.6227,
+      "loss": 1.8676,
       "step": 4215
     },
     {
       "epoch": 0.5230216273161058,
-      "grad_norm": 0.2620907560500356,
+      "grad_norm": 0.13199230862465572,
       "learning_rate": 0.00010939661660230309,
-      "loss": 3.7016,
+      "loss": 1.8328,
       "step": 4220
     },
     {
       "epoch": 0.5236413211873334,
-      "grad_norm": 0.26858792866309283,
+      "grad_norm": 0.13376789718860482,
       "learning_rate": 0.00010918121864009543,
-      "loss": 3.7938,
+      "loss": 1.8305,
       "step": 4225
     },
     {
       "epoch": 0.524261015058561,
-      "grad_norm": 0.2800359718630766,
+      "grad_norm": 0.13923517431663307,
       "learning_rate": 0.00010896577770969964,
-      "loss": 3.6867,
+      "loss": 1.8027,
       "step": 4230
     },
     {
       "epoch": 0.5248807089297887,
-      "grad_norm": 0.2571645347786101,
+      "grad_norm": 0.13044002358446727,
       "learning_rate": 0.00010875029481938126,
-      "loss": 3.6711,
+      "loss": 1.9016,
       "step": 4235
     },
     {
       "epoch": 0.5255004028010163,
-      "grad_norm": 0.2608272606052285,
+      "grad_norm": 0.1299472760176014,
       "learning_rate": 0.00010853477097760222,
-      "loss": 3.6078,
+      "loss": 1.8285,
       "step": 4240
     },
     {
       "epoch": 0.5261200966722439,
-      "grad_norm": 0.26451451779524215,
+      "grad_norm": 0.13239585756853023,
       "learning_rate": 0.0001083192071930161,
-      "loss": 3.7781,
+      "loss": 1.8016,
       "step": 4245
     },
     {
       "epoch": 0.5267397905434715,
-      "grad_norm": 0.26351086041096267,
+      "grad_norm": 0.12416531787933346,
       "learning_rate": 0.00010810360447446335,
-      "loss": 3.7883,
+      "loss": 1.852,
       "step": 4250
     },
     {
       "epoch": 0.5273594844146992,
-      "grad_norm": 0.28221768245515083,
+      "grad_norm": 0.1379243677220307,
       "learning_rate": 0.00010788796383096676,
-      "loss": 3.6695,
+      "loss": 1.8152,
       "step": 4255
     },
     {
       "epoch": 0.5279791782859268,
-      "grad_norm": 0.2639447901747608,
+      "grad_norm": 0.12963488868452042,
       "learning_rate": 0.00010767228627172645,
-      "loss": 3.8016,
+      "loss": 1.8582,
       "step": 4260
     },
     {
       "epoch": 0.5285988721571544,
-      "grad_norm": 0.2679083471470788,
+      "grad_norm": 0.1328002434282655,
       "learning_rate": 0.00010745657280611552,
-      "loss": 3.6914,
+      "loss": 1.8641,
       "step": 4265
     },
     {
       "epoch": 0.529218566028382,
-      "grad_norm": 0.27311699346488155,
+      "grad_norm": 0.13106171306296183,
       "learning_rate": 0.00010724082444367485,
-      "loss": 3.7383,
+      "loss": 1.8445,
       "step": 4270
     },
     {
       "epoch": 0.5298382598996096,
-      "grad_norm": 0.27104674035190124,
+      "grad_norm": 0.13741774367232246,
       "learning_rate": 0.00010702504219410884,
-      "loss": 3.6758,
+      "loss": 1.8777,
       "step": 4275
     },
     {
       "epoch": 0.5304579537708372,
-      "grad_norm": 0.2681807255742581,
+      "grad_norm": 0.13098908649184046,
       "learning_rate": 0.00010680922706728041,
-      "loss": 3.6836,
+      "loss": 1.7973,
       "step": 4280
     },
     {
       "epoch": 0.5310776476420648,
-      "grad_norm": 0.2704989786880266,
+      "grad_norm": 0.13693787562687978,
       "learning_rate": 0.00010659338007320632,
-      "loss": 3.6414,
+      "loss": 1.9187,
       "step": 4285
     },
     {
       "epoch": 0.5316973415132924,
-      "grad_norm": 0.2656517152994718,
+      "grad_norm": 0.13261045311512387,
       "learning_rate": 0.00010637750222205253,
-      "loss": 3.7219,
+      "loss": 1.8277,
       "step": 4290
     },
     {
       "epoch": 0.53231703538452,
-      "grad_norm": 0.27147564166779353,
+      "grad_norm": 0.1361721967821798,
       "learning_rate": 0.00010616159452412939,
-      "loss": 3.6375,
+      "loss": 1.8301,
       "step": 4295
     },
     {
       "epoch": 0.5329367292557476,
-      "grad_norm": 0.26355038804511355,
+      "grad_norm": 0.1330959419108472,
       "learning_rate": 0.00010594565798988689,
-      "loss": 3.6742,
+      "loss": 1.8496,
       "step": 4300
     },
     {
       "epoch": 0.5335564231269753,
-      "grad_norm": 0.2670813512281771,
+      "grad_norm": 0.14146841910755462,
       "learning_rate": 0.00010572969362990998,
-      "loss": 3.6531,
+      "loss": 1.8523,
       "step": 4305
     },
     {
       "epoch": 0.5341761169982029,
-      "grad_norm": 0.26063814209234343,
+      "grad_norm": 0.1335707575288798,
       "learning_rate": 0.00010551370245491394,
-      "loss": 3.6086,
+      "loss": 1.8457,
       "step": 4310
     },
     {
       "epoch": 0.5347958108694305,
-      "grad_norm": 0.2844653210765238,
+      "grad_norm": 0.1337012914277915,
       "learning_rate": 0.00010529768547573942,
-      "loss": 3.682,
+      "loss": 1.8125,
       "step": 4315
     },
     {
       "epoch": 0.5354155047406581,
-      "grad_norm": 0.26651449146907025,
+      "grad_norm": 0.13666744181424356,
       "learning_rate": 0.00010508164370334787,
-      "loss": 3.7648,
+      "loss": 1.8461,
       "step": 4320
     },
     {
       "epoch": 0.5360351986118858,
-      "grad_norm": 0.29026927384527157,
+      "grad_norm": 0.13958679117203346,
       "learning_rate": 0.00010486557814881686,
-      "loss": 3.6891,
+      "loss": 1.8363,
       "step": 4325
     },
     {
       "epoch": 0.5366548924831134,
-      "grad_norm": 0.2951899420486929,
+      "grad_norm": 0.13607366667952386,
       "learning_rate": 0.00010464948982333504,
-      "loss": 3.4922,
+      "loss": 1.8137,
       "step": 4330
     },
     {
       "epoch": 0.537274586354341,
-      "grad_norm": 0.2681518077012756,
+      "grad_norm": 0.1350440016774684,
       "learning_rate": 0.00010443337973819791,
-      "loss": 3.6727,
+      "loss": 1.8121,
       "step": 4335
     },
     {
       "epoch": 0.5378942802255686,
-      "grad_norm": 0.2677952279703973,
+      "grad_norm": 0.1322464115487655,
       "learning_rate": 0.00010421724890480258,
-      "loss": 3.6484,
+      "loss": 1.8227,
       "step": 4340
     },
     {
       "epoch": 0.5385139740967961,
-      "grad_norm": 0.26992711864448443,
+      "grad_norm": 0.13809058095945276,
       "learning_rate": 0.00010400109833464338,
-      "loss": 3.6117,
+      "loss": 1.8594,
       "step": 4345
     },
     {
       "epoch": 0.5391336679680238,
-      "grad_norm": 0.27337921580576696,
+      "grad_norm": 0.13052069620811604,
       "learning_rate": 0.00010378492903930699,
-      "loss": 3.6383,
+      "loss": 1.825,
       "step": 4350
     },
     {
       "epoch": 0.5397533618392514,
-      "grad_norm": 0.2624992989486429,
+      "grad_norm": 0.14779827571036525,
       "learning_rate": 0.00010356874203046766,
-      "loss": 3.6789,
+      "loss": 1.8391,
       "step": 4355
     },
     {
       "epoch": 0.540373055710479,
-      "grad_norm": 0.27710421143802777,
+      "grad_norm": 0.12710789410858492,
       "learning_rate": 0.00010335253831988267,
-      "loss": 3.4875,
+      "loss": 1.8184,
       "step": 4360
     },
     {
       "epoch": 0.5409927495817066,
-      "grad_norm": 0.2826706906139555,
+      "grad_norm": 0.1285834215695178,
       "learning_rate": 0.00010313631891938736,
-      "loss": 3.6797,
+      "loss": 1.8633,
       "step": 4365
     },
     {
       "epoch": 0.5416124434529342,
-      "grad_norm": 0.2707904993711854,
+      "grad_norm": 0.13350900347998482,
       "learning_rate": 0.00010292008484089047,
-      "loss": 3.7258,
+      "loss": 1.8516,
       "step": 4370
     },
     {
       "epoch": 0.5422321373241619,
-      "grad_norm": 0.2734811974169758,
+      "grad_norm": 0.1320142499534893,
       "learning_rate": 0.0001027038370963695,
-      "loss": 3.8141,
+      "loss": 1.8543,
       "step": 4375
     },
     {
       "epoch": 0.5428518311953895,
-      "grad_norm": 0.26637771423373513,
+      "grad_norm": 0.13382411859073567,
       "learning_rate": 0.00010248757669786594,
-      "loss": 3.7797,
+      "loss": 1.8438,
       "step": 4380
     },
     {
       "epoch": 0.5434715250666171,
-      "grad_norm": 0.28036695867937245,
+      "grad_norm": 0.15096282450997914,
       "learning_rate": 0.00010227130465748045,
-      "loss": 3.6367,
+      "loss": 1.8844,
       "step": 4385
     },
     {
       "epoch": 0.5440912189378447,
-      "grad_norm": 0.2641150707890887,
+      "grad_norm": 0.13050426547539765,
       "learning_rate": 0.00010205502198736816,
-      "loss": 3.6641,
+      "loss": 1.7953,
       "step": 4390
     },
     {
       "epoch": 0.5447109128090724,
-      "grad_norm": 0.2814260608847561,
+      "grad_norm": 0.12786255907842825,
       "learning_rate": 0.00010183872969973396,
-      "loss": 3.7086,
+      "loss": 1.8191,
       "step": 4395
     },
     {
       "epoch": 0.5453306066803,
-      "grad_norm": 0.2682954596928813,
+      "grad_norm": 0.14113604167832816,
       "learning_rate": 0.00010162242880682776,
-      "loss": 3.5836,
+      "loss": 1.8598,
       "step": 4400
     },
     {
       "epoch": 0.5459503005515276,
-      "grad_norm": 0.27065092520145695,
+      "grad_norm": 0.1341328866770354,
       "learning_rate": 0.00010140612032093972,
-      "loss": 3.675,
+      "loss": 1.825,
       "step": 4405
     },
     {
       "epoch": 0.5465699944227551,
-      "grad_norm": 0.2698049008297175,
+      "grad_norm": 0.13380233821056914,
       "learning_rate": 0.00010118980525439559,
-      "loss": 3.6375,
+      "loss": 1.843,
       "step": 4410
     },
     {
       "epoch": 0.5471896882939827,
-      "grad_norm": 0.2542929379997657,
+      "grad_norm": 0.1343984620510971,
       "learning_rate": 0.00010097348461955186,
-      "loss": 3.7352,
+      "loss": 1.7957,
       "step": 4415
     },
     {
       "epoch": 0.5478093821652104,
-      "grad_norm": 0.2557245988208345,
+      "grad_norm": 0.1351500470677798,
       "learning_rate": 0.00010075715942879114,
-      "loss": 3.5656,
+      "loss": 1.8129,
       "step": 4420
     },
     {
       "epoch": 0.548429076036438,
-      "grad_norm": 0.27292418147518904,
+      "grad_norm": 0.13404154577268396,
       "learning_rate": 0.00010054083069451728,
-      "loss": 3.5719,
+      "loss": 1.8422,
       "step": 4425
     },
     {
       "epoch": 0.5490487699076656,
-      "grad_norm": 0.26425016532979473,
+      "grad_norm": 0.13166534889000214,
       "learning_rate": 0.00010032449942915072,
-      "loss": 3.6578,
+      "loss": 1.8156,
       "step": 4430
     },
     {
       "epoch": 0.5496684637788932,
-      "grad_norm": 0.27604141996830334,
+      "grad_norm": 0.13506639450706173,
       "learning_rate": 0.00010010816664512389,
-      "loss": 3.718,
+      "loss": 1.8012,
       "step": 4435
     },
     {
       "epoch": 0.5502881576501208,
-      "grad_norm": 0.289694464188095,
+      "grad_norm": 0.13182983003356816,
       "learning_rate": 9.989183335487615e-05,
-      "loss": 3.7164,
+      "loss": 1.8676,
       "step": 4440
     },
     {
       "epoch": 0.5509078515213485,
-      "grad_norm": 0.2600073920423731,
+      "grad_norm": 0.14142433786552008,
       "learning_rate": 9.96755005708493e-05,
-      "loss": 3.6734,
+      "loss": 1.8723,
       "step": 4445
     },
     {
       "epoch": 0.5515275453925761,
-      "grad_norm": 0.27299233498222497,
+      "grad_norm": 0.1334151262150459,
       "learning_rate": 9.945916930548276e-05,
-      "loss": 3.6578,
+      "loss": 1.9223,
       "step": 4450
     },
     {
       "epoch": 0.5521472392638037,
-      "grad_norm": 0.2825950975586584,
+      "grad_norm": 0.13047834012682685,
       "learning_rate": 9.924284057120888e-05,
-      "loss": 3.6187,
+      "loss": 1.8035,
       "step": 4455
     },
     {
       "epoch": 0.5527669331350313,
-      "grad_norm": 0.2690506554715704,
+      "grad_norm": 0.13551238503921087,
       "learning_rate": 9.902651538044814e-05,
-      "loss": 3.6023,
+      "loss": 1.8793,
       "step": 4460
     },
     {
       "epoch": 0.553386627006259,
-      "grad_norm": 0.2692063879106752,
+      "grad_norm": 0.13009663685025677,
       "learning_rate": 9.88101947456044e-05,
-      "loss": 3.657,
+      "loss": 1.8641,
       "step": 4465
     },
     {
       "epoch": 0.5540063208774865,
-      "grad_norm": 0.2760619915971685,
+      "grad_norm": 0.1360455529484941,
       "learning_rate": 9.859387967906033e-05,
-      "loss": 3.657,
+      "loss": 1.8391,
       "step": 4470
     },
     {
       "epoch": 0.5546260147487141,
-      "grad_norm": 0.2753779861348704,
+      "grad_norm": 0.1289716476183338,
       "learning_rate": 9.837757119317228e-05,
-      "loss": 3.6414,
+      "loss": 1.8742,
       "step": 4475
     },
     {
       "epoch": 0.5552457086199417,
-      "grad_norm": 0.27011777687757504,
+      "grad_norm": 0.13275590576368201,
       "learning_rate": 9.816127030026607e-05,
-      "loss": 3.8336,
+      "loss": 1.8613,
       "step": 4480
     },
     {
       "epoch": 0.5558654024911693,
-      "grad_norm": 0.2720905676563954,
+      "grad_norm": 0.14348224249868619,
       "learning_rate": 9.794497801263185e-05,
-      "loss": 3.6656,
+      "loss": 1.8391,
       "step": 4485
     },
     {
       "epoch": 0.556485096362397,
-      "grad_norm": 0.2666666192966718,
+      "grad_norm": 0.13464069972617826,
       "learning_rate": 9.772869534251956e-05,
-      "loss": 3.643,
+      "loss": 1.827,
       "step": 4490
     },
     {
       "epoch": 0.5571047902336246,
-      "grad_norm": 0.27442530437873,
+      "grad_norm": 0.13214411129137058,
       "learning_rate": 9.751242330213407e-05,
-      "loss": 3.5914,
+      "loss": 1.8836,
       "step": 4495
     },
     {
       "epoch": 0.5577244841048522,
-      "grad_norm": 0.2762518832157627,
+      "grad_norm": 0.13401560105396315,
       "learning_rate": 9.729616290363051e-05,
-      "loss": 3.6383,
+      "loss": 1.8488,
       "step": 4500
     },
     {
       "epoch": 0.5583441779760798,
-      "grad_norm": 0.2719838257171385,
+      "grad_norm": 0.13560522706619668,
       "learning_rate": 9.707991515910957e-05,
-      "loss": 3.757,
+      "loss": 1.8621,
       "step": 4505
     },
     {
       "epoch": 0.5589638718473074,
-      "grad_norm": 0.2665995820731799,
+      "grad_norm": 0.1378604398311812,
       "learning_rate": 9.686368108061268e-05,
-      "loss": 3.6352,
+      "loss": 1.9074,
       "step": 4510
     },
     {
       "epoch": 0.5595835657185351,
-      "grad_norm": 0.2793809708656953,
+      "grad_norm": 0.12967243553989305,
       "learning_rate": 9.664746168011734e-05,
-      "loss": 3.5305,
+      "loss": 1.8211,
       "step": 4515
     },
     {
       "epoch": 0.5602032595897627,
-      "grad_norm": 0.27282729383029186,
+      "grad_norm": 0.1273548051846235,
       "learning_rate": 9.643125796953233e-05,
-      "loss": 3.7062,
+      "loss": 1.8742,
       "step": 4520
     },
     {
       "epoch": 0.5608229534609903,
-      "grad_norm": 0.26852576393578903,
+      "grad_norm": 0.13392981702989407,
       "learning_rate": 9.621507096069304e-05,
-      "loss": 3.6648,
+      "loss": 1.8578,
       "step": 4525
     },
     {
       "epoch": 0.5614426473322179,
-      "grad_norm": 0.27059227818607917,
+      "grad_norm": 0.13085951763754228,
       "learning_rate": 9.599890166535665e-05,
-      "loss": 3.5891,
+      "loss": 1.8684,
       "step": 4530
     },
     {
       "epoch": 0.5620623412034454,
-      "grad_norm": 0.26232970416408835,
+      "grad_norm": 0.14167378893366092,
       "learning_rate": 9.578275109519744e-05,
-      "loss": 3.6594,
+      "loss": 1.893,
       "step": 4535
     },
     {
       "epoch": 0.5626820350746731,
-      "grad_norm": 0.27443731038719305,
+      "grad_norm": 0.13395690664276264,
       "learning_rate": 9.556662026180212e-05,
-      "loss": 3.5516,
+      "loss": 1.8559,
       "step": 4540
     },
     {
       "epoch": 0.5633017289459007,
-      "grad_norm": 0.2618687741186096,
+      "grad_norm": 0.13785153538830483,
       "learning_rate": 9.535051017666497e-05,
-      "loss": 3.6773,
+      "loss": 1.8742,
       "step": 4545
     },
     {
       "epoch": 0.5639214228171283,
-      "grad_norm": 0.2684216768813871,
+      "grad_norm": 0.13326090153428086,
       "learning_rate": 9.513442185118319e-05,
-      "loss": 3.7344,
+      "loss": 1.8723,
       "step": 4550
     },
     {
       "epoch": 0.5645411166883559,
-      "grad_norm": 0.28067208731324905,
+      "grad_norm": 0.1325737274991862,
       "learning_rate": 9.491835629665214e-05,
-      "loss": 3.6211,
+      "loss": 1.8371,
       "step": 4555
     },
     {
       "epoch": 0.5651608105595836,
-      "grad_norm": 0.2670591286203052,
+      "grad_norm": 0.13484451999521505,
       "learning_rate": 9.470231452426059e-05,
-      "loss": 3.5281,
+      "loss": 1.7844,
       "step": 4560
     },
     {
       "epoch": 0.5657805044308112,
-      "grad_norm": 0.26120176936520656,
+      "grad_norm": 0.13667668532140465,
       "learning_rate": 9.448629754508607e-05,
-      "loss": 3.7922,
+      "loss": 1.8711,
       "step": 4565
     },
     {
       "epoch": 0.5664001983020388,
-      "grad_norm": 0.2653916879077019,
+      "grad_norm": 0.13989769457594814,
       "learning_rate": 9.427030637009003e-05,
-      "loss": 3.7422,
+      "loss": 1.8859,
       "step": 4570
     },
     {
       "epoch": 0.5670198921732664,
-      "grad_norm": 0.27047334889618924,
+      "grad_norm": 0.12883812907785516,
       "learning_rate": 9.405434201011313e-05,
-      "loss": 3.6961,
+      "loss": 1.8988,
       "step": 4575
     },
     {
       "epoch": 0.567639586044494,
-      "grad_norm": 0.26233721834749346,
+      "grad_norm": 0.130205864302805,
       "learning_rate": 9.383840547587064e-05,
-      "loss": 3.7984,
+      "loss": 1.8863,
       "step": 4580
     },
     {
       "epoch": 0.5682592799157217,
-      "grad_norm": 0.2671699568255034,
+      "grad_norm": 0.1333369490609614,
       "learning_rate": 9.362249777794749e-05,
-      "loss": 3.6398,
+      "loss": 1.8434,
       "step": 4585
     },
     {
       "epoch": 0.5688789737869493,
-      "grad_norm": 0.3075923924701826,
+      "grad_norm": 0.13277669705453948,
       "learning_rate": 9.340661992679369e-05,
-      "loss": 3.675,
+      "loss": 1.8246,
       "step": 4590
     },
     {
       "epoch": 0.5694986676581769,
-      "grad_norm": 0.29037913669304627,
+      "grad_norm": 0.1388883320912154,
       "learning_rate": 9.319077293271961e-05,
-      "loss": 3.7156,
+      "loss": 1.8754,
       "step": 4595
     },
     {
       "epoch": 0.5701183615294044,
-      "grad_norm": 0.27370521659421365,
+      "grad_norm": 0.1320971868111603,
       "learning_rate": 9.297495780589118e-05,
-      "loss": 3.5,
+      "loss": 1.85,
       "step": 4600
     },
     {
       "epoch": 0.570738055400632,
-      "grad_norm": 0.25765783085364835,
+      "grad_norm": 0.13017211313722013,
       "learning_rate": 9.27591755563252e-05,
-      "loss": 3.5992,
+      "loss": 1.7656,
       "step": 4605
     },
     {
       "epoch": 0.5713577492718597,
-      "grad_norm": 0.26943210575764764,
+      "grad_norm": 0.129888126477682,
       "learning_rate": 9.254342719388454e-05,
-      "loss": 3.6242,
+      "loss": 1.7785,
       "step": 4610
     },
     {
       "epoch": 0.5719774431430873,
-      "grad_norm": 0.2745748208620268,
+      "grad_norm": 0.14770097601386062,
       "learning_rate": 9.232771372827356e-05,
-      "loss": 3.6227,
+      "loss": 1.877,
       "step": 4615
     },
     {
       "epoch": 0.5725971370143149,
-      "grad_norm": 0.26354715484573016,
+      "grad_norm": 0.1383845014722268,
       "learning_rate": 9.211203616903328e-05,
-      "loss": 3.5875,
+      "loss": 1.8395,
       "step": 4620
     },
     {
       "epoch": 0.5732168308855425,
-      "grad_norm": 0.2850816926028819,
+      "grad_norm": 0.12825268242494714,
       "learning_rate": 9.189639552553667e-05,
-      "loss": 3.6516,
+      "loss": 1.898,
       "step": 4625
     },
     {
       "epoch": 0.5738365247567702,
-      "grad_norm": 0.2727825297247378,
+      "grad_norm": 0.14095823402273389,
       "learning_rate": 9.168079280698391e-05,
-      "loss": 3.7016,
+      "loss": 1.8254,
       "step": 4630
     },
     {
       "epoch": 0.5744562186279978,
-      "grad_norm": 0.26533216629165046,
+      "grad_norm": 0.13138557864881723,
       "learning_rate": 9.146522902239781e-05,
-      "loss": 3.6914,
+      "loss": 1.8512,
       "step": 4635
     },
     {
       "epoch": 0.5750759124992254,
-      "grad_norm": 0.2759553912069646,
+      "grad_norm": 0.1275168924292215,
       "learning_rate": 9.124970518061877e-05,
-      "loss": 3.5375,
+      "loss": 1.8637,
       "step": 4640
     },
     {
       "epoch": 0.575695606370453,
-      "grad_norm": 0.2706374034574014,
+      "grad_norm": 0.1301661963978751,
       "learning_rate": 9.103422229030038e-05,
-      "loss": 3.7633,
+      "loss": 1.8434,
       "step": 4645
     },
     {
       "epoch": 0.5763153002416807,
-      "grad_norm": 0.2730578135363883,
+      "grad_norm": 0.13374425189552858,
       "learning_rate": 9.081878135990458e-05,
-      "loss": 3.7047,
+      "loss": 1.7465,
       "step": 4650
     },
     {
       "epoch": 0.5769349941129083,
-      "grad_norm": 0.25414420934326987,
+      "grad_norm": 0.13808982686111124,
       "learning_rate": 9.06033833976969e-05,
-      "loss": 3.7484,
+      "loss": 1.8422,
       "step": 4655
     },
     {
       "epoch": 0.5775546879841358,
-      "grad_norm": 0.256400849310976,
+      "grad_norm": 0.13447543310447194,
       "learning_rate": 9.038802941174187e-05,
-      "loss": 3.7687,
+      "loss": 1.7785,
       "step": 4660
     },
     {
       "epoch": 0.5781743818553634,
-      "grad_norm": 0.29634415084122473,
+      "grad_norm": 0.13300414692423948,
       "learning_rate": 9.017272040989804e-05,
-      "loss": 3.8375,
+      "loss": 1.8574,
       "step": 4665
     },
     {
       "epoch": 0.578794075726591,
-      "grad_norm": 0.27794653757814963,
+      "grad_norm": 0.13096590711343975,
       "learning_rate": 8.995745739981355e-05,
-      "loss": 3.7055,
+      "loss": 1.8609,
       "step": 4670
     },
     {
       "epoch": 0.5794137695978187,
-      "grad_norm": 0.27219582612115456,
+      "grad_norm": 0.13820508880678925,
       "learning_rate": 8.974224138892127e-05,
-      "loss": 3.807,
+      "loss": 1.8523,
       "step": 4675
     },
     {
       "epoch": 0.5800334634690463,
-      "grad_norm": 0.2717849560452208,
+      "grad_norm": 0.12765070621427313,
       "learning_rate": 8.952707338443418e-05,
-      "loss": 3.6008,
+      "loss": 1.8145,
       "step": 4680
     },
     {
       "epoch": 0.5806531573402739,
-      "grad_norm": 0.2689613705566208,
+      "grad_norm": 0.13872971817201188,
       "learning_rate": 8.931195439334048e-05,
-      "loss": 3.6508,
+      "loss": 1.8512,
       "step": 4685
     },
     {
       "epoch": 0.5812728512115015,
-      "grad_norm": 0.2679893379960508,
+      "grad_norm": 0.13654187286164757,
       "learning_rate": 8.90968854223991e-05,
-      "loss": 3.6219,
+      "loss": 1.8188,
       "step": 4690
     },
     {
       "epoch": 0.5818925450827291,
-      "grad_norm": 0.2720978068136715,
+      "grad_norm": 0.1358565435204054,
       "learning_rate": 8.888186747813473e-05,
-      "loss": 3.5727,
+      "loss": 1.8594,
       "step": 4695
     },
     {
       "epoch": 0.5825122389539568,
-      "grad_norm": 0.2790232242548787,
+      "grad_norm": 0.1361212793989369,
       "learning_rate": 8.866690156683332e-05,
-      "loss": 3.6594,
+      "loss": 1.825,
       "step": 4700
     },
     {
       "epoch": 0.5831319328251844,
-      "grad_norm": 0.2550213840737251,
+      "grad_norm": 0.1313054079631678,
       "learning_rate": 8.845198869453742e-05,
-      "loss": 3.7766,
+      "loss": 1.8617,
       "step": 4705
     },
     {
       "epoch": 0.583751626696412,
-      "grad_norm": 0.27335294108007446,
+      "grad_norm": 0.1340543850970045,
       "learning_rate": 8.823712986704121e-05,
-      "loss": 3.6156,
+      "loss": 1.7836,
       "step": 4710
     },
     {
       "epoch": 0.5843713205676396,
-      "grad_norm": 0.25931700709755595,
+      "grad_norm": 0.13217833814152594,
       "learning_rate": 8.802232608988604e-05,
-      "loss": 3.7563,
+      "loss": 1.832,
       "step": 4715
     },
     {
       "epoch": 0.5849910144388673,
-      "grad_norm": 0.26874870206277024,
+      "grad_norm": 0.14629328645282433,
       "learning_rate": 8.780757836835544e-05,
-      "loss": 3.6766,
+      "loss": 1.823,
       "step": 4720
     },
     {
       "epoch": 0.5856107083100948,
-      "grad_norm": 0.2707225242276941,
+      "grad_norm": 0.13494154906279057,
       "learning_rate": 8.759288770747087e-05,
-      "loss": 3.7711,
+      "loss": 1.8539,
       "step": 4725
     },
     {
       "epoch": 0.5862304021813224,
-      "grad_norm": 0.2651627038636793,
+      "grad_norm": 0.14440441221946163,
       "learning_rate": 8.737825511198654e-05,
-      "loss": 3.6562,
+      "loss": 1.8441,
       "step": 4730
     },
     {
       "epoch": 0.58685009605255,
-      "grad_norm": 0.2785945180431172,
+      "grad_norm": 0.13863154183602425,
       "learning_rate": 8.7163681586385e-05,
-      "loss": 3.6945,
+      "loss": 1.8242,
       "step": 4735
     },
     {
       "epoch": 0.5874697899237776,
-      "grad_norm": 0.2653182542622149,
+      "grad_norm": 0.1318855989670804,
       "learning_rate": 8.694916813487233e-05,
-      "loss": 3.632,
+      "loss": 1.834,
       "step": 4740
     },
     {
       "epoch": 0.5880894837950053,
-      "grad_norm": 0.26490316164430944,
+      "grad_norm": 0.13920618174796423,
       "learning_rate": 8.67347157613735e-05,
-      "loss": 3.65,
+      "loss": 1.7969,
       "step": 4745
     },
     {
       "epoch": 0.5887091776662329,
-      "grad_norm": 0.2709427564963539,
+      "grad_norm": 0.13767154490310451,
       "learning_rate": 8.652032546952754e-05,
-      "loss": 3.693,
+      "loss": 1.807,
       "step": 4750
     },
     {
       "epoch": 0.5893288715374605,
-      "grad_norm": 0.2748718169619932,
+      "grad_norm": 0.1354204158207126,
       "learning_rate": 8.630599826268303e-05,
-      "loss": 3.757,
+      "loss": 1.9137,
       "step": 4755
     },
     {
       "epoch": 0.5899485654086881,
-      "grad_norm": 0.30416816756324155,
+      "grad_norm": 0.13637608173257612,
       "learning_rate": 8.609173514389328e-05,
-      "loss": 3.618,
+      "loss": 1.8375,
       "step": 4760
     },
     {
       "epoch": 0.5905682592799157,
-      "grad_norm": 0.2739395450601053,
+      "grad_norm": 0.1392456473512201,
       "learning_rate": 8.587753711591166e-05,
-      "loss": 3.7812,
+      "loss": 1.7957,
       "step": 4765
     },
     {
       "epoch": 0.5911879531511434,
-      "grad_norm": 0.26917898434493065,
+      "grad_norm": 0.14079378358935743,
       "learning_rate": 8.566340518118695e-05,
-      "loss": 3.7789,
+      "loss": 1.8789,
       "step": 4770
     },
     {
       "epoch": 0.591807647022371,
-      "grad_norm": 0.26497623397022974,
+      "grad_norm": 0.1326673419979531,
       "learning_rate": 8.544934034185854e-05,
-      "loss": 3.6297,
+      "loss": 1.8301,
       "step": 4775
     },
     {
       "epoch": 0.5924273408935986,
-      "grad_norm": 0.257955707351069,
+      "grad_norm": 0.1324300142166372,
       "learning_rate": 8.523534359975189e-05,
-      "loss": 3.6953,
+      "loss": 1.8129,
       "step": 4780
     },
     {
       "epoch": 0.5930470347648262,
-      "grad_norm": 0.26226027788634904,
+      "grad_norm": 0.13436159136571374,
       "learning_rate": 8.502141595637371e-05,
-      "loss": 3.6797,
+      "loss": 1.8234,
       "step": 4785
     },
     {
       "epoch": 0.5936667286360537,
-      "grad_norm": 0.2695043028263795,
+      "grad_norm": 0.1340824171889328,
       "learning_rate": 8.480755841290736e-05,
-      "loss": 3.6016,
+      "loss": 1.8719,
       "step": 4790
     },
     {
       "epoch": 0.5942864225072814,
-      "grad_norm": 0.2710516349098267,
+      "grad_norm": 0.13942656890716446,
       "learning_rate": 8.459377197020813e-05,
-      "loss": 3.7289,
+      "loss": 1.8418,
       "step": 4795
     },
     {
       "epoch": 0.594906116378509,
-      "grad_norm": 0.25498187017040835,
+      "grad_norm": 0.1375776284383645,
       "learning_rate": 8.438005762879856e-05,
-      "loss": 3.7164,
+      "loss": 1.8352,
       "step": 4800
     },
     {
       "epoch": 0.5955258102497366,
-      "grad_norm": 0.2662837360767922,
+      "grad_norm": 0.13817749230266843,
       "learning_rate": 8.416641638886369e-05,
-      "loss": 3.582,
+      "loss": 1.8715,
       "step": 4805
     },
     {
       "epoch": 0.5961455041209642,
-      "grad_norm": 0.2700127101757652,
+      "grad_norm": 0.13644813853970703,
       "learning_rate": 8.395284925024654e-05,
-      "loss": 3.5852,
+      "loss": 1.8141,
       "step": 4810
     },
     {
       "epoch": 0.5967651979921919,
-      "grad_norm": 0.2678327716698091,
+      "grad_norm": 0.14009468198120875,
       "learning_rate": 8.373935721244329e-05,
-      "loss": 3.8133,
+      "loss": 1.8398,
       "step": 4815
     },
     {
       "epoch": 0.5973848918634195,
-      "grad_norm": 0.26206854136484303,
+      "grad_norm": 0.13539696078697683,
       "learning_rate": 8.352594127459865e-05,
-      "loss": 3.7641,
+      "loss": 1.7586,
       "step": 4820
     },
     {
       "epoch": 0.5980045857346471,
-      "grad_norm": 0.27440252125300807,
+      "grad_norm": 0.1366948032582978,
       "learning_rate": 8.331260243550119e-05,
-      "loss": 3.6336,
+      "loss": 1.8266,
       "step": 4825
     },
     {
       "epoch": 0.5986242796058747,
-      "grad_norm": 0.27610081992251445,
+      "grad_norm": 0.12944284870629252,
       "learning_rate": 8.309934169357862e-05,
-      "loss": 3.6305,
+      "loss": 1.8117,
       "step": 4830
     },
     {
       "epoch": 0.5992439734771023,
-      "grad_norm": 0.27199189429278114,
+      "grad_norm": 0.13536533441301316,
       "learning_rate": 8.28861600468932e-05,
-      "loss": 3.7133,
+      "loss": 1.7957,
       "step": 4835
     },
     {
       "epoch": 0.59986366734833,
-      "grad_norm": 0.26713808906578573,
+      "grad_norm": 0.13946065741059344,
       "learning_rate": 8.267305849313702e-05,
-      "loss": 3.85,
+      "loss": 1.8605,
       "step": 4840
     },
     {
       "epoch": 0.6004833612195576,
-      "grad_norm": 0.2646602068441273,
+      "grad_norm": 0.1389191099799506,
       "learning_rate": 8.246003802962732e-05,
-      "loss": 3.7367,
+      "loss": 1.7836,
       "step": 4845
     },
     {
       "epoch": 0.6011030550907851,
-      "grad_norm": 0.2834428909362969,
+      "grad_norm": 0.1379459598325412,
       "learning_rate": 8.224709965330182e-05,
-      "loss": 3.6453,
+      "loss": 1.9152,
       "step": 4850
     },
     {
       "epoch": 0.6017227489620127,
-      "grad_norm": 0.26601716548743887,
+      "grad_norm": 0.13789049940846776,
       "learning_rate": 8.203424436071413e-05,
-      "loss": 3.6734,
+      "loss": 1.8289,
       "step": 4855
     },
     {
       "epoch": 0.6023424428332403,
-      "grad_norm": 0.27536953883017307,
+      "grad_norm": 0.13702308254276288,
       "learning_rate": 8.182147314802892e-05,
-      "loss": 3.6914,
+      "loss": 1.8297,
       "step": 4860
     },
     {
       "epoch": 0.602962136704468,
-      "grad_norm": 0.2863152536982749,
+      "grad_norm": 0.13119036631588685,
       "learning_rate": 8.160878701101751e-05,
-      "loss": 3.5758,
+      "loss": 1.8105,
       "step": 4865
     },
     {
       "epoch": 0.6035818305756956,
-      "grad_norm": 0.2597341825241679,
+      "grad_norm": 0.14154762542667904,
       "learning_rate": 8.139618694505292e-05,
-      "loss": 3.7016,
+      "loss": 1.7879,
       "step": 4870
     },
     {
       "epoch": 0.6042015244469232,
-      "grad_norm": 0.2677923848745103,
+      "grad_norm": 0.1353547074450921,
       "learning_rate": 8.118367394510544e-05,
-      "loss": 3.543,
+      "loss": 1.8203,
       "step": 4875
     },
     {
       "epoch": 0.6048212183181508,
-      "grad_norm": 0.27300541950447416,
+      "grad_norm": 0.1268917466267913,
       "learning_rate": 8.097124900573795e-05,
-      "loss": 3.6367,
+      "loss": 1.7988,
       "step": 4880
     },
     {
       "epoch": 0.6054409121893785,
-      "grad_norm": 0.2775292613482833,
+      "grad_norm": 0.1307148187671773,
       "learning_rate": 8.075891312110104e-05,
-      "loss": 3.6172,
+      "loss": 1.8664,
       "step": 4885
     },
     {
       "epoch": 0.6060606060606061,
-      "grad_norm": 0.271824184100202,
+      "grad_norm": 0.1380952234229187,
       "learning_rate": 8.054666728492864e-05,
-      "loss": 3.7484,
+      "loss": 1.8156,
       "step": 4890
     },
     {
       "epoch": 0.6066802999318337,
-      "grad_norm": 0.26384159534759966,
+      "grad_norm": 0.13996339797842827,
       "learning_rate": 8.033451249053324e-05,
-      "loss": 3.5281,
+      "loss": 1.8059,
       "step": 4895
     },
     {
       "epoch": 0.6072999938030613,
-      "grad_norm": 0.27157171737048585,
+      "grad_norm": 0.13226203390711466,
       "learning_rate": 8.01224497308012e-05,
-      "loss": 3.6742,
+      "loss": 1.9059,
       "step": 4900
     },
     {
       "epoch": 0.6079196876742889,
-      "grad_norm": 0.25821685280582,
+      "grad_norm": 0.13671539535832183,
       "learning_rate": 7.991047999818825e-05,
-      "loss": 3.7383,
+      "loss": 1.8492,
       "step": 4905
     },
     {
       "epoch": 0.6085393815455166,
-      "grad_norm": 0.2659945632497991,
+      "grad_norm": 0.13858915396512983,
       "learning_rate": 7.969860428471472e-05,
-      "loss": 3.6859,
+      "loss": 1.8258,
       "step": 4910
     },
     {
       "epoch": 0.6091590754167441,
-      "grad_norm": 0.2701267645891113,
+      "grad_norm": 0.13457679501799066,
       "learning_rate": 7.948682358196081e-05,
-      "loss": 3.5867,
+      "loss": 1.8109,
       "step": 4915
     },
     {
       "epoch": 0.6097787692879717,
-      "grad_norm": 0.2764927840431044,
+      "grad_norm": 0.13773955152308393,
       "learning_rate": 7.927513888106222e-05,
-      "loss": 3.8344,
+      "loss": 1.868,
       "step": 4920
     },
     {
       "epoch": 0.6103984631591993,
-      "grad_norm": 0.2714043191480644,
+      "grad_norm": 0.13608898492240817,
       "learning_rate": 7.90635511727053e-05,
-      "loss": 3.6148,
+      "loss": 1.8598,
       "step": 4925
     },
     {
       "epoch": 0.611018157030427,
-      "grad_norm": 0.2609093098453354,
+      "grad_norm": 0.13594465543273473,
       "learning_rate": 7.88520614471225e-05,
-      "loss": 3.6852,
+      "loss": 1.823,
       "step": 4930
     },
     {
       "epoch": 0.6116378509016546,
-      "grad_norm": 0.2674835719736552,
+      "grad_norm": 0.13620784093910132,
       "learning_rate": 7.864067069408773e-05,
-      "loss": 3.6867,
+      "loss": 1.8395,
       "step": 4935
     },
     {
       "epoch": 0.6122575447728822,
-      "grad_norm": 0.26158771055150254,
+      "grad_norm": 0.13036549008869344,
       "learning_rate": 7.842937990291157e-05,
-      "loss": 3.6469,
+      "loss": 1.7992,
       "step": 4940
     },
     {
       "epoch": 0.6128772386441098,
-      "grad_norm": 0.26450556985138535,
+      "grad_norm": 0.14264302125929879,
       "learning_rate": 7.821819006243691e-05,
-      "loss": 3.6164,
+      "loss": 1.8539,
       "step": 4945
     },
     {
       "epoch": 0.6134969325153374,
-      "grad_norm": 0.25698316515907754,
+      "grad_norm": 0.14528351932829806,
       "learning_rate": 7.80071021610342e-05,
-      "loss": 3.7086,
+      "loss": 1.9066,
       "step": 4950
     },
     {
       "epoch": 0.6141166263865651,
-      "grad_norm": 0.27112001318405343,
+      "grad_norm": 0.13669847537271232,
       "learning_rate": 7.779611718659674e-05,
-      "loss": 3.7711,
+      "loss": 1.8215,
       "step": 4955
     },
     {
       "epoch": 0.6147363202577927,
-      "grad_norm": 0.26272113575637257,
+      "grad_norm": 0.13166972553564527,
       "learning_rate": 7.758523612653617e-05,
-      "loss": 3.6203,
+      "loss": 1.7715,
       "step": 4960
     },
     {
       "epoch": 0.6153560141290203,
-      "grad_norm": 0.267094681815031,
+      "grad_norm": 0.13422436268794732,
       "learning_rate": 7.737445996777784e-05,
-      "loss": 3.7563,
+      "loss": 1.8004,
       "step": 4965
     },
     {
       "epoch": 0.6159757080002479,
-      "grad_norm": 0.26865139865456006,
+      "grad_norm": 0.13621849103473319,
       "learning_rate": 7.716378969675599e-05,
-      "loss": 3.6984,
+      "loss": 1.834,
       "step": 4970
     },
     {
       "epoch": 0.6165954018714755,
-      "grad_norm": 0.2656146898765198,
+      "grad_norm": 0.1396825107082279,
       "learning_rate": 7.695322629940957e-05,
-      "loss": 3.5805,
+      "loss": 1.8137,
       "step": 4975
     },
     {
       "epoch": 0.6172150957427031,
-      "grad_norm": 0.27153591932306037,
+      "grad_norm": 0.14110952896496498,
       "learning_rate": 7.674277076117718e-05,
-      "loss": 3.6555,
+      "loss": 1.8672,
       "step": 4980
     },
     {
       "epoch": 0.6178347896139307,
-      "grad_norm": 0.2639320933604655,
+      "grad_norm": 0.13523516678932618,
       "learning_rate": 7.653242406699267e-05,
-      "loss": 3.6633,
+      "loss": 1.8195,
       "step": 4985
     },
     {
       "epoch": 0.6184544834851583,
-      "grad_norm": 0.2827465508666598,
+      "grad_norm": 0.13297527973482937,
       "learning_rate": 7.632218720128056e-05,
-      "loss": 3.5836,
+      "loss": 1.8781,
       "step": 4990
     },
     {
       "epoch": 0.6190741773563859,
-      "grad_norm": 0.25845031918040845,
+      "grad_norm": 0.13350350195977134,
       "learning_rate": 7.611206114795126e-05,
-      "loss": 3.6258,
+      "loss": 1.8102,
       "step": 4995
     },
     {
       "epoch": 0.6196938712276135,
-      "grad_norm": 0.27017652307124396,
+      "grad_norm": 0.13726854441704386,
       "learning_rate": 7.590204689039671e-05,
-      "loss": 3.7398,
+      "loss": 1.8469,
       "step": 5000
     },
     {
       "epoch": 0.6203135650988412,
-      "grad_norm": 0.263829705804654,
+      "grad_norm": 0.12880233502711264,
       "learning_rate": 7.569214541148551e-05,
-      "loss": 3.7102,
+      "loss": 1.8051,
       "step": 5005
     },
     {
       "epoch": 0.6209332589700688,
-      "grad_norm": 0.272388249038324,
+      "grad_norm": 0.14314377591100924,
       "learning_rate": 7.548235769355858e-05,
-      "loss": 3.7344,
+      "loss": 1.8531,
       "step": 5010
     },
     {
       "epoch": 0.6215529528412964,
-      "grad_norm": 0.2740607682775529,
+      "grad_norm": 0.1398214144314476,
       "learning_rate": 7.527268471842436e-05,
-      "loss": 3.6148,
+      "loss": 1.8633,
       "step": 5015
     },
     {
       "epoch": 0.622172646712524,
-      "grad_norm": 0.261982458328612,
+      "grad_norm": 0.1422631562696818,
       "learning_rate": 7.506312746735432e-05,
-      "loss": 3.6453,
+      "loss": 1.8121,
       "step": 5020
     },
     {
       "epoch": 0.6227923405837517,
-      "grad_norm": 0.30048336140769255,
+      "grad_norm": 0.13683576892585453,
       "learning_rate": 7.48536869210783e-05,
-      "loss": 3.6078,
+      "loss": 1.8523,
       "step": 5025
     },
     {
       "epoch": 0.6234120344549793,
-      "grad_norm": 0.2605099143553377,
+      "grad_norm": 0.13252236443713647,
       "learning_rate": 7.464436405978002e-05,
-      "loss": 3.7477,
+      "loss": 1.8219,
       "step": 5030
     },
     {
       "epoch": 0.6240317283262069,
-      "grad_norm": 0.2783089384043496,
+      "grad_norm": 0.13856901697884766,
       "learning_rate": 7.44351598630924e-05,
-      "loss": 3.6391,
+      "loss": 1.8586,
       "step": 5035
     },
     {
       "epoch": 0.6246514221974344,
-      "grad_norm": 0.269056158083971,
+      "grad_norm": 0.13467211382407307,
       "learning_rate": 7.422607531009302e-05,
-      "loss": 3.7281,
+      "loss": 1.8246,
       "step": 5040
     },
     {
       "epoch": 0.625271116068662,
-      "grad_norm": 0.2719744983748762,
+      "grad_norm": 0.13455069844136516,
       "learning_rate": 7.401711137929955e-05,
-      "loss": 3.7648,
+      "loss": 1.8809,
       "step": 5045
     },
     {
       "epoch": 0.6258908099398897,
-      "grad_norm": 0.27808306980704667,
+      "grad_norm": 0.12705255181918326,
       "learning_rate": 7.380826904866504e-05,
-      "loss": 3.6828,
+      "loss": 1.8727,
       "step": 5050
     },
     {
       "epoch": 0.6265105038111173,
-      "grad_norm": 0.2785298104106332,
+      "grad_norm": 0.14168378260835907,
       "learning_rate": 7.35995492955736e-05,
-      "loss": 3.6023,
+      "loss": 1.8375,
       "step": 5055
     },
     {
       "epoch": 0.6271301976823449,
-      "grad_norm": 0.25365798748393725,
+      "grad_norm": 0.12904228750519411,
       "learning_rate": 7.339095309683557e-05,
-      "loss": 3.7969,
+      "loss": 1.8516,
       "step": 5060
     },
     {
       "epoch": 0.6277498915535725,
-      "grad_norm": 0.258311153426152,
+      "grad_norm": 0.12951799383226725,
       "learning_rate": 7.31824814286831e-05,
-      "loss": 3.7062,
+      "loss": 1.827,
       "step": 5065
     },
     {
       "epoch": 0.6283695854248001,
-      "grad_norm": 0.2681029758178154,
+      "grad_norm": 0.1425730592753484,
       "learning_rate": 7.297413526676553e-05,
-      "loss": 3.6703,
+      "loss": 1.7684,
       "step": 5070
     },
     {
       "epoch": 0.6289892792960278,
-      "grad_norm": 0.25425870016804963,
+      "grad_norm": 0.13509911728359397,
       "learning_rate": 7.276591558614482e-05,
-      "loss": 3.7852,
+      "loss": 1.8562,
       "step": 5075
     },
     {
       "epoch": 0.6296089731672554,
-      "grad_norm": 0.28200262079953675,
+      "grad_norm": 0.12664229753196302,
       "learning_rate": 7.2557823361291e-05,
-      "loss": 3.7766,
+      "loss": 1.8227,
       "step": 5080
     },
     {
       "epoch": 0.630228667038483,
-      "grad_norm": 0.2602859160465504,
+      "grad_norm": 0.1351445427240646,
       "learning_rate": 7.23498595660776e-05,
-      "loss": 3.6867,
+      "loss": 1.8434,
       "step": 5085
     },
     {
       "epoch": 0.6308483609097106,
-      "grad_norm": 0.2705199490904154,
+      "grad_norm": 0.13550229832786473,
       "learning_rate": 7.21420251737771e-05,
-      "loss": 3.75,
+      "loss": 1.8648,
       "step": 5090
     },
     {
       "epoch": 0.6314680547809383,
-      "grad_norm": 0.25985423252580486,
+      "grad_norm": 0.13783515484834394,
       "learning_rate": 7.19343211570564e-05,
-      "loss": 3.7039,
+      "loss": 1.8687,
       "step": 5095
     },
     {
       "epoch": 0.6320877486521659,
-      "grad_norm": 0.26575649360663545,
+      "grad_norm": 0.13698632676880806,
       "learning_rate": 7.172674848797219e-05,
-      "loss": 3.7219,
+      "loss": 1.7918,
       "step": 5100
     },
     {
       "epoch": 0.6327074425233934,
-      "grad_norm": 0.269094262766888,
+      "grad_norm": 0.13469212714793175,
       "learning_rate": 7.151930813796655e-05,
-      "loss": 3.6102,
+      "loss": 1.8109,
       "step": 5105
     },
     {
       "epoch": 0.633327136394621,
-      "grad_norm": 0.26672713090869327,
+      "grad_norm": 0.13371221943648842,
       "learning_rate": 7.131200107786214e-05,
-      "loss": 3.5633,
+      "loss": 1.8078,
       "step": 5110
     },
     {
       "epoch": 0.6339468302658486,
-      "grad_norm": 0.2676506827919143,
+      "grad_norm": 0.13978406533114363,
       "learning_rate": 7.110482827785796e-05,
-      "loss": 3.6844,
+      "loss": 1.7805,
       "step": 5115
     },
     {
       "epoch": 0.6345665241370763,
-      "grad_norm": 0.2680216210875524,
+      "grad_norm": 0.14360037321686958,
       "learning_rate": 7.089779070752463e-05,
-      "loss": 3.8469,
+      "loss": 1.8004,
       "step": 5120
     },
     {
       "epoch": 0.6351862180083039,
-      "grad_norm": 0.27889275766782334,
+      "grad_norm": 0.1441559480748562,
       "learning_rate": 7.069088933579988e-05,
-      "loss": 3.5789,
+      "loss": 1.816,
       "step": 5125
     },
     {
       "epoch": 0.6358059118795315,
-      "grad_norm": 0.27602158130456417,
+      "grad_norm": 0.13537029010980783,
       "learning_rate": 7.048412513098412e-05,
-      "loss": 3.6328,
+      "loss": 1.8727,
       "step": 5130
     },
     {
       "epoch": 0.6364256057507591,
-      "grad_norm": 0.27748934261100744,
+      "grad_norm": 0.1394840279411983,
       "learning_rate": 7.027749906073564e-05,
-      "loss": 3.5906,
+      "loss": 1.7949,
       "step": 5135
     },
     {
       "epoch": 0.6370452996219867,
-      "grad_norm": 0.2573107306456965,
+      "grad_norm": 0.14014978043715506,
       "learning_rate": 7.007101209206639e-05,
-      "loss": 3.6711,
+      "loss": 1.8586,
       "step": 5140
     },
     {
       "epoch": 0.6376649934932144,
-      "grad_norm": 0.2781670907626585,
+      "grad_norm": 0.13162962388412522,
       "learning_rate": 6.98646651913373e-05,
-      "loss": 3.7609,
+      "loss": 1.8465,
       "step": 5145
     },
     {
       "epoch": 0.638284687364442,
-      "grad_norm": 0.2871360053272032,
+      "grad_norm": 0.14382551661597553,
       "learning_rate": 6.965845932425377e-05,
-      "loss": 3.6289,
+      "loss": 1.866,
       "step": 5150
     },
     {
       "epoch": 0.6389043812356696,
-      "grad_norm": 0.28246011585382963,
+      "grad_norm": 0.13458729772042977,
       "learning_rate": 6.945239545586117e-05,
-      "loss": 3.5617,
+      "loss": 1.8449,
       "step": 5155
     },
     {
       "epoch": 0.6395240751068972,
-      "grad_norm": 0.27900694670602694,
+      "grad_norm": 0.13804106195535212,
       "learning_rate": 6.924647455054036e-05,
-      "loss": 3.6117,
+      "loss": 1.7883,
       "step": 5160
     },
     {
       "epoch": 0.6401437689781249,
-      "grad_norm": 0.2723062853542401,
+      "grad_norm": 0.13515409775532647,
       "learning_rate": 6.9040697572003e-05,
-      "loss": 3.7,
+      "loss": 1.8359,
       "step": 5165
     },
     {
       "epoch": 0.6407634628493524,
-      "grad_norm": 0.2914722506156454,
+      "grad_norm": 0.13624779550221264,
       "learning_rate": 6.883506548328724e-05,
-      "loss": 3.6047,
+      "loss": 1.8066,
       "step": 5170
     },
     {
       "epoch": 0.64138315672058,
-      "grad_norm": 0.25979291503704416,
+      "grad_norm": 0.1404713378731896,
       "learning_rate": 6.862957924675323e-05,
-      "loss": 3.6648,
+      "loss": 1.8582,
       "step": 5175
     },
     {
       "epoch": 0.6420028505918076,
-      "grad_norm": 0.2759588390942882,
+      "grad_norm": 0.13353167504613667,
       "learning_rate": 6.842423982407842e-05,
-      "loss": 3.8023,
+      "loss": 1.843,
       "step": 5180
     },
     {
       "epoch": 0.6426225444630352,
-      "grad_norm": 0.26539685236538607,
+      "grad_norm": 0.13711887655413635,
       "learning_rate": 6.821904817625326e-05,
-      "loss": 3.7352,
+      "loss": 1.8637,
       "step": 5185
     },
     {
       "epoch": 0.6432422383342629,
-      "grad_norm": 0.28057212369468476,
+      "grad_norm": 0.13286338586049254,
       "learning_rate": 6.801400526357647e-05,
-      "loss": 3.6742,
+      "loss": 1.8516,
       "step": 5190
     },
     {
       "epoch": 0.6438619322054905,
-      "grad_norm": 0.25859286944728005,
+      "grad_norm": 0.1362928199879506,
       "learning_rate": 6.78091120456508e-05,
-      "loss": 3.6766,
+      "loss": 1.798,
       "step": 5195
     },
     {
       "epoch": 0.6444816260767181,
-      "grad_norm": 0.2811565662064264,
+      "grad_norm": 0.13578384732652682,
       "learning_rate": 6.760436948137843e-05,
-      "loss": 3.8312,
+      "loss": 1.8605,
       "step": 5200
     },
     {
       "epoch": 0.6451013199479457,
-      "grad_norm": 0.2848133060880273,
+      "grad_norm": 0.13894639333958217,
       "learning_rate": 6.739977852895647e-05,
-      "loss": 3.6781,
+      "loss": 1.8391,
       "step": 5205
     },
     {
       "epoch": 0.6457210138191734,
-      "grad_norm": 0.26811279612194816,
+      "grad_norm": 0.1444595514101014,
       "learning_rate": 6.719534014587243e-05,
-      "loss": 3.6383,
+      "loss": 1.8305,
       "step": 5210
     },
     {
       "epoch": 0.646340707690401,
-      "grad_norm": 0.2690681612279308,
+      "grad_norm": 0.129775059828516,
       "learning_rate": 6.699105528889988e-05,
-      "loss": 3.6758,
+      "loss": 1.8738,
       "step": 5215
     },
     {
       "epoch": 0.6469604015616286,
-      "grad_norm": 0.2739057803867215,
+      "grad_norm": 0.13988858403098833,
       "learning_rate": 6.678692491409378e-05,
-      "loss": 3.682,
+      "loss": 1.8102,
       "step": 5220
     },
     {
       "epoch": 0.6475800954328562,
-      "grad_norm": 0.2641595804456671,
+      "grad_norm": 0.1344900153408345,
       "learning_rate": 6.658294997678621e-05,
-      "loss": 3.8078,
+      "loss": 1.8879,
       "step": 5225
     },
     {
       "epoch": 0.6481997893040837,
-      "grad_norm": 0.25907535741498083,
+      "grad_norm": 0.13501587523263156,
       "learning_rate": 6.637913143158175e-05,
-      "loss": 3.6281,
+      "loss": 1.823,
       "step": 5230
     },
     {
       "epoch": 0.6488194831753114,
-      "grad_norm": 0.26304684862546324,
+      "grad_norm": 0.1322718696914636,
       "learning_rate": 6.617547023235309e-05,
-      "loss": 3.7375,
+      "loss": 1.8813,
       "step": 5235
     },
     {
       "epoch": 0.649439177046539,
-      "grad_norm": 0.2563673164937869,
+      "grad_norm": 0.13942300547170608,
       "learning_rate": 6.597196733223651e-05,
-      "loss": 3.6063,
+      "loss": 1.8574,
       "step": 5240
     },
     {
       "epoch": 0.6500588709177666,
-      "grad_norm": 0.2746685068321447,
+      "grad_norm": 0.14428546074299917,
       "learning_rate": 6.576862368362747e-05,
-      "loss": 3.6219,
+      "loss": 1.8062,
       "step": 5245
     },
     {
       "epoch": 0.6506785647889942,
-      "grad_norm": 0.26941330753939924,
+      "grad_norm": 0.13480317318245413,
       "learning_rate": 6.556544023817613e-05,
-      "loss": 3.6773,
+      "loss": 1.9059,
       "step": 5250
     },
     {
       "epoch": 0.6512982586602218,
-      "grad_norm": 0.3084803609497407,
+      "grad_norm": 0.1312291352484845,
       "learning_rate": 6.536241794678288e-05,
-      "loss": 3.6852,
+      "loss": 1.8102,
       "step": 5255
     },
     {
       "epoch": 0.6519179525314495,
-      "grad_norm": 0.2681737705088487,
+      "grad_norm": 0.13249440712367896,
       "learning_rate": 6.515955775959394e-05,
-      "loss": 3.7117,
+      "loss": 1.8719,
       "step": 5260
     },
     {
       "epoch": 0.6525376464026771,
-      "grad_norm": 0.27332131990349773,
+      "grad_norm": 0.13525668669855212,
       "learning_rate": 6.495686062599684e-05,
-      "loss": 3.6141,
+      "loss": 1.7945,
       "step": 5265
     },
     {
       "epoch": 0.6531573402739047,
-      "grad_norm": 0.26817048963551954,
+      "grad_norm": 0.13953894070499534,
       "learning_rate": 6.475432749461607e-05,
-      "loss": 3.757,
+      "loss": 1.8211,
       "step": 5270
     },
     {
       "epoch": 0.6537770341451323,
-      "grad_norm": 0.2912555620129492,
+      "grad_norm": 0.13339702671582024,
       "learning_rate": 6.455195931330855e-05,
-      "loss": 3.7703,
+      "loss": 1.8375,
       "step": 5275
     },
     {
       "epoch": 0.65439672801636,
-      "grad_norm": 0.2715643651310061,
+      "grad_norm": 0.1308872237787671,
       "learning_rate": 6.43497570291592e-05,
-      "loss": 3.6313,
+      "loss": 1.8645,
       "step": 5280
     },
     {
       "epoch": 0.6550164218875876,
-      "grad_norm": 0.25897755858739324,
+      "grad_norm": 0.13823305004742287,
       "learning_rate": 6.414772158847661e-05,
-      "loss": 3.7648,
+      "loss": 1.8766,
       "step": 5285
     },
     {
       "epoch": 0.6556361157588152,
-      "grad_norm": 0.27462922238791165,
+      "grad_norm": 0.1399613751358962,
       "learning_rate": 6.394585393678851e-05,
-      "loss": 3.5742,
+      "loss": 1.8031,
       "step": 5290
     },
     {
       "epoch": 0.6562558096300427,
-      "grad_norm": 0.27731728573308645,
+      "grad_norm": 0.13564743794929152,
       "learning_rate": 6.374415501883741e-05,
-      "loss": 3.6219,
+      "loss": 1.7996,
       "step": 5295
     },
     {
       "epoch": 0.6568755035012703,
-      "grad_norm": 0.26603376474150264,
+      "grad_norm": 0.13777383842407462,
       "learning_rate": 6.354262577857606e-05,
-      "loss": 3.7352,
+      "loss": 1.8645,
       "step": 5300
     },
     {
       "epoch": 0.657495197372498,
-      "grad_norm": 0.2786861537303143,
+      "grad_norm": 0.14386060040722395,
       "learning_rate": 6.334126715916318e-05,
-      "loss": 3.5867,
+      "loss": 1.8031,
       "step": 5305
     },
     {
       "epoch": 0.6581148912437256,
-      "grad_norm": 0.2801961113701643,
+      "grad_norm": 0.13632842622265176,
       "learning_rate": 6.314008010295897e-05,
-      "loss": 3.7102,
+      "loss": 1.7953,
       "step": 5310
     },
     {
       "epoch": 0.6587345851149532,
-      "grad_norm": 0.2787675183338051,
+      "grad_norm": 0.14031049401636042,
       "learning_rate": 6.293906555152072e-05,
-      "loss": 3.7641,
+      "loss": 1.8473,
       "step": 5315
     },
     {
       "epoch": 0.6593542789861808,
-      "grad_norm": 0.2682963267821453,
+      "grad_norm": 0.13222425685504624,
       "learning_rate": 6.273822444559839e-05,
-      "loss": 3.6117,
+      "loss": 1.8613,
       "step": 5320
     },
     {
       "epoch": 0.6599739728574084,
-      "grad_norm": 0.27492368327956523,
+      "grad_norm": 0.13291257291641337,
       "learning_rate": 6.253755772513024e-05,
-      "loss": 3.7219,
+      "loss": 1.7691,
       "step": 5325
     },
     {
       "epoch": 0.6605936667286361,
-      "grad_norm": 0.2759166364793097,
+      "grad_norm": 0.13144689432166498,
       "learning_rate": 6.233706632923832e-05,
-      "loss": 3.6437,
+      "loss": 1.8609,
       "step": 5330
     },
     {
       "epoch": 0.6612133605998637,
-      "grad_norm": 0.2653509999216259,
+      "grad_norm": 0.13941925756899518,
       "learning_rate": 6.213675119622425e-05,
-      "loss": 3.7406,
+      "loss": 1.8164,
       "step": 5335
     },
     {
       "epoch": 0.6618330544710913,
-      "grad_norm": 0.28142671319418383,
+      "grad_norm": 0.13100150642405495,
       "learning_rate": 6.19366132635647e-05,
-      "loss": 3.6461,
+      "loss": 1.8211,
       "step": 5340
     },
     {
       "epoch": 0.6624527483423189,
-      "grad_norm": 0.2831668256834555,
+      "grad_norm": 0.13942514680969523,
       "learning_rate": 6.173665346790704e-05,
-      "loss": 3.6984,
+      "loss": 1.9023,
       "step": 5345
     },
     {
       "epoch": 0.6630724422135466,
-      "grad_norm": 0.2665448241837328,
+      "grad_norm": 0.13192303146868606,
       "learning_rate": 6.153687274506501e-05,
-      "loss": 3.7305,
+      "loss": 1.8059,
       "step": 5350
     },
     {
       "epoch": 0.6636921360847742,
-      "grad_norm": 0.29779814887341804,
+      "grad_norm": 0.1388421354260116,
       "learning_rate": 6.133727203001415e-05,
-      "loss": 3.757,
+      "loss": 1.8008,
       "step": 5355
     },
     {
       "epoch": 0.6643118299560017,
-      "grad_norm": 0.28358375762398536,
+      "grad_norm": 0.1333146563999724,
       "learning_rate": 6.113785225688772e-05,
-      "loss": 3.693,
+      "loss": 1.8508,
       "step": 5360
     },
     {
       "epoch": 0.6649315238272293,
-      "grad_norm": 0.27688829546795823,
+      "grad_norm": 0.13829025809259454,
       "learning_rate": 6.093861435897208e-05,
-      "loss": 3.6906,
+      "loss": 1.8555,
       "step": 5365
     },
     {
       "epoch": 0.6655512176984569,
-      "grad_norm": 0.27159258119460017,
+      "grad_norm": 0.13128060200833697,
       "learning_rate": 6.073955926870243e-05,
-      "loss": 3.6133,
+      "loss": 1.8176,
       "step": 5370
     },
     {
       "epoch": 0.6661709115696846,
-      "grad_norm": 0.2706578605527812,
+      "grad_norm": 0.14008029597105764,
       "learning_rate": 6.0540687917658445e-05,
-      "loss": 3.6516,
+      "loss": 1.7871,
       "step": 5375
     },
     {
       "epoch": 0.6667906054409122,
-      "grad_norm": 0.272240993622507,
+      "grad_norm": 0.13309428884917443,
       "learning_rate": 6.034200123655993e-05,
-      "loss": 3.7875,
+      "loss": 1.8508,
       "step": 5380
     },
     {
       "epoch": 0.6674102993121398,
-      "grad_norm": 0.2691414978331236,
+      "grad_norm": 0.13778281479016122,
       "learning_rate": 6.01435001552623e-05,
-      "loss": 3.707,
+      "loss": 1.8551,
       "step": 5385
     },
     {
       "epoch": 0.6680299931833674,
-      "grad_norm": 0.2748226868840819,
+      "grad_norm": 0.13712329360654402,
       "learning_rate": 5.9945185602752496e-05,
-      "loss": 3.7852,
+      "loss": 1.8121,
       "step": 5390
     },
     {
       "epoch": 0.668649687054595,
-      "grad_norm": 0.2796562783963983,
+      "grad_norm": 0.14056090297114665,
       "learning_rate": 5.974705850714444e-05,
-      "loss": 3.6086,
+      "loss": 1.859,
       "step": 5395
     },
     {
       "epoch": 0.6692693809258227,
-      "grad_norm": 0.2741689382602694,
+      "grad_norm": 0.1347457309130447,
       "learning_rate": 5.95491197956748e-05,
-      "loss": 3.6648,
+      "loss": 1.8328,
       "step": 5400
     },
     {
       "epoch": 0.6698890747970503,
-      "grad_norm": 0.2647913630087514,
+      "grad_norm": 0.1298320820999991,
       "learning_rate": 5.9351370394698604e-05,
-      "loss": 3.6727,
+      "loss": 1.834,
       "step": 5405
     },
     {
       "epoch": 0.6705087686682779,
-      "grad_norm": 0.26773543962456603,
+      "grad_norm": 0.13333710741549507,
       "learning_rate": 5.9153811229684794e-05,
-      "loss": 3.7328,
+      "loss": 1.8566,
       "step": 5410
     },
     {
       "epoch": 0.6711284625395055,
-      "grad_norm": 0.27461350736207607,
+      "grad_norm": 0.14198552870822626,
       "learning_rate": 5.895644322521212e-05,
-      "loss": 3.7398,
+      "loss": 1.8516,
       "step": 5415
     },
     {
       "epoch": 0.671748156410733,
-      "grad_norm": 0.261829577537663,
+      "grad_norm": 0.13856885989348303,
       "learning_rate": 5.875926730496471e-05,
-      "loss": 3.7109,
+      "loss": 1.8672,
       "step": 5420
     },
     {
       "epoch": 0.6723678502819607,
-      "grad_norm": 0.27459196700473204,
+      "grad_norm": 0.13510293511167396,
       "learning_rate": 5.856228439172764e-05,
-      "loss": 3.7156,
+      "loss": 1.9039,
       "step": 5425
     },
     {
       "epoch": 0.6729875441531883,
-      "grad_norm": 0.2671129219642882,
+      "grad_norm": 0.13788198959803624,
       "learning_rate": 5.836549540738281e-05,
-      "loss": 3.6805,
+      "loss": 1.8254,
       "step": 5430
     },
     {
       "epoch": 0.6736072380244159,
-      "grad_norm": 0.2720255207257116,
+      "grad_norm": 0.13159584100189814,
       "learning_rate": 5.816890127290446e-05,
-      "loss": 3.7313,
+      "loss": 1.8254,
       "step": 5435
     },
     {
       "epoch": 0.6742269318956435,
-      "grad_norm": 0.2779000239190421,
+      "grad_norm": 0.14690626686001587,
       "learning_rate": 5.7972502908354954e-05,
-      "loss": 3.618,
+      "loss": 1.8363,
       "step": 5440
     },
     {
       "epoch": 0.6748466257668712,
-      "grad_norm": 0.28122200811943143,
+      "grad_norm": 0.13059037203311238,
       "learning_rate": 5.777630123288046e-05,
-      "loss": 3.7664,
+      "loss": 1.8855,
       "step": 5445
     },
     {
       "epoch": 0.6754663196380988,
-      "grad_norm": 0.2816358882326997,
+      "grad_norm": 0.13878103289346766,
       "learning_rate": 5.758029716470664e-05,
-      "loss": 3.6461,
+      "loss": 1.8887,
       "step": 5450
     },
     {
       "epoch": 0.6760860135093264,
-      "grad_norm": 0.2772970420175764,
+      "grad_norm": 0.14633351758116236,
       "learning_rate": 5.738449162113435e-05,
-      "loss": 3.6594,
+      "loss": 1.8047,
       "step": 5455
     },
     {
       "epoch": 0.676705707380554,
-      "grad_norm": 0.27756919708943467,
+      "grad_norm": 0.13885286212193357,
       "learning_rate": 5.7188885518535365e-05,
-      "loss": 3.7172,
+      "loss": 1.8578,
       "step": 5460
     },
     {
       "epoch": 0.6773254012517816,
-      "grad_norm": 0.25998511603464325,
+      "grad_norm": 0.13247881906826084,
       "learning_rate": 5.699347977234799e-05,
-      "loss": 3.5633,
+      "loss": 1.8477,
       "step": 5465
     },
     {
       "epoch": 0.6779450951230093,
-      "grad_norm": 0.2774423540805963,
+      "grad_norm": 0.1368360748545682,
       "learning_rate": 5.679827529707295e-05,
-      "loss": 3.693,
+      "loss": 1.7785,
       "step": 5470
     },
     {
       "epoch": 0.6785647889942369,
-      "grad_norm": 0.27067129052342487,
+      "grad_norm": 0.13529477063650758,
       "learning_rate": 5.660327300626902e-05,
-      "loss": 3.6727,
+      "loss": 1.8754,
       "step": 5475
     },
     {
       "epoch": 0.6791844828654645,
-      "grad_norm": 0.2777602302202561,
+      "grad_norm": 0.14429065728590754,
       "learning_rate": 5.640847381254869e-05,
-      "loss": 3.6094,
+      "loss": 1.827,
       "step": 5480
     },
     {
       "epoch": 0.679804176736692,
-      "grad_norm": 0.2767422532288637,
+      "grad_norm": 0.13699866016353732,
       "learning_rate": 5.621387862757397e-05,
-      "loss": 3.6922,
+      "loss": 1.8109,
       "step": 5485
     },
     {
       "epoch": 0.6804238706079196,
-      "grad_norm": 0.26569803803864545,
+      "grad_norm": 0.13888564340023535,
       "learning_rate": 5.6019488362052255e-05,
-      "loss": 3.6375,
+      "loss": 1.8152,
       "step": 5490
     },
     {
       "epoch": 0.6810435644791473,
-      "grad_norm": 0.2685320593943573,
+      "grad_norm": 0.13583014916724537,
       "learning_rate": 5.582530392573164e-05,
-      "loss": 3.6742,
+      "loss": 1.8211,
       "step": 5495
     },
     {
       "epoch": 0.6816632583503749,
-      "grad_norm": 0.2696298711513978,
+      "grad_norm": 0.13614465884313762,
       "learning_rate": 5.563132622739713e-05,
-      "loss": 3.693,
+      "loss": 1.841,
       "step": 5500
     },
     {
       "epoch": 0.6822829522216025,
-      "grad_norm": 0.27986715683283847,
+      "grad_norm": 0.13628815897778207,
       "learning_rate": 5.5437556174866156e-05,
-      "loss": 3.6586,
+      "loss": 1.7945,
       "step": 5505
     },
     {
       "epoch": 0.6829026460928301,
-      "grad_norm": 0.2846125032247335,
+      "grad_norm": 0.13622985157113032,
       "learning_rate": 5.5243994674984345e-05,
-      "loss": 3.6461,
+      "loss": 1.8762,
       "step": 5510
     },
     {
       "epoch": 0.6835223399640578,
-      "grad_norm": 0.2679407955321904,
+      "grad_norm": 0.13892313184585137,
       "learning_rate": 5.505064263362136e-05,
-      "loss": 3.6672,
+      "loss": 1.8637,
       "step": 5515
     },
     {
       "epoch": 0.6841420338352854,
-      "grad_norm": 0.27825325641781,
+      "grad_norm": 0.133688611710239,
       "learning_rate": 5.485750095566644e-05,
-      "loss": 3.693,
+      "loss": 1.9133,
       "step": 5520
     },
     {
       "epoch": 0.684761727706513,
-      "grad_norm": 0.26863788090470064,
+      "grad_norm": 0.14133553441876628,
       "learning_rate": 5.46645705450245e-05,
-      "loss": 3.6656,
+      "loss": 1.8555,
       "step": 5525
     },
     {
       "epoch": 0.6853814215777406,
-      "grad_norm": 0.2635749210177757,
+      "grad_norm": 0.13349705900448725,
       "learning_rate": 5.447185230461156e-05,
-      "loss": 3.807,
+      "loss": 1.7844,
       "step": 5530
     },
     {
       "epoch": 0.6860011154489682,
-      "grad_norm": 0.2712057403712623,
+      "grad_norm": 0.1328239605606481,
       "learning_rate": 5.427934713635088e-05,
-      "loss": 3.7086,
+      "loss": 1.8617,
       "step": 5535
     },
     {
       "epoch": 0.6866208093201959,
-      "grad_norm": 0.27093839270676695,
+      "grad_norm": 0.1381643142277639,
       "learning_rate": 5.4087055941168384e-05,
-      "loss": 3.7414,
+      "loss": 1.825,
       "step": 5540
     },
     {
       "epoch": 0.6872405031914235,
-      "grad_norm": 0.2713959823426367,
+      "grad_norm": 0.13528845623515193,
       "learning_rate": 5.389497961898866e-05,
-      "loss": 3.6578,
+      "loss": 1.8598,
       "step": 5545
     },
     {
       "epoch": 0.687860197062651,
-      "grad_norm": 0.28889436188917833,
+      "grad_norm": 0.12988354878682856,
       "learning_rate": 5.370311906873062e-05,
-      "loss": 3.5922,
+      "loss": 1.8645,
       "step": 5550
     },
     {
       "epoch": 0.6884798909338786,
-      "grad_norm": 0.27027978550422393,
+      "grad_norm": 0.1278301642155368,
       "learning_rate": 5.351147518830345e-05,
-      "loss": 3.7133,
+      "loss": 1.8238,
       "step": 5555
     },
     {
       "epoch": 0.6890995848051062,
-      "grad_norm": 0.26119809455788323,
+      "grad_norm": 0.1323076496864634,
       "learning_rate": 5.3320048874602266e-05,
-      "loss": 3.7734,
+      "loss": 1.8457,
       "step": 5560
     },
     {
       "epoch": 0.6897192786763339,
-      "grad_norm": 0.2535064884203829,
+      "grad_norm": 0.13812568199928557,
       "learning_rate": 5.3128841023504e-05,
-      "loss": 3.7375,
+      "loss": 1.8723,
       "step": 5565
     },
     {
       "epoch": 0.6903389725475615,
-      "grad_norm": 0.2793430975177648,
+      "grad_norm": 0.139934405228208,
       "learning_rate": 5.293785252986321e-05,
-      "loss": 3.6523,
+      "loss": 1.8379,
       "step": 5570
     },
     {
       "epoch": 0.6909586664187891,
-      "grad_norm": 0.27778118811253677,
+      "grad_norm": 0.13715770643194863,
       "learning_rate": 5.274708428750765e-05,
-      "loss": 3.5797,
+      "loss": 1.7949,
       "step": 5575
     },
     {
       "epoch": 0.6915783602900167,
-      "grad_norm": 0.270323450033622,
+      "grad_norm": 0.1298711106881199,
       "learning_rate": 5.255653718923463e-05,
-      "loss": 3.6961,
+      "loss": 1.8375,
       "step": 5580
     },
     {
       "epoch": 0.6921980541612444,
-      "grad_norm": 0.2653637163384434,
+      "grad_norm": 0.13859980183407575,
       "learning_rate": 5.236621212680628e-05,
-      "loss": 3.7023,
+      "loss": 1.8297,
       "step": 5585
     },
     {
       "epoch": 0.692817748032472,
-      "grad_norm": 0.2753589669008408,
+      "grad_norm": 0.1318933840751103,
       "learning_rate": 5.217610999094563e-05,
-      "loss": 3.7102,
+      "loss": 1.841,
       "step": 5590
     },
     {
       "epoch": 0.6934374419036996,
-      "grad_norm": 0.29022430253276155,
+      "grad_norm": 0.14800086449754893,
       "learning_rate": 5.1986231671332454e-05,
-      "loss": 3.6742,
+      "loss": 1.843,
       "step": 5595
     },
     {
       "epoch": 0.6940571357749272,
-      "grad_norm": 0.27688395623809187,
+      "grad_norm": 0.1323008676469033,
       "learning_rate": 5.179657805659908e-05,
-      "loss": 3.5992,
+      "loss": 1.8613,
       "step": 5600
     },
     {
       "epoch": 0.6946768296461548,
-      "grad_norm": 0.26855552012781275,
+      "grad_norm": 0.13870752821590712,
       "learning_rate": 5.160715003432608e-05,
-      "loss": 3.718,
+      "loss": 1.8664,
       "step": 5605
     },
     {
       "epoch": 0.6952965235173824,
-      "grad_norm": 0.2605313018963682,
+      "grad_norm": 0.13883802765782519,
       "learning_rate": 5.1417948491038416e-05,
-      "loss": 3.7578,
+      "loss": 1.8098,
       "step": 5610
     },
     {
       "epoch": 0.69591621738861,
-      "grad_norm": 0.2664551240286889,
+      "grad_norm": 0.14505650835163597,
       "learning_rate": 5.122897431220104e-05,
-      "loss": 3.7234,
+      "loss": 1.848,
       "step": 5615
     },
     {
       "epoch": 0.6965359112598376,
-      "grad_norm": 0.27718395586925093,
+      "grad_norm": 0.142610612049516,
       "learning_rate": 5.104022838221487e-05,
-      "loss": 3.6727,
+      "loss": 1.7668,
       "step": 5620
     },
     {
       "epoch": 0.6971556051310652,
-      "grad_norm": 0.2779483743020303,
+      "grad_norm": 0.13663598600996665,
       "learning_rate": 5.085171158441261e-05,
-      "loss": 3.6977,
+      "loss": 1.8543,
       "step": 5625
     },
     {
       "epoch": 0.6977752990022928,
-      "grad_norm": 0.2851749067821478,
+      "grad_norm": 0.14135398916556138,
       "learning_rate": 5.0663424801054595e-05,
-      "loss": 3.5562,
+      "loss": 1.8148,
       "step": 5630
     },
     {
       "epoch": 0.6983949928735205,
-      "grad_norm": 0.2735140799612901,
+      "grad_norm": 0.1382568356418709,
       "learning_rate": 5.047536891332473e-05,
-      "loss": 3.7687,
+      "loss": 1.8004,
       "step": 5635
     },
     {
       "epoch": 0.6990146867447481,
-      "grad_norm": 0.27891630984844623,
+      "grad_norm": 0.13404657874607154,
       "learning_rate": 5.0287544801326293e-05,
-      "loss": 3.6484,
+      "loss": 1.8324,
       "step": 5640
     },
     {
       "epoch": 0.6996343806159757,
-      "grad_norm": 0.2798336852376535,
+      "grad_norm": 0.13905564445774873,
       "learning_rate": 5.0099953344077885e-05,
-      "loss": 3.6914,
+      "loss": 1.8457,
       "step": 5645
     },
     {
       "epoch": 0.7002540744872033,
-      "grad_norm": 0.28114121108671125,
+      "grad_norm": 0.1369113594997885,
       "learning_rate": 4.991259541950924e-05,
-      "loss": 3.6063,
+      "loss": 1.8586,
       "step": 5650
     },
     {
       "epoch": 0.700873768358431,
-      "grad_norm": 0.27673349011973397,
+      "grad_norm": 0.13278800041844127,
       "learning_rate": 4.972547190445723e-05,
-      "loss": 3.7367,
+      "loss": 1.8719,
       "step": 5655
     },
     {
       "epoch": 0.7014934622296586,
-      "grad_norm": 0.2662079818812042,
+      "grad_norm": 0.14229177852733657,
       "learning_rate": 4.953858367466155e-05,
-      "loss": 3.6422,
+      "loss": 1.8949,
       "step": 5660
     },
     {
       "epoch": 0.7021131561008862,
-      "grad_norm": 0.27756807998773575,
+      "grad_norm": 0.14078175269659313,
       "learning_rate": 4.9351931604760907e-05,
-      "loss": 3.7484,
+      "loss": 1.8633,
       "step": 5665
     },
     {
       "epoch": 0.7027328499721138,
-      "grad_norm": 0.27804509429596413,
+      "grad_norm": 0.1317958863467384,
       "learning_rate": 4.9165516568288674e-05,
-      "loss": 3.6445,
+      "loss": 1.8641,
       "step": 5670
     },
     {
       "epoch": 0.7033525438433413,
-      "grad_norm": 0.275651335130277,
+      "grad_norm": 0.13544413510947215,
       "learning_rate": 4.897933943766897e-05,
-      "loss": 3.6531,
+      "loss": 1.816,
       "step": 5675
     },
     {
       "epoch": 0.703972237714569,
-      "grad_norm": 0.27122700127379756,
+      "grad_norm": 0.1298237481765858,
       "learning_rate": 4.879340108421248e-05,
-      "loss": 3.7055,
+      "loss": 1.8805,
       "step": 5680
     },
     {
       "epoch": 0.7045919315857966,
-      "grad_norm": 0.26980941705603506,
+      "grad_norm": 0.13472711627814213,
       "learning_rate": 4.8607702378112415e-05,
-      "loss": 3.7414,
+      "loss": 1.841,
       "step": 5685
     },
     {
       "epoch": 0.7052116254570242,
-      "grad_norm": 0.26003012135298476,
+      "grad_norm": 0.13535347127585984,
       "learning_rate": 4.842224418844045e-05,
-      "loss": 3.5023,
+      "loss": 1.8094,
       "step": 5690
     },
     {
       "epoch": 0.7058313193282518,
-      "grad_norm": 0.2647503412298433,
+      "grad_norm": 0.13436886422476685,
       "learning_rate": 4.823702738314262e-05,
-      "loss": 3.7102,
+      "loss": 1.823,
       "step": 5695
     },
     {
       "epoch": 0.7064510131994794,
-      "grad_norm": 0.2712871372151752,
+      "grad_norm": 0.14068253039093517,
       "learning_rate": 4.8052052829035275e-05,
-      "loss": 3.8047,
+      "loss": 1.8824,
       "step": 5700
     },
     {
       "epoch": 0.7070707070707071,
-      "grad_norm": 0.27859849030011535,
+      "grad_norm": 0.1346479936388671,
       "learning_rate": 4.7867321391801065e-05,
-      "loss": 3.682,
+      "loss": 1.8781,
       "step": 5705
     },
     {
       "epoch": 0.7076904009419347,
-      "grad_norm": 0.2820394324659631,
+      "grad_norm": 0.13435815599940173,
       "learning_rate": 4.768283393598484e-05,
-      "loss": 3.5344,
+      "loss": 1.8207,
       "step": 5710
     },
     {
       "epoch": 0.7083100948131623,
-      "grad_norm": 0.2591429454872442,
+      "grad_norm": 0.14223328089430068,
       "learning_rate": 4.749859132498953e-05,
-      "loss": 3.6195,
+      "loss": 1.8543,
       "step": 5715
     },
     {
       "epoch": 0.7089297886843899,
-      "grad_norm": 0.3010598054872879,
+      "grad_norm": 0.13941692924625854,
       "learning_rate": 4.73145944210723e-05,
-      "loss": 3.6203,
+      "loss": 1.8285,
       "step": 5720
     },
     {
       "epoch": 0.7095494825556176,
-      "grad_norm": 0.2722081655502301,
+      "grad_norm": 0.13441710577473923,
       "learning_rate": 4.713084408534035e-05,
-      "loss": 3.6688,
+      "loss": 1.7984,
       "step": 5725
     },
     {
       "epoch": 0.7101691764268452,
-      "grad_norm": 0.27122867523097466,
+      "grad_norm": 0.14059083732086994,
       "learning_rate": 4.6947341177746926e-05,
-      "loss": 3.7344,
+      "loss": 1.7797,
       "step": 5730
     },
     {
       "epoch": 0.7107888702980728,
-      "grad_norm": 0.27824464760436307,
+      "grad_norm": 0.13080419041067848,
       "learning_rate": 4.6764086557087406e-05,
-      "loss": 3.7719,
+      "loss": 1.8066,
       "step": 5735
     },
     {
       "epoch": 0.7114085641693003,
-      "grad_norm": 0.27414327511872144,
+      "grad_norm": 0.14130261841489294,
       "learning_rate": 4.65810810809951e-05,
-      "loss": 3.6859,
+      "loss": 1.8453,
       "step": 5740
     },
     {
       "epoch": 0.7120282580405279,
-      "grad_norm": 0.27915537657137607,
+      "grad_norm": 0.14387364542120532,
       "learning_rate": 4.6398325605937265e-05,
-      "loss": 3.6406,
+      "loss": 1.8383,
       "step": 5745
     },
     {
       "epoch": 0.7126479519117556,
-      "grad_norm": 0.2781760005338474,
+      "grad_norm": 0.138013788809857,
       "learning_rate": 4.621582098721124e-05,
-      "loss": 3.7109,
+      "loss": 1.8703,
       "step": 5750
     },
     {
       "epoch": 0.7132676457829832,
-      "grad_norm": 0.279762376148418,
+      "grad_norm": 0.13149717266725255,
       "learning_rate": 4.6033568078940345e-05,
-      "loss": 3.6594,
+      "loss": 1.8422,
       "step": 5755
     },
     {
       "epoch": 0.7138873396542108,
-      "grad_norm": 0.2635731530602869,
+      "grad_norm": 0.14300678827280267,
       "learning_rate": 4.585156773406986e-05,
-      "loss": 3.5781,
+      "loss": 1.8402,
       "step": 5760
     },
     {
       "epoch": 0.7145070335254384,
-      "grad_norm": 0.2807646872525827,
+      "grad_norm": 0.1369909491023122,
       "learning_rate": 4.5669820804363116e-05,
-      "loss": 3.7719,
+      "loss": 1.827,
       "step": 5765
     },
     {
       "epoch": 0.715126727396666,
-      "grad_norm": 0.28003214694151435,
+      "grad_norm": 0.1381026280394612,
       "learning_rate": 4.5488328140397364e-05,
-      "loss": 3.6719,
+      "loss": 1.8656,
       "step": 5770
     },
     {
       "epoch": 0.7157464212678937,
-      "grad_norm": 0.2679072939914116,
+      "grad_norm": 0.13344045108890723,
       "learning_rate": 4.530709059155994e-05,
-      "loss": 3.7164,
+      "loss": 1.8281,
       "step": 5775
     },
     {
       "epoch": 0.7163661151391213,
-      "grad_norm": 0.26558897424509803,
+      "grad_norm": 0.14375868703270098,
       "learning_rate": 4.512610900604434e-05,
-      "loss": 3.6875,
+      "loss": 1.8422,
       "step": 5780
     },
     {
       "epoch": 0.7169858090103489,
-      "grad_norm": 0.2738273465942126,
+      "grad_norm": 0.1373988139845734,
       "learning_rate": 4.4945384230846e-05,
-      "loss": 3.7906,
+      "loss": 1.8176,
       "step": 5785
     },
     {
       "epoch": 0.7176055028815765,
-      "grad_norm": 0.27969995555479055,
+      "grad_norm": 0.13871113648194647,
       "learning_rate": 4.476491711175854e-05,
-      "loss": 3.6391,
+      "loss": 1.8008,
       "step": 5790
     },
     {
       "epoch": 0.7182251967528042,
-      "grad_norm": 0.25880179682133825,
+      "grad_norm": 0.1358378570022186,
       "learning_rate": 4.45847084933698e-05,
-      "loss": 3.7211,
+      "loss": 1.8734,
       "step": 5795
     },
     {
       "epoch": 0.7188448906240317,
-      "grad_norm": 0.25811067963296236,
+      "grad_norm": 0.1331598981484069,
       "learning_rate": 4.440475921905768e-05,
-      "loss": 3.6898,
+      "loss": 1.7898,
       "step": 5800
     },
     {
       "epoch": 0.7194645844952593,
-      "grad_norm": 0.27238020603219465,
+      "grad_norm": 0.1332569979541051,
       "learning_rate": 4.422507013098651e-05,
-      "loss": 3.5922,
+      "loss": 1.8148,
       "step": 5805
     },
     {
       "epoch": 0.7200842783664869,
-      "grad_norm": 0.2763272460657019,
+      "grad_norm": 0.1401241076701764,
       "learning_rate": 4.404564207010288e-05,
-      "loss": 3.6836,
+      "loss": 1.8598,
       "step": 5810
     },
     {
       "epoch": 0.7207039722377145,
-      "grad_norm": 0.27148027454186435,
+      "grad_norm": 0.13606072750770695,
       "learning_rate": 4.3866475876131764e-05,
-      "loss": 3.7961,
+      "loss": 1.7977,
       "step": 5815
     },
     {
       "epoch": 0.7213236661089422,
-      "grad_norm": 0.25605901536244036,
+      "grad_norm": 0.12799996771306443,
       "learning_rate": 4.3687572387572605e-05,
-      "loss": 3.7375,
+      "loss": 1.8617,
       "step": 5820
     },
     {
       "epoch": 0.7219433599801698,
-      "grad_norm": 0.26265291674637486,
+      "grad_norm": 0.1392372680256112,
       "learning_rate": 4.350893244169541e-05,
-      "loss": 3.7172,
+      "loss": 1.8543,
       "step": 5825
     },
     {
       "epoch": 0.7225630538513974,
-      "grad_norm": 0.28286131134250764,
+      "grad_norm": 0.13632894278294244,
       "learning_rate": 4.333055687453673e-05,
-      "loss": 3.557,
+      "loss": 1.8211,
       "step": 5830
     },
     {
       "epoch": 0.723182747722625,
-      "grad_norm": 0.27841967359618486,
+      "grad_norm": 0.14283083422916973,
       "learning_rate": 4.315244652089592e-05,
-      "loss": 3.6977,
+      "loss": 1.7918,
       "step": 5835
     },
     {
       "epoch": 0.7238024415938527,
-      "grad_norm": 0.2608987536285254,
+      "grad_norm": 0.14036541324610963,
       "learning_rate": 4.297460221433104e-05,
-      "loss": 3.6414,
+      "loss": 1.8301,
       "step": 5840
     },
     {
       "epoch": 0.7244221354650803,
-      "grad_norm": 0.2825656150573623,
+      "grad_norm": 0.13411932671691454,
       "learning_rate": 4.2797024787155114e-05,
-      "loss": 3.6492,
+      "loss": 1.8434,
       "step": 5845
     },
     {
       "epoch": 0.7250418293363079,
-      "grad_norm": 0.2876457246577729,
+      "grad_norm": 0.13644407967970396,
       "learning_rate": 4.2619715070432174e-05,
-      "loss": 3.7148,
+      "loss": 1.8492,
       "step": 5850
     },
     {
       "epoch": 0.7256615232075355,
-      "grad_norm": 0.27880007263838713,
+      "grad_norm": 0.13314774631995285,
       "learning_rate": 4.244267389397326e-05,
-      "loss": 3.6133,
+      "loss": 1.825,
       "step": 5855
     },
     {
       "epoch": 0.7262812170787631,
-      "grad_norm": 0.2720496922839693,
+      "grad_norm": 0.1300981208309135,
       "learning_rate": 4.226590208633275e-05,
-      "loss": 3.7305,
+      "loss": 1.8262,
       "step": 5860
     },
     {
       "epoch": 0.7269009109499907,
-      "grad_norm": 0.272676520442399,
+      "grad_norm": 0.14053955987130684,
       "learning_rate": 4.208940047480434e-05,
-      "loss": 3.718,
+      "loss": 1.8016,
       "step": 5865
     },
     {
       "epoch": 0.7275206048212183,
-      "grad_norm": 0.26497822927460635,
+      "grad_norm": 0.13840346422495164,
       "learning_rate": 4.191316988541721e-05,
-      "loss": 3.632,
+      "loss": 1.8449,
       "step": 5870
     },
     {
       "epoch": 0.7281402986924459,
-      "grad_norm": 0.27723156490255046,
+      "grad_norm": 0.14187233849517403,
       "learning_rate": 4.173721114293214e-05,
-      "loss": 3.8031,
+      "loss": 1.8168,
       "step": 5875
     },
     {
       "epoch": 0.7287599925636735,
-      "grad_norm": 0.26094788474282904,
+      "grad_norm": 0.13386643476403284,
       "learning_rate": 4.156152507083767e-05,
-      "loss": 3.6758,
+      "loss": 1.8105,
       "step": 5880
     },
     {
       "epoch": 0.7293796864349011,
-      "grad_norm": 0.28172055855879213,
+      "grad_norm": 0.1345475089808684,
       "learning_rate": 4.1386112491346255e-05,
-      "loss": 3.743,
+      "loss": 1.8125,
       "step": 5885
     },
     {
       "epoch": 0.7299993803061288,
-      "grad_norm": 0.2732741509695242,
+      "grad_norm": 0.13643936735112303,
       "learning_rate": 4.121097422539036e-05,
-      "loss": 3.6992,
+      "loss": 1.8852,
       "step": 5890
     },
     {
       "epoch": 0.7306190741773564,
-      "grad_norm": 0.28094804569295945,
+      "grad_norm": 0.12903168333708062,
       "learning_rate": 4.1036111092618725e-05,
-      "loss": 3.5977,
+      "loss": 1.8113,
       "step": 5895
     },
     {
       "epoch": 0.731238768048584,
-      "grad_norm": 0.2677662106993077,
+      "grad_norm": 0.14243171873479077,
       "learning_rate": 4.0861523911392406e-05,
-      "loss": 3.6539,
+      "loss": 1.7836,
       "step": 5900
     },
     {
       "epoch": 0.7318584619198116,
-      "grad_norm": 0.2697207576664988,
+      "grad_norm": 0.14039456232514363,
       "learning_rate": 4.068721349878107e-05,
-      "loss": 3.5273,
+      "loss": 1.8797,
       "step": 5905
     },
     {
       "epoch": 0.7324781557910393,
-      "grad_norm": 0.2707629838608235,
+      "grad_norm": 0.13211859990032615,
       "learning_rate": 4.051318067055898e-05,
-      "loss": 3.632,
+      "loss": 1.8375,
       "step": 5910
     },
     {
       "epoch": 0.7330978496622669,
-      "grad_norm": 0.2674857148566874,
+      "grad_norm": 0.14206687574027346,
       "learning_rate": 4.033942624120143e-05,
-      "loss": 3.7164,
+      "loss": 1.8238,
       "step": 5915
     },
     {
       "epoch": 0.7337175435334945,
-      "grad_norm": 0.2655492853840789,
+      "grad_norm": 0.13499243739697084,
       "learning_rate": 4.0165951023880746e-05,
-      "loss": 3.782,
+      "loss": 1.8441,
       "step": 5920
     },
     {
       "epoch": 0.7343372374047221,
-      "grad_norm": 0.27416095873987867,
+      "grad_norm": 0.13559241628088486,
       "learning_rate": 3.999275583046256e-05,
-      "loss": 3.7094,
+      "loss": 1.7824,
       "step": 5925
     },
     {
       "epoch": 0.7349569312759496,
-      "grad_norm": 0.267483841770565,
+      "grad_norm": 0.14192010466529414,
       "learning_rate": 3.981984147150196e-05,
-      "loss": 3.7852,
+      "loss": 1.8527,
       "step": 5930
     },
     {
       "epoch": 0.7355766251471773,
-      "grad_norm": 0.2745830184563242,
+      "grad_norm": 0.1318846856274811,
       "learning_rate": 3.964720875623976e-05,
-      "loss": 3.6359,
+      "loss": 1.8039,
       "step": 5935
     },
     {
       "epoch": 0.7361963190184049,
-      "grad_norm": 0.2757045668181282,
+      "grad_norm": 0.13396322343104464,
       "learning_rate": 3.9474858492598653e-05,
-      "loss": 3.7078,
+      "loss": 1.8039,
       "step": 5940
     },
     {
       "epoch": 0.7368160128896325,
-      "grad_norm": 0.27129935273188194,
+      "grad_norm": 0.13529179713111514,
       "learning_rate": 3.930279148717948e-05,
-      "loss": 3.7453,
+      "loss": 1.7961,
       "step": 5945
     },
     {
       "epoch": 0.7374357067608601,
-      "grad_norm": 0.2652954671649276,
+      "grad_norm": 0.13808062326374956,
       "learning_rate": 3.913100854525742e-05,
-      "loss": 3.6898,
+      "loss": 1.857,
       "step": 5950
     },
     {
       "epoch": 0.7380554006320877,
-      "grad_norm": 0.2660265511643317,
+      "grad_norm": 0.14259484377808126,
       "learning_rate": 3.895951047077821e-05,
-      "loss": 3.5609,
+      "loss": 1.8074,
       "step": 5955
     },
     {
       "epoch": 0.7386750945033154,
-      "grad_norm": 0.2850360173732016,
+      "grad_norm": 0.14048933061505342,
       "learning_rate": 3.8788298066354464e-05,
-      "loss": 3.6125,
+      "loss": 1.8113,
       "step": 5960
     },
     {
       "epoch": 0.739294788374543,
-      "grad_norm": 0.2747779595842181,
+      "grad_norm": 0.1365199990868293,
       "learning_rate": 3.8617372133261766e-05,
-      "loss": 3.7523,
+      "loss": 1.8387,
       "step": 5965
     },
     {
       "epoch": 0.7399144822457706,
-      "grad_norm": 0.27791872773574944,
+      "grad_norm": 0.1304295788973774,
       "learning_rate": 3.844673347143507e-05,
-      "loss": 3.6641,
+      "loss": 1.8598,
       "step": 5970
     },
     {
       "epoch": 0.7405341761169982,
-      "grad_norm": 0.26936680667842383,
+      "grad_norm": 0.14192605565965738,
       "learning_rate": 3.827638287946489e-05,
-      "loss": 3.5922,
+      "loss": 1.8555,
       "step": 5975
     },
     {
       "epoch": 0.7411538699882259,
-      "grad_norm": 0.2676597010665189,
+      "grad_norm": 0.13697173512407929,
       "learning_rate": 3.8106321154593605e-05,
-      "loss": 3.6336,
+      "loss": 1.8055,
       "step": 5980
     },
     {
       "epoch": 0.7417735638594535,
-      "grad_norm": 0.27102690603506263,
+      "grad_norm": 0.13183561867638896,
       "learning_rate": 3.793654909271169e-05,
-      "loss": 3.6437,
+      "loss": 1.7809,
       "step": 5985
     },
     {
       "epoch": 0.742393257730681,
-      "grad_norm": 0.267102384870593,
+      "grad_norm": 0.14024605247591226,
       "learning_rate": 3.776706748835388e-05,
-      "loss": 3.6969,
+      "loss": 1.8723,
       "step": 5990
     },
     {
       "epoch": 0.7430129516019086,
-      "grad_norm": 0.2764451949708146,
+      "grad_norm": 0.14086892950568974,
       "learning_rate": 3.759787713469569e-05,
-      "loss": 3.7094,
+      "loss": 1.8406,
       "step": 5995
     },
     {
       "epoch": 0.7436326454731362,
-      "grad_norm": 0.27181670294451116,
+      "grad_norm": 0.14334020779281637,
       "learning_rate": 3.7428978823549545e-05,
-      "loss": 3.7445,
+      "loss": 1.8184,
       "step": 6000
     },
     {
       "epoch": 0.7442523393443639,
-      "grad_norm": 0.2700157262463554,
+      "grad_norm": 0.13571907865119778,
       "learning_rate": 3.726037334536109e-05,
-      "loss": 3.6367,
+      "loss": 1.8562,
       "step": 6005
     },
     {
       "epoch": 0.7448720332155915,
-      "grad_norm": 0.2649639847023256,
+      "grad_norm": 0.14184536059545447,
       "learning_rate": 3.709206148920553e-05,
-      "loss": 3.6531,
+      "loss": 1.9035,
       "step": 6010
     },
     {
       "epoch": 0.7454917270868191,
-      "grad_norm": 0.2661623158171092,
+      "grad_norm": 0.13624732293876154,
       "learning_rate": 3.692404404278395e-05,
-      "loss": 3.7008,
+      "loss": 1.8234,
       "step": 6015
     },
     {
       "epoch": 0.7461114209580467,
-      "grad_norm": 0.2626176135511345,
+      "grad_norm": 0.13276441124031632,
       "learning_rate": 3.675632179241946e-05,
-      "loss": 3.6617,
+      "loss": 1.8305,
       "step": 6020
     },
     {
       "epoch": 0.7467311148292743,
-      "grad_norm": 0.2733085135357621,
+      "grad_norm": 0.13930808879264342,
       "learning_rate": 3.658889552305376e-05,
-      "loss": 3.6055,
+      "loss": 1.7789,
       "step": 6025
     },
     {
       "epoch": 0.747350808700502,
-      "grad_norm": 0.2699184735568989,
+      "grad_norm": 0.13764635628250024,
       "learning_rate": 3.642176601824341e-05,
-      "loss": 3.5484,
+      "loss": 1.8172,
       "step": 6030
     },
     {
       "epoch": 0.7479705025717296,
-      "grad_norm": 0.2753762018171419,
+      "grad_norm": 0.13461489368112944,
       "learning_rate": 3.6254934060156e-05,
-      "loss": 3.7484,
+      "loss": 1.8633,
       "step": 6035
     },
     {
       "epoch": 0.7485901964429572,
-      "grad_norm": 0.26788292526469076,
+      "grad_norm": 0.13966985047575237,
       "learning_rate": 3.608840042956666e-05,
-      "loss": 3.6727,
+      "loss": 1.8715,
       "step": 6040
     },
     {
       "epoch": 0.7492098903141848,
-      "grad_norm": 0.2797673176500415,
+      "grad_norm": 0.13990874030023231,
       "learning_rate": 3.592216590585427e-05,
-      "loss": 3.6391,
+      "loss": 1.8723,
       "step": 6045
     },
     {
       "epoch": 0.7498295841854125,
-      "grad_norm": 0.26026046751945825,
+      "grad_norm": 0.14361890024985108,
       "learning_rate": 3.5756231266997965e-05,
-      "loss": 3.5984,
+      "loss": 1.8285,
       "step": 6050
     },
     {
       "epoch": 0.75044927805664,
-      "grad_norm": 0.26334684058291796,
+      "grad_norm": 0.14084856041667215,
       "learning_rate": 3.559059728957338e-05,
-      "loss": 3.6688,
+      "loss": 1.8383,
       "step": 6055
     },
     {
       "epoch": 0.7510689719278676,
-      "grad_norm": 0.2816745822735808,
+      "grad_norm": 0.13120761896400793,
       "learning_rate": 3.5425264748749074e-05,
-      "loss": 3.7313,
+      "loss": 1.8734,
       "step": 6060
     },
     {
       "epoch": 0.7516886657990952,
-      "grad_norm": 0.28352625683115584,
+      "grad_norm": 0.13658964579723912,
       "learning_rate": 3.5260234418282865e-05,
-      "loss": 3.6672,
+      "loss": 1.8109,
       "step": 6065
     },
     {
       "epoch": 0.7523083596703228,
-      "grad_norm": 0.27308255212562715,
+      "grad_norm": 0.1344443341684798,
       "learning_rate": 3.509550707051823e-05,
-      "loss": 3.5969,
+      "loss": 1.8398,
       "step": 6070
     },
     {
       "epoch": 0.7529280535415505,
-      "grad_norm": 0.28045714684906514,
+      "grad_norm": 0.14012917508847528,
       "learning_rate": 3.493108347638067e-05,
-      "loss": 3.7687,
+      "loss": 1.8484,
       "step": 6075
     },
     {
       "epoch": 0.7535477474127781,
-      "grad_norm": 0.28038747649942236,
+      "grad_norm": 0.13379672444004473,
       "learning_rate": 3.476696440537413e-05,
-      "loss": 3.6875,
+      "loss": 1.7824,
       "step": 6080
     },
     {
       "epoch": 0.7541674412840057,
-      "grad_norm": 0.2757804471211753,
+      "grad_norm": 0.1434039130004496,
       "learning_rate": 3.460315062557737e-05,
-      "loss": 3.6906,
+      "loss": 1.7945,
       "step": 6085
     },
     {
       "epoch": 0.7547871351552333,
-      "grad_norm": 0.27039100222498835,
+      "grad_norm": 0.13398319194874977,
       "learning_rate": 3.443964290364041e-05,
-      "loss": 3.6086,
+      "loss": 1.8031,
       "step": 6090
     },
     {
       "epoch": 0.755406829026461,
-      "grad_norm": 0.27574840398743766,
+      "grad_norm": 0.13976767423578978,
       "learning_rate": 3.4276442004780916e-05,
-      "loss": 3.6977,
+      "loss": 1.9234,
       "step": 6095
     },
     {
       "epoch": 0.7560265228976886,
-      "grad_norm": 0.26908095664325493,
+      "grad_norm": 0.14215579154672933,
       "learning_rate": 3.411354869278056e-05,
-      "loss": 3.7055,
+      "loss": 1.8199,
       "step": 6100
     },
     {
       "epoch": 0.7566462167689162,
-      "grad_norm": 0.2871430737464743,
+      "grad_norm": 0.14305847528188373,
       "learning_rate": 3.3950963729981565e-05,
-      "loss": 3.7453,
+      "loss": 1.8391,
       "step": 6105
     },
     {
       "epoch": 0.7572659106401438,
-      "grad_norm": 0.27111861031633594,
+      "grad_norm": 0.13159704286365811,
       "learning_rate": 3.378868787728308e-05,
-      "loss": 3.5867,
+      "loss": 1.8926,
       "step": 6110
     },
     {
       "epoch": 0.7578856045113714,
-      "grad_norm": 0.276205387609455,
+      "grad_norm": 0.13826500141162315,
       "learning_rate": 3.362672189413756e-05,
-      "loss": 3.7328,
+      "loss": 1.832,
       "step": 6115
     },
     {
       "epoch": 0.758505298382599,
-      "grad_norm": 0.26012347649205353,
+      "grad_norm": 0.13932153252369295,
       "learning_rate": 3.346506653854734e-05,
-      "loss": 3.5984,
+      "loss": 1.8195,
       "step": 6120
     },
     {
       "epoch": 0.7591249922538266,
-      "grad_norm": 0.29007047304566813,
+      "grad_norm": 0.1408428855634542,
       "learning_rate": 3.3303722567060956e-05,
-      "loss": 3.6797,
+      "loss": 1.7531,
       "step": 6125
     },
     {
       "epoch": 0.7597446861250542,
-      "grad_norm": 0.2690965755326681,
+      "grad_norm": 0.1310748687649542,
       "learning_rate": 3.31426907347697e-05,
-      "loss": 3.7141,
+      "loss": 1.8559,
       "step": 6130
     },
     {
       "epoch": 0.7603643799962818,
-      "grad_norm": 0.2677627631712178,
+      "grad_norm": 0.13962627369469138,
       "learning_rate": 3.2981971795304026e-05,
-      "loss": 3.6375,
+      "loss": 1.8816,
       "step": 6135
     },
     {
       "epoch": 0.7609840738675094,
-      "grad_norm": 0.2661547296225876,
+      "grad_norm": 0.1341856783680393,
       "learning_rate": 3.282156650083006e-05,
-      "loss": 3.6875,
+      "loss": 1.8637,
       "step": 6140
     },
     {
       "epoch": 0.7616037677387371,
-      "grad_norm": 0.27694800622542926,
+      "grad_norm": 0.13593015684887874,
       "learning_rate": 3.266147560204608e-05,
-      "loss": 3.6703,
+      "loss": 1.8684,
       "step": 6145
     },
     {
       "epoch": 0.7622234616099647,
-      "grad_norm": 0.2744597463214922,
+      "grad_norm": 0.1416671199439836,
       "learning_rate": 3.250169984817897e-05,
-      "loss": 3.7047,
+      "loss": 1.8223,
       "step": 6150
     },
     {
       "epoch": 0.7628431554811923,
-      "grad_norm": 0.2760498499832667,
+      "grad_norm": 0.13085965612004039,
       "learning_rate": 3.23422399869807e-05,
-      "loss": 3.7289,
+      "loss": 1.8141,
       "step": 6155
     },
     {
       "epoch": 0.7634628493524199,
-      "grad_norm": 0.3144507061509538,
+      "grad_norm": 0.13830254030035186,
       "learning_rate": 3.2183096764724915e-05,
-      "loss": 3.6688,
+      "loss": 1.8324,
       "step": 6160
     },
     {
       "epoch": 0.7640825432236475,
-      "grad_norm": 0.2719439357170228,
+      "grad_norm": 0.13986388265543723,
       "learning_rate": 3.2024270926203384e-05,
-      "loss": 3.6875,
+      "loss": 1.7957,
       "step": 6165
     },
     {
       "epoch": 0.7647022370948752,
-      "grad_norm": 0.2782653971635696,
+      "grad_norm": 0.14393270333700767,
       "learning_rate": 3.1865763214722474e-05,
-      "loss": 3.8164,
+      "loss": 1.848,
       "step": 6170
     },
     {
       "epoch": 0.7653219309661028,
-      "grad_norm": 0.26283063441760457,
+      "grad_norm": 0.12961867709571917,
       "learning_rate": 3.1707574372099754e-05,
-      "loss": 3.7227,
+      "loss": 1.8449,
       "step": 6175
     },
     {
       "epoch": 0.7659416248373303,
-      "grad_norm": 0.2640248567397006,
+      "grad_norm": 0.13331591764693343,
       "learning_rate": 3.154970513866047e-05,
-      "loss": 3.7109,
+      "loss": 1.8234,
       "step": 6180
     },
     {
       "epoch": 0.7665613187085579,
-      "grad_norm": 0.2821077631231624,
+      "grad_norm": 0.13854732063824962,
       "learning_rate": 3.1392156253234086e-05,
-      "loss": 3.7023,
+      "loss": 1.8215,
       "step": 6185
     },
     {
       "epoch": 0.7671810125797855,
-      "grad_norm": 0.2647061292808031,
+      "grad_norm": 0.13593317269346725,
       "learning_rate": 3.123492845315086e-05,
-      "loss": 3.6844,
+      "loss": 1.8738,
       "step": 6190
     },
     {
       "epoch": 0.7678007064510132,
-      "grad_norm": 0.25522158800287487,
+      "grad_norm": 0.1366378836050727,
       "learning_rate": 3.1078022474238334e-05,
-      "loss": 3.6484,
+      "loss": 1.8262,
       "step": 6195
     },
     {
       "epoch": 0.7684204003222408,
-      "grad_norm": 0.27734156829652845,
+      "grad_norm": 0.13829425232888887,
       "learning_rate": 3.092143905081794e-05,
-      "loss": 3.7102,
+      "loss": 1.8305,
       "step": 6200
     },
     {
       "epoch": 0.7690400941934684,
-      "grad_norm": 0.26325851682400353,
+      "grad_norm": 0.1353286080001516,
       "learning_rate": 3.07651789157016e-05,
-      "loss": 3.7055,
+      "loss": 1.8023,
       "step": 6205
     },
     {
       "epoch": 0.769659788064696,
-      "grad_norm": 0.27806975585433374,
+      "grad_norm": 0.13398727438252167,
       "learning_rate": 3.060924280018811e-05,
-      "loss": 3.6555,
+      "loss": 1.7555,
       "step": 6210
     },
     {
       "epoch": 0.7702794819359237,
-      "grad_norm": 0.2752537883343241,
+      "grad_norm": 0.14202328524094357,
       "learning_rate": 3.0453631434059958e-05,
-      "loss": 3.6266,
+      "loss": 1.7934,
       "step": 6215
     },
     {
       "epoch": 0.7708991758071513,
-      "grad_norm": 0.2718545931189434,
+      "grad_norm": 0.1333220975895458,
       "learning_rate": 3.0298345545579787e-05,
-      "loss": 3.6258,
+      "loss": 1.832,
       "step": 6220
     },
     {
       "epoch": 0.7715188696783789,
-      "grad_norm": 0.2650532143764226,
+      "grad_norm": 0.1320352616233952,
       "learning_rate": 3.0143385861486974e-05,
-      "loss": 3.5984,
+      "loss": 1.8793,
       "step": 6225
     },
     {
       "epoch": 0.7721385635496065,
-      "grad_norm": 0.2748769017897296,
+      "grad_norm": 0.1492565680814568,
       "learning_rate": 2.9988753106994306e-05,
-      "loss": 3.7234,
+      "loss": 1.8461,
       "step": 6230
     },
     {
       "epoch": 0.7727582574208341,
-      "grad_norm": 0.28224340503720746,
+      "grad_norm": 0.1341082625454802,
       "learning_rate": 2.983444800578452e-05,
-      "loss": 3.7195,
+      "loss": 1.8508,
       "step": 6235
     },
     {
       "epoch": 0.7733779512920618,
-      "grad_norm": 0.27479324651216624,
+      "grad_norm": 0.1357404576853704,
       "learning_rate": 2.9680471280006848e-05,
-      "loss": 3.6797,
+      "loss": 1.8469,
       "step": 6240
     },
     {
       "epoch": 0.7739976451632893,
-      "grad_norm": 0.2767199114775174,
+      "grad_norm": 0.13561038043919632,
       "learning_rate": 2.9526823650273837e-05,
-      "loss": 3.6172,
+      "loss": 1.8293,
       "step": 6245
     },
     {
       "epoch": 0.7746173390345169,
-      "grad_norm": 0.26562638908851804,
+      "grad_norm": 0.14624823865011252,
       "learning_rate": 2.93735058356578e-05,
-      "loss": 3.6367,
+      "loss": 1.7902,
       "step": 6250
     },
     {
       "epoch": 0.7752370329057445,
-      "grad_norm": 0.26872418612230703,
+      "grad_norm": 0.1361302394715922,
       "learning_rate": 2.9220518553687526e-05,
-      "loss": 3.6898,
+      "loss": 1.8559,
       "step": 6255
     },
     {
       "epoch": 0.7758567267769721,
-      "grad_norm": 0.27873653338300347,
+      "grad_norm": 0.14017086141892965,
       "learning_rate": 2.9067862520344956e-05,
-      "loss": 3.7484,
+      "loss": 1.8379,
       "step": 6260
     },
     {
       "epoch": 0.7764764206481998,
-      "grad_norm": 0.263032563358348,
+      "grad_norm": 0.1338549821915282,
       "learning_rate": 2.891553845006165e-05,
-      "loss": 3.6781,
+      "loss": 1.8273,
       "step": 6265
     },
     {
       "epoch": 0.7770961145194274,
-      "grad_norm": 0.2777153283779717,
+      "grad_norm": 0.14098907928260412,
       "learning_rate": 2.87635470557157e-05,
-      "loss": 3.7594,
+      "loss": 1.8465,
       "step": 6270
     },
     {
       "epoch": 0.777715808390655,
-      "grad_norm": 0.2702436881365533,
+      "grad_norm": 0.14080305362194845,
       "learning_rate": 2.861188904862827e-05,
-      "loss": 3.7055,
+      "loss": 1.8457,
       "step": 6275
     },
     {
       "epoch": 0.7783355022618826,
-      "grad_norm": 0.26759599717089805,
+      "grad_norm": 0.12929469590085493,
       "learning_rate": 2.8460565138560212e-05,
-      "loss": 3.7031,
+      "loss": 1.8199,
       "step": 6280
     },
     {
       "epoch": 0.7789551961331103,
-      "grad_norm": 0.26819723161508713,
+      "grad_norm": 0.13426337891967977,
       "learning_rate": 2.830957603370883e-05,
-      "loss": 3.6867,
+      "loss": 1.8402,
       "step": 6285
     },
     {
       "epoch": 0.7795748900043379,
-      "grad_norm": 0.28385416525089446,
+      "grad_norm": 0.13768387356351205,
       "learning_rate": 2.815892244070455e-05,
-      "loss": 3.65,
+      "loss": 1.8809,
       "step": 6290
     },
     {
       "epoch": 0.7801945838755655,
-      "grad_norm": 0.26991854803623094,
+      "grad_norm": 0.1357497474261774,
       "learning_rate": 2.8008605064607528e-05,
-      "loss": 3.6922,
+      "loss": 1.8629,
       "step": 6295
     },
     {
       "epoch": 0.7808142777467931,
-      "grad_norm": 0.2669082534769859,
+      "grad_norm": 0.1380951933528559,
       "learning_rate": 2.7858624608904515e-05,
-      "loss": 3.6484,
+      "loss": 1.9027,
       "step": 6300
     },
     {
       "epoch": 0.7814339716180208,
-      "grad_norm": 0.268626321217136,
+      "grad_norm": 0.14271323304318345,
       "learning_rate": 2.7708981775505416e-05,
-      "loss": 3.6273,
+      "loss": 1.8309,
       "step": 6305
     },
     {
       "epoch": 0.7820536654892483,
-      "grad_norm": 0.2673209837044881,
+      "grad_norm": 0.1421418046102633,
       "learning_rate": 2.755967726474007e-05,
-      "loss": 3.7453,
+      "loss": 1.8266,
       "step": 6310
     },
     {
       "epoch": 0.7826733593604759,
-      "grad_norm": 0.26924472996042415,
+      "grad_norm": 0.14136246949703826,
       "learning_rate": 2.741071177535499e-05,
-      "loss": 3.6211,
+      "loss": 1.8238,
       "step": 6315
     },
     {
       "epoch": 0.7832930532317035,
-      "grad_norm": 0.2849096803234581,
+      "grad_norm": 0.1287667654475233,
       "learning_rate": 2.7262086004510023e-05,
-      "loss": 3.6344,
+      "loss": 1.8285,
       "step": 6320
     },
     {
       "epoch": 0.7839127471029311,
-      "grad_norm": 0.2685985355407374,
+      "grad_norm": 0.1363243768230169,
       "learning_rate": 2.7113800647775156e-05,
-      "loss": 3.7352,
+      "loss": 1.7789,
       "step": 6325
     },
     {
       "epoch": 0.7845324409741588,
-      "grad_norm": 0.2783351123681416,
+      "grad_norm": 0.1403800799180287,
       "learning_rate": 2.6965856399127232e-05,
-      "loss": 3.6258,
+      "loss": 1.8434,
       "step": 6330
     },
     {
       "epoch": 0.7851521348453864,
-      "grad_norm": 0.28724851570545096,
+      "grad_norm": 0.14473756811626665,
       "learning_rate": 2.6818253950946704e-05,
-      "loss": 3.6836,
+      "loss": 1.8477,
       "step": 6335
     },
     {
       "epoch": 0.785771828716614,
-      "grad_norm": 0.2644860014989463,
+      "grad_norm": 0.15235633299839338,
       "learning_rate": 2.6670993994014394e-05,
-      "loss": 3.6313,
+      "loss": 1.8754,
       "step": 6340
     },
     {
       "epoch": 0.7863915225878416,
-      "grad_norm": 0.2781026506109821,
+      "grad_norm": 0.13349212068682054,
       "learning_rate": 2.6524077217508292e-05,
-      "loss": 3.5844,
+      "loss": 1.8309,
       "step": 6345
     },
     {
       "epoch": 0.7870112164590692,
-      "grad_norm": 0.28371233138789237,
+      "grad_norm": 0.14063273951161426,
       "learning_rate": 2.63775043090002e-05,
-      "loss": 3.6945,
+      "loss": 1.8309,
       "step": 6350
     },
     {
       "epoch": 0.7876309103302969,
-      "grad_norm": 0.2676866471094794,
+      "grad_norm": 0.13554326916726495,
       "learning_rate": 2.623127595445274e-05,
-      "loss": 3.6805,
+      "loss": 1.7812,
       "step": 6355
     },
     {
       "epoch": 0.7882506042015245,
-      "grad_norm": 0.2804037276692652,
+      "grad_norm": 0.14617594465762834,
       "learning_rate": 2.6085392838215938e-05,
-      "loss": 3.7062,
+      "loss": 1.8766,
       "step": 6360
     },
     {
       "epoch": 0.7888702980727521,
-      "grad_norm": 0.2677893021797037,
+      "grad_norm": 0.14063474885828448,
       "learning_rate": 2.5939855643024136e-05,
-      "loss": 3.6555,
+      "loss": 1.8336,
       "step": 6365
     },
     {
       "epoch": 0.7894899919439796,
-      "grad_norm": 0.2726351905018635,
+      "grad_norm": 0.13631025672810165,
       "learning_rate": 2.5794665049992762e-05,
-      "loss": 3.6461,
+      "loss": 1.7414,
       "step": 6370
     },
     {
       "epoch": 0.7901096858152072,
-      "grad_norm": 0.2900933246217791,
+      "grad_norm": 0.13754469382158213,
       "learning_rate": 2.564982173861512e-05,
-      "loss": 3.675,
+      "loss": 1.8371,
       "step": 6375
     },
     {
       "epoch": 0.7907293796864349,
-      "grad_norm": 0.2838750953944875,
+      "grad_norm": 0.1373525895674881,
       "learning_rate": 2.5505326386759254e-05,
-      "loss": 3.6828,
+      "loss": 1.8539,
       "step": 6380
     },
     {
       "epoch": 0.7913490735576625,
-      "grad_norm": 0.2694169455941463,
+      "grad_norm": 0.1331753501860663,
       "learning_rate": 2.536117967066476e-05,
-      "loss": 3.6039,
+      "loss": 1.7859,
       "step": 6385
     },
     {
       "epoch": 0.7919687674288901,
-      "grad_norm": 0.26213248615454743,
+      "grad_norm": 0.1320713578647196,
       "learning_rate": 2.521738226493957e-05,
-      "loss": 3.5781,
+      "loss": 1.8109,
       "step": 6390
     },
     {
       "epoch": 0.7925884613001177,
-      "grad_norm": 0.260333774889878,
+      "grad_norm": 0.1383762808515008,
       "learning_rate": 2.50739348425569e-05,
-      "loss": 3.6797,
+      "loss": 1.7887,
       "step": 6395
     },
     {
       "epoch": 0.7932081551713454,
-      "grad_norm": 0.273921652812745,
+      "grad_norm": 0.14212639343800593,
       "learning_rate": 2.4930838074852026e-05,
-      "loss": 3.6359,
+      "loss": 1.8687,
       "step": 6400
     },
     {
       "epoch": 0.793827849042573,
-      "grad_norm": 0.26712516101630807,
+      "grad_norm": 0.1434562089301648,
       "learning_rate": 2.47880926315191e-05,
-      "loss": 3.6078,
+      "loss": 1.8445,
       "step": 6405
     },
     {
       "epoch": 0.7944475429138006,
-      "grad_norm": 0.2693988596271278,
+      "grad_norm": 0.13947244740426135,
       "learning_rate": 2.4645699180608127e-05,
-      "loss": 3.5023,
+      "loss": 1.8414,
       "step": 6410
     },
     {
       "epoch": 0.7950672367850282,
-      "grad_norm": 0.26822140724293647,
+      "grad_norm": 0.13237054669581869,
       "learning_rate": 2.45036583885218e-05,
-      "loss": 3.6805,
+      "loss": 1.8285,
       "step": 6415
     },
     {
       "epoch": 0.7956869306562558,
-      "grad_norm": 0.26510464834957814,
+      "grad_norm": 0.12680327271666422,
       "learning_rate": 2.4361970920012313e-05,
-      "loss": 3.5883,
+      "loss": 1.9254,
       "step": 6420
     },
     {
       "epoch": 0.7963066245274835,
-      "grad_norm": 0.27406804031937193,
+      "grad_norm": 0.1384006795821082,
       "learning_rate": 2.4220637438178317e-05,
-      "loss": 3.7289,
+      "loss": 1.8344,
       "step": 6425
     },
     {
       "epoch": 0.7969263183987111,
-      "grad_norm": 0.2783616044061502,
+      "grad_norm": 0.13493566470479693,
       "learning_rate": 2.4079658604461896e-05,
-      "loss": 3.6969,
+      "loss": 1.8164,
       "step": 6430
     },
     {
       "epoch": 0.7975460122699386,
-      "grad_norm": 0.27499764910548563,
+      "grad_norm": 0.13904637123993435,
       "learning_rate": 2.393903507864521e-05,
-      "loss": 3.75,
+      "loss": 1.8328,
       "step": 6435
     },
     {
       "epoch": 0.7981657061411662,
-      "grad_norm": 0.2810591412318894,
+      "grad_norm": 0.1363832017283468,
       "learning_rate": 2.3798767518847687e-05,
-      "loss": 3.5781,
+      "loss": 1.8594,
       "step": 6440
     },
     {
       "epoch": 0.7987854000123938,
-      "grad_norm": 0.2896760112930254,
+      "grad_norm": 0.13774336173857357,
       "learning_rate": 2.3658856581522804e-05,
-      "loss": 3.6688,
+      "loss": 1.8348,
       "step": 6445
     },
     {
       "epoch": 0.7994050938836215,
-      "grad_norm": 0.27371011149072194,
+      "grad_norm": 0.13986649111637628,
       "learning_rate": 2.3519302921455033e-05,
-      "loss": 3.6703,
+      "loss": 1.8211,
       "step": 6450
     },
     {
       "epoch": 0.8000247877548491,
-      "grad_norm": 0.27322521002152894,
+      "grad_norm": 0.13669443840005838,
       "learning_rate": 2.338010719175684e-05,
-      "loss": 3.4781,
+      "loss": 1.7703,
       "step": 6455
     },
     {
       "epoch": 0.8006444816260767,
-      "grad_norm": 0.27492595731794456,
+      "grad_norm": 0.13007801549105374,
       "learning_rate": 2.324127004386546e-05,
-      "loss": 3.6719,
+      "loss": 1.852,
       "step": 6460
     },
     {
       "epoch": 0.8012641754973043,
-      "grad_norm": 0.269841379201818,
+      "grad_norm": 0.13317280097098233,
       "learning_rate": 2.310279212754006e-05,
-      "loss": 3.6742,
+      "loss": 1.8387,
       "step": 6465
     },
     {
       "epoch": 0.801883869368532,
-      "grad_norm": 0.2631112072013383,
+      "grad_norm": 0.13720871260598752,
       "learning_rate": 2.296467409085853e-05,
-      "loss": 3.7,
+      "loss": 1.8492,
       "step": 6470
     },
     {
       "epoch": 0.8025035632397596,
-      "grad_norm": 0.26544959609730795,
+      "grad_norm": 0.13677530269293067,
       "learning_rate": 2.2826916580214632e-05,
-      "loss": 3.7,
+      "loss": 1.7738,
       "step": 6475
     },
     {
       "epoch": 0.8031232571109872,
-      "grad_norm": 0.27135804528233304,
+      "grad_norm": 0.1383749917899896,
       "learning_rate": 2.2689520240314755e-05,
-      "loss": 3.6781,
+      "loss": 1.8594,
       "step": 6480
     },
     {
       "epoch": 0.8037429509822148,
-      "grad_norm": 0.2653641462255707,
+      "grad_norm": 0.13930908242456305,
       "learning_rate": 2.2552485714175064e-05,
-      "loss": 3.6906,
+      "loss": 1.8461,
       "step": 6485
     },
     {
       "epoch": 0.8043626448534424,
-      "grad_norm": 0.28084635001022096,
+      "grad_norm": 0.1323961629115279,
       "learning_rate": 2.2415813643118356e-05,
-      "loss": 3.8,
+      "loss": 1.8254,
       "step": 6490
     },
     {
       "epoch": 0.8049823387246701,
-      "grad_norm": 0.27853856158817114,
+      "grad_norm": 0.13640022237848362,
       "learning_rate": 2.227950466677121e-05,
-      "loss": 3.5984,
+      "loss": 1.8008,
       "step": 6495
     },
     {
       "epoch": 0.8056020325958976,
-      "grad_norm": 0.2799281087638109,
+      "grad_norm": 0.13393523254678819,
       "learning_rate": 2.21435594230609e-05,
-      "loss": 3.6453,
+      "loss": 1.8148,
       "step": 6500
     },
     {
       "epoch": 0.8062217264671252,
-      "grad_norm": 0.2864919996155976,
+      "grad_norm": 0.13856918377591282,
       "learning_rate": 2.2007978548212425e-05,
-      "loss": 3.7008,
+      "loss": 1.7848,
       "step": 6505
     },
     {
       "epoch": 0.8068414203383528,
-      "grad_norm": 0.2749310670018361,
+      "grad_norm": 0.1467820578644545,
       "learning_rate": 2.1872762676745563e-05,
-      "loss": 3.7477,
+      "loss": 1.8687,
       "step": 6510
     },
     {
       "epoch": 0.8074611142095804,
-      "grad_norm": 0.2694571853494997,
+      "grad_norm": 0.13199784721447735,
       "learning_rate": 2.1737912441471787e-05,
-      "loss": 3.7133,
+      "loss": 1.7953,
       "step": 6515
     },
     {
       "epoch": 0.8080808080808081,
-      "grad_norm": 0.2618758865915513,
+      "grad_norm": 0.13521097371996904,
       "learning_rate": 2.160342847349144e-05,
-      "loss": 3.6539,
+      "loss": 1.8305,
       "step": 6520
     },
     {
       "epoch": 0.8087005019520357,
-      "grad_norm": 0.2733166377034909,
+      "grad_norm": 0.14340242068765546,
       "learning_rate": 2.1469311402190794e-05,
-      "loss": 3.8234,
+      "loss": 1.8246,
       "step": 6525
     },
     {
       "epoch": 0.8093201958232633,
-      "grad_norm": 0.26976744475111125,
+      "grad_norm": 0.13128847998828025,
       "learning_rate": 2.133556185523895e-05,
-      "loss": 3.6914,
+      "loss": 1.8203,
       "step": 6530
     },
     {
       "epoch": 0.8099398896944909,
-      "grad_norm": 0.2714542565616823,
+      "grad_norm": 0.13591682961831256,
       "learning_rate": 2.120218045858503e-05,
-      "loss": 3.7508,
+      "loss": 1.8637,
       "step": 6535
     },
     {
       "epoch": 0.8105595835657186,
-      "grad_norm": 0.2669981069887573,
+      "grad_norm": 0.14188949741756163,
       "learning_rate": 2.1069167836455228e-05,
-      "loss": 3.6297,
+      "loss": 1.823,
       "step": 6540
     },
     {
       "epoch": 0.8111792774369462,
-      "grad_norm": 0.2634714817754169,
+      "grad_norm": 0.14951548768393097,
       "learning_rate": 2.0936524611349795e-05,
-      "loss": 3.5695,
+      "loss": 1.7926,
       "step": 6545
     },
     {
       "epoch": 0.8117989713081738,
-      "grad_norm": 0.2800560817358675,
+      "grad_norm": 0.1463305546882131,
       "learning_rate": 2.080425140404029e-05,
-      "loss": 3.6023,
+      "loss": 1.8512,
       "step": 6550
     },
     {
       "epoch": 0.8124186651794014,
-      "grad_norm": 0.26936666732046705,
+      "grad_norm": 0.1370880820734347,
       "learning_rate": 2.0672348833566512e-05,
-      "loss": 3.7266,
+      "loss": 1.8703,
       "step": 6555
     },
     {
       "epoch": 0.813038359050629,
-      "grad_norm": 0.2782298698283211,
+      "grad_norm": 0.13265798046306984,
       "learning_rate": 2.0540817517233735e-05,
-      "loss": 3.7586,
+      "loss": 1.8309,
       "step": 6560
     },
     {
       "epoch": 0.8136580529218566,
-      "grad_norm": 0.27492647831446826,
+      "grad_norm": 0.1298504140699306,
       "learning_rate": 2.0409658070609738e-05,
-      "loss": 3.6,
+      "loss": 1.8246,
       "step": 6565
     },
     {
       "epoch": 0.8142777467930842,
-      "grad_norm": 0.2675626790996732,
+      "grad_norm": 0.13644051403236102,
       "learning_rate": 2.0278871107521936e-05,
-      "loss": 3.6469,
+      "loss": 1.8023,
       "step": 6570
     },
     {
       "epoch": 0.8148974406643118,
-      "grad_norm": 0.27782396820372657,
+      "grad_norm": 0.1397736858054863,
       "learning_rate": 2.014845724005453e-05,
-      "loss": 3.6766,
+      "loss": 1.8457,
       "step": 6575
     },
     {
       "epoch": 0.8155171345355394,
-      "grad_norm": 0.26540581302587424,
+      "grad_norm": 0.13415177608850457,
       "learning_rate": 2.0018417078545614e-05,
-      "loss": 3.7437,
+      "loss": 1.8301,
       "step": 6580
     },
     {
       "epoch": 0.816136828406767,
-      "grad_norm": 0.270984981257796,
+      "grad_norm": 0.13204065568378437,
       "learning_rate": 1.988875123158437e-05,
-      "loss": 3.6602,
+      "loss": 1.802,
       "step": 6585
     },
     {
       "epoch": 0.8167565222779947,
-      "grad_norm": 0.26379727764864,
+      "grad_norm": 0.1372522407858421,
       "learning_rate": 1.975946030600814e-05,
-      "loss": 3.7078,
+      "loss": 1.7984,
       "step": 6590
     },
     {
       "epoch": 0.8173762161492223,
-      "grad_norm": 0.28352021928974896,
+      "grad_norm": 0.1402427316379218,
       "learning_rate": 1.9630544906899672e-05,
-      "loss": 3.6172,
+      "loss": 1.7902,
       "step": 6595
     },
     {
       "epoch": 0.8179959100204499,
-      "grad_norm": 0.2929059747351139,
+      "grad_norm": 0.13504947552206878,
       "learning_rate": 1.9502005637584198e-05,
-      "loss": 3.7469,
+      "loss": 1.8594,
       "step": 6600
     },
     {
       "epoch": 0.8186156038916775,
-      "grad_norm": 0.2621616613487277,
+      "grad_norm": 0.13671633407154715,
       "learning_rate": 1.93738430996267e-05,
-      "loss": 3.6984,
+      "loss": 1.8473,
       "step": 6605
     },
     {
       "epoch": 0.8192352977629052,
-      "grad_norm": 0.2543683174187999,
+      "grad_norm": 0.14505886127883016,
       "learning_rate": 1.9246057892829038e-05,
-      "loss": 3.6883,
+      "loss": 1.8297,
       "step": 6610
     },
     {
       "epoch": 0.8198549916341328,
-      "grad_norm": 0.2676990427144938,
+      "grad_norm": 0.13744286662175678,
       "learning_rate": 1.9118650615227162e-05,
-      "loss": 3.6273,
+      "loss": 1.8254,
       "step": 6615
     },
     {
       "epoch": 0.8204746855053604,
-      "grad_norm": 0.26701918525313256,
+      "grad_norm": 0.1378170278876299,
       "learning_rate": 1.8991621863088315e-05,
-      "loss": 3.7531,
+      "loss": 1.8301,
       "step": 6620
     },
     {
       "epoch": 0.8210943793765879,
-      "grad_norm": 0.2735829555403975,
+      "grad_norm": 0.13425156743572028,
       "learning_rate": 1.886497223090823e-05,
-      "loss": 3.6688,
+      "loss": 1.8391,
       "step": 6625
     },
     {
       "epoch": 0.8217140732478155,
-      "grad_norm": 0.2586013500555307,
+      "grad_norm": 0.136925795997487,
       "learning_rate": 1.8738702311408352e-05,
-      "loss": 3.8141,
+      "loss": 1.8523,
       "step": 6630
     },
     {
       "epoch": 0.8223337671190432,
-      "grad_norm": 0.2682042573349656,
+      "grad_norm": 0.13555305693474248,
       "learning_rate": 1.8612812695533077e-05,
-      "loss": 3.668,
+      "loss": 1.9012,
       "step": 6635
     },
     {
       "epoch": 0.8229534609902708,
-      "grad_norm": 0.2798025021737809,
+      "grad_norm": 0.14021386873564234,
       "learning_rate": 1.8487303972446966e-05,
-      "loss": 3.6023,
+      "loss": 1.8211,
       "step": 6640
     },
     {
       "epoch": 0.8235731548614984,
-      "grad_norm": 0.2665439914850495,
+      "grad_norm": 0.13435340169695256,
       "learning_rate": 1.836217672953201e-05,
-      "loss": 3.7883,
+      "loss": 1.8738,
       "step": 6645
     },
     {
       "epoch": 0.824192848732726,
-      "grad_norm": 0.2782919179548854,
+      "grad_norm": 0.13825299288157739,
       "learning_rate": 1.8237431552384887e-05,
-      "loss": 3.5961,
+      "loss": 1.8168,
       "step": 6650
     },
     {
       "epoch": 0.8248125426039536,
-      "grad_norm": 0.2654144345290904,
+      "grad_norm": 0.1365578497978942,
       "learning_rate": 1.811306902481412e-05,
-      "loss": 3.7172,
+      "loss": 1.7988,
       "step": 6655
     },
     {
       "epoch": 0.8254322364751813,
-      "grad_norm": 0.27250768303187856,
+      "grad_norm": 0.14191275010866652,
       "learning_rate": 1.798908972883754e-05,
-      "loss": 3.6125,
+      "loss": 1.8621,
       "step": 6660
     },
     {
       "epoch": 0.8260519303464089,
-      "grad_norm": 0.2679383566934312,
+      "grad_norm": 0.14261769291356297,
       "learning_rate": 1.786549424467936e-05,
-      "loss": 3.7078,
+      "loss": 1.95,
       "step": 6665
     },
     {
       "epoch": 0.8266716242176365,
-      "grad_norm": 0.270235333933691,
+      "grad_norm": 0.1476188503443379,
       "learning_rate": 1.7742283150767614e-05,
-      "loss": 3.6305,
+      "loss": 1.9102,
       "step": 6670
     },
     {
       "epoch": 0.8272913180888641,
-      "grad_norm": 0.27558887009789595,
+      "grad_norm": 0.130356698831072,
       "learning_rate": 1.7619457023731355e-05,
-      "loss": 3.6703,
+      "loss": 1.8477,
       "step": 6675
     },
     {
       "epoch": 0.8279110119600918,
-      "grad_norm": 0.2683680281474502,
+      "grad_norm": 0.13967859191640508,
       "learning_rate": 1.7497016438397984e-05,
-      "loss": 3.6719,
+      "loss": 1.8336,
       "step": 6680
     },
     {
       "epoch": 0.8285307058313194,
-      "grad_norm": 0.2730520742473007,
+      "grad_norm": 0.1445625801135568,
       "learning_rate": 1.737496196779059e-05,
-      "loss": 3.6461,
+      "loss": 1.7988,
       "step": 6685
     },
     {
       "epoch": 0.8291503997025469,
-      "grad_norm": 0.2936511481286826,
+      "grad_norm": 0.13772616272180707,
       "learning_rate": 1.7253294183125223e-05,
-      "loss": 3.5742,
+      "loss": 1.8387,
       "step": 6690
     },
     {
       "epoch": 0.8297700935737745,
-      "grad_norm": 0.28855084456291186,
+      "grad_norm": 0.13927654860478775,
       "learning_rate": 1.7132013653808222e-05,
-      "loss": 3.6719,
+      "loss": 1.8344,
       "step": 6695
     },
     {
       "epoch": 0.8303897874450021,
-      "grad_norm": 0.2597746796894285,
+      "grad_norm": 0.1367478820289552,
       "learning_rate": 1.70111209474336e-05,
-      "loss": 3.6102,
+      "loss": 1.8711,
       "step": 6700
     },
     {
       "epoch": 0.8310094813162298,
-      "grad_norm": 0.27814469005937525,
+      "grad_norm": 0.13778305851562409,
       "learning_rate": 1.6890616629780364e-05,
-      "loss": 3.7359,
+      "loss": 1.8105,
       "step": 6705
     },
     {
       "epoch": 0.8316291751874574,
-      "grad_norm": 0.277325774566865,
+      "grad_norm": 0.13937638411283737,
       "learning_rate": 1.6770501264809778e-05,
-      "loss": 3.6242,
+      "loss": 1.8516,
       "step": 6710
     },
     {
       "epoch": 0.832248869058685,
-      "grad_norm": 0.2693934876663354,
+      "grad_norm": 0.14142715677631995,
       "learning_rate": 1.665077541466289e-05,
-      "loss": 3.7117,
+      "loss": 1.848,
       "step": 6715
     },
     {
       "epoch": 0.8328685629299126,
-      "grad_norm": 0.2663304575953115,
+      "grad_norm": 0.13547775806902937,
       "learning_rate": 1.6531439639657776e-05,
-      "loss": 3.7172,
+      "loss": 1.8512,
       "step": 6720
     },
     {
       "epoch": 0.8334882568011402,
-      "grad_norm": 0.2680430225306164,
+      "grad_norm": 0.13259567267028294,
       "learning_rate": 1.641249449828699e-05,
-      "loss": 3.7172,
+      "loss": 1.9055,
       "step": 6725
     },
     {
       "epoch": 0.8341079506723679,
-      "grad_norm": 0.28908859826508604,
+      "grad_norm": 0.13430439200938232,
       "learning_rate": 1.6293940547214905e-05,
-      "loss": 3.8102,
+      "loss": 1.743,
       "step": 6730
     },
     {
       "epoch": 0.8347276445435955,
-      "grad_norm": 0.2723242580613196,
+      "grad_norm": 0.14323647170257087,
       "learning_rate": 1.617577834127506e-05,
-      "loss": 3.7539,
+      "loss": 1.7785,
       "step": 6735
     },
     {
       "epoch": 0.8353473384148231,
-      "grad_norm": 0.26396208033874435,
+      "grad_norm": 0.13887966152843348,
       "learning_rate": 1.6058008433467698e-05,
-      "loss": 3.618,
+      "loss": 1.7848,
       "step": 6740
     },
     {
       "epoch": 0.8359670322860507,
-      "grad_norm": 0.2690220541626815,
+      "grad_norm": 0.12926377677093945,
       "learning_rate": 1.594063137495707e-05,
-      "loss": 3.807,
+      "loss": 1.8285,
       "step": 6745
     },
     {
       "epoch": 0.8365867261572784,
-      "grad_norm": 0.28665826800813404,
+      "grad_norm": 0.13985627925415461,
       "learning_rate": 1.582364771506891e-05,
-      "loss": 3.6242,
+      "loss": 1.8617,
       "step": 6750
     },
     {
       "epoch": 0.8372064200285059,
-      "grad_norm": 0.27065792545829087,
+      "grad_norm": 0.14071537495548447,
       "learning_rate": 1.570705800128781e-05,
-      "loss": 3.6313,
+      "loss": 1.798,
       "step": 6755
     },
     {
       "epoch": 0.8378261138997335,
-      "grad_norm": 0.2607272849200065,
+      "grad_norm": 0.14005570372758716,
       "learning_rate": 1.5590862779254746e-05,
-      "loss": 3.6539,
+      "loss": 1.807,
       "step": 6760
     },
     {
       "epoch": 0.8384458077709611,
-      "grad_norm": 0.2787653779356175,
+      "grad_norm": 0.1373432585304391,
       "learning_rate": 1.5475062592764346e-05,
-      "loss": 3.682,
+      "loss": 1.8082,
       "step": 6765
     },
     {
       "epoch": 0.8390655016421887,
-      "grad_norm": 0.27568525546442,
+      "grad_norm": 0.13648916484535767,
       "learning_rate": 1.5359657983762632e-05,
-      "loss": 3.6875,
+      "loss": 1.8035,
       "step": 6770
     },
     {
       "epoch": 0.8396851955134164,
-      "grad_norm": 0.2950892796467241,
+      "grad_norm": 0.13659166896958833,
       "learning_rate": 1.524464949234422e-05,
-      "loss": 3.6586,
+      "loss": 1.8188,
       "step": 6775
     },
     {
       "epoch": 0.840304889384644,
-      "grad_norm": 0.2764190380094884,
+      "grad_norm": 0.13285103674527363,
       "learning_rate": 1.5130037656749918e-05,
-      "loss": 3.5883,
+      "loss": 1.7789,
       "step": 6780
     },
     {
       "epoch": 0.8409245832558716,
-      "grad_norm": 0.26761586390041553,
+      "grad_norm": 0.13234536077305287,
       "learning_rate": 1.5015823013364183e-05,
-      "loss": 3.7297,
+      "loss": 1.8586,
       "step": 6785
     },
     {
       "epoch": 0.8415442771270992,
-      "grad_norm": 0.2912541676210491,
+      "grad_norm": 0.13222815236741142,
       "learning_rate": 1.4902006096712572e-05,
-      "loss": 3.5719,
+      "loss": 1.7836,
       "step": 6790
     },
     {
       "epoch": 0.8421639709983268,
-      "grad_norm": 0.27254549706733305,
+      "grad_norm": 0.13495804055854296,
       "learning_rate": 1.4788587439459323e-05,
-      "loss": 3.6945,
+      "loss": 1.8051,
       "step": 6795
     },
     {
       "epoch": 0.8427836648695545,
-      "grad_norm": 0.2620221582714912,
+      "grad_norm": 0.13630743372332232,
       "learning_rate": 1.4675567572404803e-05,
-      "loss": 3.7812,
+      "loss": 1.8426,
       "step": 6800
     },
     {
       "epoch": 0.8434033587407821,
-      "grad_norm": 0.289922770515983,
+      "grad_norm": 0.13164111897888933,
       "learning_rate": 1.4562947024483031e-05,
-      "loss": 3.6508,
+      "loss": 1.7629,
       "step": 6805
     },
     {
       "epoch": 0.8440230526120097,
-      "grad_norm": 0.2616919452385473,
+      "grad_norm": 0.13579891156886684,
       "learning_rate": 1.4450726322759223e-05,
-      "loss": 3.6617,
+      "loss": 1.8094,
       "step": 6810
     },
     {
       "epoch": 0.8446427464832372,
-      "grad_norm": 0.2847412415269478,
+      "grad_norm": 0.13553237661955567,
       "learning_rate": 1.4338905992427287e-05,
-      "loss": 3.6508,
+      "loss": 1.8687,
       "step": 6815
     },
     {
       "epoch": 0.8452624403544649,
-      "grad_norm": 0.2791306174520287,
+      "grad_norm": 0.14720160991862427,
       "learning_rate": 1.4227486556807412e-05,
-      "loss": 3.693,
+      "loss": 1.8113,
       "step": 6820
     },
     {
       "epoch": 0.8458821342256925,
-      "grad_norm": 0.27972768023171085,
+      "grad_norm": 0.1340030456048872,
       "learning_rate": 1.4116468537343585e-05,
-      "loss": 3.7508,
+      "loss": 1.8234,
       "step": 6825
     },
     {
       "epoch": 0.8465018280969201,
-      "grad_norm": 0.27808447135210174,
+      "grad_norm": 0.14666047167725307,
       "learning_rate": 1.4005852453601164e-05,
-      "loss": 3.6789,
+      "loss": 1.8621,
       "step": 6830
     },
     {
       "epoch": 0.8471215219681477,
-      "grad_norm": 0.26347659861052786,
+      "grad_norm": 0.13668555746075953,
       "learning_rate": 1.3895638823264446e-05,
-      "loss": 3.7016,
+      "loss": 1.848,
       "step": 6835
     },
     {
       "epoch": 0.8477412158393753,
-      "grad_norm": 0.2860968542159929,
+      "grad_norm": 0.1339418741254649,
       "learning_rate": 1.3785828162134252e-05,
-      "loss": 3.6992,
+      "loss": 1.834,
       "step": 6840
     },
     {
       "epoch": 0.848360909710603,
-      "grad_norm": 0.2711258258595594,
+      "grad_norm": 0.12912937848665926,
       "learning_rate": 1.367642098412546e-05,
-      "loss": 3.6148,
+      "loss": 1.8633,
       "step": 6845
     },
     {
       "epoch": 0.8489806035818306,
-      "grad_norm": 0.29543579306133466,
+      "grad_norm": 0.13665464782157216,
       "learning_rate": 1.3567417801264692e-05,
-      "loss": 3.6797,
+      "loss": 1.8273,
       "step": 6850
     },
     {
       "epoch": 0.8496002974530582,
-      "grad_norm": 0.26902475678532334,
+      "grad_norm": 0.14072966290231814,
       "learning_rate": 1.345881912368785e-05,
-      "loss": 3.6586,
+      "loss": 1.8637,
       "step": 6855
     },
     {
       "epoch": 0.8502199913242858,
-      "grad_norm": 0.25812138980263255,
+      "grad_norm": 0.139383826652858,
       "learning_rate": 1.3350625459637744e-05,
-      "loss": 3.7273,
+      "loss": 1.8113,
       "step": 6860
     },
     {
       "epoch": 0.8508396851955135,
-      "grad_norm": 0.2830901391479715,
+      "grad_norm": 0.1347401215990617,
       "learning_rate": 1.3242837315461732e-05,
-      "loss": 3.6719,
+      "loss": 1.8414,
       "step": 6865
     },
     {
       "epoch": 0.8514593790667411,
-      "grad_norm": 0.29031597061519,
+      "grad_norm": 0.14070257303503275,
       "learning_rate": 1.3135455195609325e-05,
-      "loss": 3.7188,
+      "loss": 1.8148,
       "step": 6870
     },
     {
       "epoch": 0.8520790729379687,
-      "grad_norm": 0.2728925825069603,
+      "grad_norm": 0.1317491498855389,
       "learning_rate": 1.3028479602629839e-05,
-      "loss": 3.6187,
+      "loss": 1.8465,
       "step": 6875
     },
     {
       "epoch": 0.8526987668091962,
-      "grad_norm": 0.2764576115800626,
+      "grad_norm": 0.13406734047768462,
       "learning_rate": 1.292191103717002e-05,
-      "loss": 3.7492,
+      "loss": 1.8473,
       "step": 6880
     },
     {
       "epoch": 0.8533184606804238,
-      "grad_norm": 0.2660636993335465,
+      "grad_norm": 0.13835653776561033,
       "learning_rate": 1.281574999797176e-05,
-      "loss": 3.5781,
+      "loss": 1.8797,
       "step": 6885
     },
     {
       "epoch": 0.8539381545516515,
-      "grad_norm": 0.28013707569590557,
+      "grad_norm": 0.1376875291822998,
       "learning_rate": 1.2709996981869699e-05,
-      "loss": 3.8156,
+      "loss": 1.7785,
       "step": 6890
     },
     {
       "epoch": 0.8545578484228791,
-      "grad_norm": 0.27399258918575153,
+      "grad_norm": 0.13107009499549113,
       "learning_rate": 1.2604652483788948e-05,
-      "loss": 3.575,
+      "loss": 1.8418,
       "step": 6895
     },
     {
       "epoch": 0.8551775422941067,
-      "grad_norm": 0.2731456828634934,
+      "grad_norm": 0.1439061599075864,
       "learning_rate": 1.2499716996742694e-05,
-      "loss": 3.5609,
+      "loss": 1.8465,
       "step": 6900
     },
     {
       "epoch": 0.8557972361653343,
-      "grad_norm": 0.26547407841220094,
+      "grad_norm": 0.13615157558842733,
       "learning_rate": 1.2395191011829999e-05,
-      "loss": 3.6492,
+      "loss": 1.8504,
       "step": 6905
     },
     {
       "epoch": 0.8564169300365619,
-      "grad_norm": 0.27355370923211936,
+      "grad_norm": 0.13740008939070456,
       "learning_rate": 1.2291075018233445e-05,
-      "loss": 3.6117,
+      "loss": 1.8633,
       "step": 6910
     },
     {
       "epoch": 0.8570366239077896,
-      "grad_norm": 0.2722276850451908,
+      "grad_norm": 0.1416192880003319,
       "learning_rate": 1.218736950321685e-05,
-      "loss": 3.7898,
+      "loss": 1.773,
       "step": 6915
     },
     {
       "epoch": 0.8576563177790172,
-      "grad_norm": 0.2711882167084163,
+      "grad_norm": 0.13447661457180524,
       "learning_rate": 1.2084074952122959e-05,
-      "loss": 3.6336,
+      "loss": 1.8652,
       "step": 6920
     },
     {
       "epoch": 0.8582760116502448,
-      "grad_norm": 0.2657796423409746,
+      "grad_norm": 0.13856068621293663,
       "learning_rate": 1.1981191848371287e-05,
-      "loss": 3.6664,
+      "loss": 1.8359,
       "step": 6925
     },
     {
       "epoch": 0.8588957055214724,
-      "grad_norm": 0.2672916194424286,
+      "grad_norm": 0.13705953632821657,
       "learning_rate": 1.1878720673455645e-05,
-      "loss": 3.782,
+      "loss": 1.8625,
       "step": 6930
     },
     {
       "epoch": 0.8595153993927,
-      "grad_norm": 0.2775303155331989,
+      "grad_norm": 0.14394551527944272,
       "learning_rate": 1.1776661906942099e-05,
-      "loss": 3.5602,
+      "loss": 1.8199,
       "step": 6935
     },
     {
       "epoch": 0.8601350932639277,
-      "grad_norm": 0.28179048470431456,
+      "grad_norm": 0.14385500400158302,
       "learning_rate": 1.1675016026466633e-05,
-      "loss": 3.6703,
+      "loss": 1.7957,
       "step": 6940
     },
     {
       "epoch": 0.8607547871351552,
-      "grad_norm": 0.2721522958602641,
+      "grad_norm": 0.13886377592614751,
       "learning_rate": 1.1573783507732893e-05,
-      "loss": 3.5633,
+      "loss": 1.8582,
       "step": 6945
     },
     {
       "epoch": 0.8613744810063828,
-      "grad_norm": 0.29373625990302477,
+      "grad_norm": 0.14377812118726135,
       "learning_rate": 1.1472964824510035e-05,
-      "loss": 3.6336,
+      "loss": 1.8348,
       "step": 6950
     },
     {
       "epoch": 0.8619941748776104,
-      "grad_norm": 0.28485020077333817,
+      "grad_norm": 0.13169196792358331,
       "learning_rate": 1.1372560448630376e-05,
-      "loss": 3.7281,
+      "loss": 1.8074,
       "step": 6955
     },
     {
       "epoch": 0.862613868748838,
-      "grad_norm": 0.29189846663791436,
+      "grad_norm": 0.13793344384621267,
       "learning_rate": 1.1272570849987351e-05,
-      "loss": 3.5758,
+      "loss": 1.8148,
       "step": 6960
     },
     {
       "epoch": 0.8632335626200657,
-      "grad_norm": 0.2727089581422032,
+      "grad_norm": 0.13768118568152948,
       "learning_rate": 1.1172996496533194e-05,
-      "loss": 3.5297,
+      "loss": 1.7746,
       "step": 6965
     },
     {
       "epoch": 0.8638532564912933,
-      "grad_norm": 0.27053269853938855,
+      "grad_norm": 0.1372503445366227,
       "learning_rate": 1.1073837854276826e-05,
-      "loss": 3.6812,
+      "loss": 1.8055,
       "step": 6970
     },
     {
       "epoch": 0.8644729503625209,
-      "grad_norm": 0.270257705336746,
+      "grad_norm": 0.14304524613047395,
       "learning_rate": 1.0975095387281587e-05,
-      "loss": 3.6547,
+      "loss": 1.8211,
       "step": 6975
     },
     {
       "epoch": 0.8650926442337485,
-      "grad_norm": 0.2709254834993136,
+      "grad_norm": 0.1341063071760674,
       "learning_rate": 1.087676955766318e-05,
-      "loss": 3.6617,
+      "loss": 1.8406,
       "step": 6980
     },
     {
       "epoch": 0.8657123381049762,
-      "grad_norm": 0.26787101687269665,
+      "grad_norm": 0.1369014351632983,
       "learning_rate": 1.0778860825587323e-05,
-      "loss": 3.7102,
+      "loss": 1.8379,
       "step": 6985
     },
     {
       "epoch": 0.8663320319762038,
-      "grad_norm": 0.25834763408294725,
+      "grad_norm": 0.13986771142969012,
       "learning_rate": 1.0681369649267836e-05,
-      "loss": 3.5211,
+      "loss": 1.8508,
       "step": 6990
     },
     {
       "epoch": 0.8669517258474314,
-      "grad_norm": 0.2772978504813741,
+      "grad_norm": 0.13871013013154287,
       "learning_rate": 1.0584296484964318e-05,
-      "loss": 3.6898,
+      "loss": 1.8746,
       "step": 6995
     },
     {
       "epoch": 0.867571419718659,
-      "grad_norm": 0.2672970881451939,
+      "grad_norm": 0.13536162110331093,
       "learning_rate": 1.0487641786980063e-05,
-      "loss": 3.6641,
+      "loss": 1.8387,
       "step": 7000
     },
     {
       "epoch": 0.8681911135898865,
-      "grad_norm": 0.27556223642075184,
+      "grad_norm": 0.13688561948696315,
       "learning_rate": 1.0391406007659964e-05,
-      "loss": 3.6203,
+      "loss": 1.7699,
       "step": 7005
     },
     {
       "epoch": 0.8688108074611142,
-      "grad_norm": 0.27886193281631777,
+      "grad_norm": 0.1379553663310307,
       "learning_rate": 1.0295589597388355e-05,
-      "loss": 3.6961,
+      "loss": 1.8297,
       "step": 7010
     },
     {
       "epoch": 0.8694305013323418,
-      "grad_norm": 0.2730774804147792,
+      "grad_norm": 0.1351138643848652,
       "learning_rate": 1.0200193004586922e-05,
-      "loss": 3.6781,
+      "loss": 1.8023,
       "step": 7015
     },
     {
       "epoch": 0.8700501952035694,
-      "grad_norm": 0.28077069068129035,
+      "grad_norm": 0.1352093710005571,
       "learning_rate": 1.0105216675712592e-05,
-      "loss": 3.6867,
+      "loss": 1.827,
       "step": 7020
     },
     {
       "epoch": 0.870669889074797,
-      "grad_norm": 0.2703429141608011,
+      "grad_norm": 0.14066110824426709,
       "learning_rate": 1.0010661055255488e-05,
-      "loss": 3.7523,
+      "loss": 1.8387,
       "step": 7025
     },
     {
       "epoch": 0.8712895829460247,
-      "grad_norm": 0.2623300718988336,
+      "grad_norm": 0.14804843977254,
       "learning_rate": 9.916526585736763e-06,
-      "loss": 3.5602,
+      "loss": 1.8551,
       "step": 7030
     },
     {
       "epoch": 0.8719092768172523,
-      "grad_norm": 0.27726179794151795,
+      "grad_norm": 0.13708112000990683,
       "learning_rate": 9.822813707706625e-06,
-      "loss": 3.5547,
+      "loss": 1.8391,
       "step": 7035
     },
     {
       "epoch": 0.8725289706884799,
-      "grad_norm": 0.273938282589832,
+      "grad_norm": 0.13764098800291605,
       "learning_rate": 9.729522859742191e-06,
-      "loss": 3.6773,
+      "loss": 1.8969,
       "step": 7040
     },
     {
       "epoch": 0.8731486645597075,
-      "grad_norm": 0.2626253209294718,
+      "grad_norm": 0.14148033017124853,
       "learning_rate": 9.636654478445494e-06,
-      "loss": 3.7633,
+      "loss": 1.8402,
       "step": 7045
     },
     {
       "epoch": 0.8737683584309351,
-      "grad_norm": 0.2709833878561481,
+      "grad_norm": 0.14686796135710695,
       "learning_rate": 9.544208998441428e-06,
-      "loss": 3.5719,
+      "loss": 1.8852,
       "step": 7050
     },
     {
       "epoch": 0.8743880523021628,
-      "grad_norm": 0.2688365440025024,
+      "grad_norm": 0.12880894029142761,
       "learning_rate": 9.452186852375678e-06,
-      "loss": 3.6688,
+      "loss": 1.8914,
       "step": 7055
     },
     {
       "epoch": 0.8750077461733904,
-      "grad_norm": 0.27760423536798684,
+      "grad_norm": 0.13556387811025775,
       "learning_rate": 9.360588470912756e-06,
-      "loss": 3.6711,
+      "loss": 1.8605,
       "step": 7060
     },
     {
       "epoch": 0.875627440044618,
-      "grad_norm": 0.2715108252926431,
+      "grad_norm": 0.142830646840542,
       "learning_rate": 9.269414282733924e-06,
-      "loss": 3.6313,
+      "loss": 1.8789,
       "step": 7065
     },
     {
       "epoch": 0.8762471339158455,
-      "grad_norm": 0.2820261278373799,
+      "grad_norm": 0.14113733132481313,
       "learning_rate": 9.178664714535235e-06,
-      "loss": 3.6602,
+      "loss": 1.8367,
       "step": 7070
     },
     {
       "epoch": 0.8768668277870731,
-      "grad_norm": 0.27157932121191247,
+      "grad_norm": 0.1380731856123245,
       "learning_rate": 9.088340191025501e-06,
-      "loss": 3.5891,
+      "loss": 1.859,
       "step": 7075
     },
     {
       "epoch": 0.8774865216583008,
-      "grad_norm": 0.26208545069180406,
+      "grad_norm": 0.13483398239386993,
       "learning_rate": 8.998441134924318e-06,
-      "loss": 3.8312,
+      "loss": 1.8738,
       "step": 7080
     },
     {
       "epoch": 0.8781062155295284,
-      "grad_norm": 0.2869454691805871,
+      "grad_norm": 0.13921045630136186,
       "learning_rate": 8.908967966960124e-06,
-      "loss": 3.6219,
+      "loss": 1.891,
       "step": 7085
     },
     {
       "epoch": 0.878725909400756,
-      "grad_norm": 0.2774369353780167,
+      "grad_norm": 0.14820885945750578,
       "learning_rate": 8.819921105868167e-06,
-      "loss": 3.7266,
+      "loss": 1.8176,
       "step": 7090
     },
     {
       "epoch": 0.8793456032719836,
-      "grad_norm": 0.2741228692101217,
+      "grad_norm": 0.13074251738222936,
       "learning_rate": 8.731300968388556e-06,
-      "loss": 3.7305,
+      "loss": 1.8359,
       "step": 7095
     },
     {
       "epoch": 0.8799652971432113,
-      "grad_norm": 0.27480937404345795,
+      "grad_norm": 0.13913021647845775,
       "learning_rate": 8.643107969264375e-06,
-      "loss": 3.6227,
+      "loss": 1.877,
       "step": 7100
     },
     {
       "epoch": 0.8805849910144389,
-      "grad_norm": 0.27131019564431835,
+      "grad_norm": 0.15160693761931768,
       "learning_rate": 8.555342521239662e-06,
-      "loss": 3.6844,
+      "loss": 1.8176,
       "step": 7105
     },
     {
       "epoch": 0.8812046848856665,
-      "grad_norm": 0.28727731987544974,
+      "grad_norm": 0.14319289119952988,
       "learning_rate": 8.468005035057536e-06,
-      "loss": 3.7648,
+      "loss": 1.8672,
       "step": 7110
     },
     {
       "epoch": 0.8818243787568941,
-      "grad_norm": 0.2772638259965115,
+      "grad_norm": 0.144700531964563,
       "learning_rate": 8.381095919458226e-06,
-      "loss": 3.5984,
+      "loss": 1.8527,
       "step": 7115
     },
     {
       "epoch": 0.8824440726281217,
-      "grad_norm": 0.2767919200596619,
+      "grad_norm": 0.14274877382833165,
       "learning_rate": 8.294615581177223e-06,
-      "loss": 3.6305,
+      "loss": 1.8434,
       "step": 7120
     },
     {
       "epoch": 0.8830637664993494,
-      "grad_norm": 0.275136982546241,
+      "grad_norm": 0.14240350591746134,
       "learning_rate": 8.208564424943288e-06,
-      "loss": 3.6969,
+      "loss": 1.8141,
       "step": 7125
     },
     {
       "epoch": 0.883683460370577,
-      "grad_norm": 0.2820723985783487,
+      "grad_norm": 0.13397820323823306,
       "learning_rate": 8.122942853476633e-06,
-      "loss": 3.6609,
+      "loss": 1.8988,
       "step": 7130
     },
     {
       "epoch": 0.8843031542418045,
-      "grad_norm": 0.2668391059350675,
+      "grad_norm": 0.14005322447201954,
       "learning_rate": 8.037751267487003e-06,
-      "loss": 3.657,
+      "loss": 1.8488,
       "step": 7135
     },
     {
       "epoch": 0.8849228481130321,
-      "grad_norm": 0.2721769752050762,
+      "grad_norm": 0.13988626846631608,
       "learning_rate": 7.952990065671817e-06,
-      "loss": 3.7703,
+      "loss": 1.8379,
       "step": 7140
     },
     {
       "epoch": 0.8855425419842597,
-      "grad_norm": 0.26632991478502493,
+      "grad_norm": 0.13867297243276785,
       "learning_rate": 7.868659644714294e-06,
-      "loss": 3.7148,
+      "loss": 1.8148,
       "step": 7145
     },
     {
       "epoch": 0.8861622358554874,
-      "grad_norm": 0.27844675439554767,
+      "grad_norm": 0.13631205393737392,
       "learning_rate": 7.784760399281554e-06,
-      "loss": 3.6,
+      "loss": 1.873,
       "step": 7150
     },
     {
       "epoch": 0.886781929726715,
-      "grad_norm": 0.27248180350893103,
+      "grad_norm": 0.13939741587279572,
       "learning_rate": 7.701292722022846e-06,
-      "loss": 3.7742,
+      "loss": 1.8008,
       "step": 7155
     },
     {
       "epoch": 0.8874016235979426,
-      "grad_norm": 0.29819852968664107,
+      "grad_norm": 0.13560938210869783,
       "learning_rate": 7.618257003567675e-06,
-      "loss": 3.6305,
+      "loss": 1.8172,
       "step": 7160
     },
     {
       "epoch": 0.8880213174691702,
-      "grad_norm": 0.2731747004186133,
+      "grad_norm": 0.13381257956234077,
       "learning_rate": 7.5356536325239755e-06,
-      "loss": 3.7445,
+      "loss": 1.773,
       "step": 7165
     },
     {
       "epoch": 0.8886410113403979,
-      "grad_norm": 0.2862676566426175,
+      "grad_norm": 0.1356484036949059,
       "learning_rate": 7.453482995476291e-06,
-      "loss": 3.6805,
+      "loss": 1.8176,
       "step": 7170
     },
     {
       "epoch": 0.8892607052116255,
-      "grad_norm": 0.27154514041437766,
+      "grad_norm": 0.13704309120390254,
       "learning_rate": 7.371745476983982e-06,
-      "loss": 3.5953,
+      "loss": 1.7602,
       "step": 7175
     },
     {
       "epoch": 0.8898803990828531,
-      "grad_norm": 0.271667748213158,
+      "grad_norm": 0.13713003743053534,
       "learning_rate": 7.2904414595793556e-06,
-      "loss": 3.5953,
+      "loss": 1.8395,
       "step": 7180
     },
     {
       "epoch": 0.8905000929540807,
-      "grad_norm": 0.25748281840553405,
+      "grad_norm": 0.13231072977298505,
       "learning_rate": 7.209571323765973e-06,
-      "loss": 3.7242,
+      "loss": 1.7906,
       "step": 7185
     },
     {
       "epoch": 0.8911197868253083,
-      "grad_norm": 0.27283419697356404,
+      "grad_norm": 0.1334467482508184,
       "learning_rate": 7.129135448016821e-06,
-      "loss": 3.6867,
+      "loss": 1.8305,
       "step": 7190
     },
     {
       "epoch": 0.8917394806965359,
-      "grad_norm": 0.2747934640984711,
+      "grad_norm": 0.1315592276353881,
       "learning_rate": 7.049134208772545e-06,
-      "loss": 3.6289,
+      "loss": 1.8535,
       "step": 7195
     },
     {
       "epoch": 0.8923591745677635,
-      "grad_norm": 0.2593985093047879,
+      "grad_norm": 0.13605396800853758,
       "learning_rate": 6.969567980439706e-06,
-      "loss": 3.7414,
+      "loss": 1.8613,
       "step": 7200
     },
     {
       "epoch": 0.8929788684389911,
-      "grad_norm": 0.27123165084229295,
+      "grad_norm": 0.13865586242794894,
       "learning_rate": 6.890437135388939e-06,
-      "loss": 3.5883,
+      "loss": 1.8625,
       "step": 7205
     },
     {
       "epoch": 0.8935985623102187,
-      "grad_norm": 0.284080180234527,
+      "grad_norm": 0.13560722459607658,
       "learning_rate": 6.8117420439533615e-06,
-      "loss": 3.6562,
+      "loss": 1.8523,
       "step": 7210
     },
     {
       "epoch": 0.8942182561814463,
-      "grad_norm": 0.26671975217197447,
+      "grad_norm": 0.13599381595733698,
       "learning_rate": 6.733483074426716e-06,
-      "loss": 3.6891,
+      "loss": 1.8551,
       "step": 7215
     },
     {
       "epoch": 0.894837950052674,
-      "grad_norm": 0.2781899517093553,
+      "grad_norm": 0.14264115635666896,
       "learning_rate": 6.655660593061719e-06,
-      "loss": 3.5781,
+      "loss": 1.8402,
       "step": 7220
     },
     {
       "epoch": 0.8954576439239016,
-      "grad_norm": 0.27489213812878466,
+      "grad_norm": 0.13748490683095682,
       "learning_rate": 6.578274964068298e-06,
-      "loss": 3.6289,
+      "loss": 1.8867,
       "step": 7225
     },
     {
       "epoch": 0.8960773377951292,
-      "grad_norm": 0.2711321321359575,
+      "grad_norm": 0.13058170557642237,
       "learning_rate": 6.50132654961193e-06,
-      "loss": 3.7266,
+      "loss": 1.8395,
       "step": 7230
     },
     {
       "epoch": 0.8966970316663568,
-      "grad_norm": 0.28198864613751984,
+      "grad_norm": 0.14230736706254918,
       "learning_rate": 6.424815709811871e-06,
-      "loss": 3.7094,
+      "loss": 1.8887,
       "step": 7235
     },
     {
       "epoch": 0.8973167255375845,
-      "grad_norm": 0.2769375687826022,
+      "grad_norm": 0.1407417474259601,
       "learning_rate": 6.3487428027395715e-06,
-      "loss": 3.6281,
+      "loss": 1.7859,
       "step": 7240
     },
     {
       "epoch": 0.8979364194088121,
-      "grad_norm": 0.2728007227425074,
+      "grad_norm": 0.13411033665330713,
       "learning_rate": 6.273108184416943e-06,
-      "loss": 3.6445,
+      "loss": 1.8422,
       "step": 7245
     },
     {
       "epoch": 0.8985561132800397,
-      "grad_norm": 0.27970919433014235,
+      "grad_norm": 0.13860443760395488,
       "learning_rate": 6.197912208814694e-06,
-      "loss": 3.6945,
+      "loss": 1.8723,
       "step": 7250
     },
     {
       "epoch": 0.8991758071512673,
-      "grad_norm": 0.28984382958053845,
+      "grad_norm": 0.1516944965846428,
       "learning_rate": 6.123155227850708e-06,
-      "loss": 3.6227,
+      "loss": 1.8492,
       "step": 7255
     },
     {
       "epoch": 0.8997955010224948,
-      "grad_norm": 0.26511116487605124,
+      "grad_norm": 0.13130989891564823,
       "learning_rate": 6.048837591388301e-06,
-      "loss": 3.575,
+      "loss": 1.8359,
       "step": 7260
     },
     {
       "epoch": 0.9004151948937225,
-      "grad_norm": 0.2591993698926708,
+      "grad_norm": 0.13496732630623953,
       "learning_rate": 5.974959647234746e-06,
-      "loss": 3.6633,
+      "loss": 1.8262,
       "step": 7265
     },
     {
       "epoch": 0.9010348887649501,
-      "grad_norm": 0.2631505045251582,
+      "grad_norm": 0.1398604330990237,
       "learning_rate": 5.901521741139482e-06,
-      "loss": 3.7563,
+      "loss": 1.8078,
       "step": 7270
     },
     {
       "epoch": 0.9016545826361777,
-      "grad_norm": 0.26728853198677144,
+      "grad_norm": 0.13392388488639656,
       "learning_rate": 5.828524216792586e-06,
-      "loss": 3.6656,
+      "loss": 1.7805,
       "step": 7275
     },
     {
       "epoch": 0.9022742765074053,
-      "grad_norm": 0.27823983832771093,
+      "grad_norm": 0.14015109449246452,
       "learning_rate": 5.75596741582316e-06,
-      "loss": 3.593,
+      "loss": 1.85,
       "step": 7280
     },
     {
       "epoch": 0.902893970378633,
-      "grad_norm": 0.26263872602397725,
+      "grad_norm": 0.1366604111163144,
       "learning_rate": 5.6838516777977135e-06,
-      "loss": 3.7344,
+      "loss": 1.8242,
       "step": 7285
     },
     {
       "epoch": 0.9035136642498606,
-      "grad_norm": 0.2794394137406013,
+      "grad_norm": 0.13873999414384125,
       "learning_rate": 5.6121773402185385e-06,
-      "loss": 3.7031,
+      "loss": 1.784,
       "step": 7290
     },
     {
       "epoch": 0.9041333581210882,
-      "grad_norm": 0.2733651674052135,
+      "grad_norm": 0.12957247714445108,
       "learning_rate": 5.540944738522203e-06,
-      "loss": 3.7102,
+      "loss": 1.8805,
       "step": 7295
     },
     {
       "epoch": 0.9047530519923158,
-      "grad_norm": 0.27487780611573903,
+      "grad_norm": 0.1375066888534788,
       "learning_rate": 5.470154206077949e-06,
-      "loss": 3.5016,
+      "loss": 1.8293,
       "step": 7300
     },
     {
       "epoch": 0.9053727458635434,
-      "grad_norm": 0.27113279352532604,
+      "grad_norm": 0.14306785500804953,
       "learning_rate": 5.3998060741861314e-06,
-      "loss": 3.6992,
+      "loss": 1.9141,
       "step": 7305
     },
     {
       "epoch": 0.9059924397347711,
-      "grad_norm": 0.2727201470546044,
+      "grad_norm": 0.13984693487259178,
       "learning_rate": 5.329900672076637e-06,
-      "loss": 3.6477,
+      "loss": 1.8313,
       "step": 7310
     },
     {
       "epoch": 0.9066121336059987,
-      "grad_norm": 0.27094646523483445,
+      "grad_norm": 0.132439352315789,
       "learning_rate": 5.260438326907413e-06,
-      "loss": 3.6203,
+      "loss": 1.8492,
       "step": 7315
     },
     {
       "epoch": 0.9072318274772263,
-      "grad_norm": 0.2631067653335526,
+      "grad_norm": 0.13944473906087765,
       "learning_rate": 5.191419363762873e-06,
-      "loss": 3.682,
+      "loss": 1.8215,
       "step": 7320
     },
     {
       "epoch": 0.9078515213484538,
-      "grad_norm": 0.28831558778089184,
+      "grad_norm": 0.13074588703804052,
       "learning_rate": 5.122844105652402e-06,
-      "loss": 3.6875,
+      "loss": 1.9125,
       "step": 7325
     },
     {
       "epoch": 0.9084712152196814,
-      "grad_norm": 0.2801500582422861,
+      "grad_norm": 0.14840324829722784,
       "learning_rate": 5.054712873508827e-06,
-      "loss": 3.8047,
+      "loss": 1.8781,
       "step": 7330
     },
     {
       "epoch": 0.9090909090909091,
-      "grad_norm": 0.2723292619898439,
+      "grad_norm": 0.13734906506051153,
       "learning_rate": 4.987025986186966e-06,
-      "loss": 3.7938,
+      "loss": 1.8535,
       "step": 7335
     },
     {
       "epoch": 0.9097106029621367,
-      "grad_norm": 0.27867354050039317,
+      "grad_norm": 0.1390678545194565,
       "learning_rate": 4.919783760462082e-06,
-      "loss": 3.5711,
+      "loss": 1.7563,
       "step": 7340
     },
     {
       "epoch": 0.9103302968333643,
-      "grad_norm": 0.27019084773565777,
+      "grad_norm": 0.1345470887887599,
       "learning_rate": 4.85298651102839e-06,
-      "loss": 3.7297,
+      "loss": 1.875,
       "step": 7345
     },
     {
       "epoch": 0.9109499907045919,
-      "grad_norm": 0.2811769070748566,
+      "grad_norm": 0.14077957222738707,
       "learning_rate": 4.786634550497637e-06,
-      "loss": 3.7406,
+      "loss": 1.8637,
       "step": 7350
     },
     {
       "epoch": 0.9115696845758195,
-      "grad_norm": 0.2737420543087705,
+      "grad_norm": 0.14011299364588545,
       "learning_rate": 4.720728189397628e-06,
-      "loss": 3.6914,
+      "loss": 1.7785,
       "step": 7355
     },
     {
       "epoch": 0.9121893784470472,
-      "grad_norm": 0.2752784244407715,
+      "grad_norm": 0.1407516641368298,
       "learning_rate": 4.655267736170732e-06,
-      "loss": 3.6805,
+      "loss": 1.8277,
       "step": 7360
     },
     {
       "epoch": 0.9128090723182748,
-      "grad_norm": 0.27025007110186383,
+      "grad_norm": 0.14111369095031798,
       "learning_rate": 4.5902534971724806e-06,
-      "loss": 3.8078,
+      "loss": 1.8016,
       "step": 7365
     },
     {
       "epoch": 0.9134287661895024,
-      "grad_norm": 0.27393454394470435,
+      "grad_norm": 0.13620410776159086,
       "learning_rate": 4.525685776670108e-06,
-      "loss": 3.6594,
+      "loss": 1.7734,
       "step": 7370
     },
     {
       "epoch": 0.91404846006073,
-      "grad_norm": 0.2727325638604466,
+      "grad_norm": 0.14478479182293405,
       "learning_rate": 4.46156487684114e-06,
-      "loss": 3.657,
+      "loss": 1.8207,
       "step": 7375
     },
     {
       "epoch": 0.9146681539319577,
-      "grad_norm": 0.27444588062659564,
+      "grad_norm": 0.1404958421396234,
       "learning_rate": 4.397891097771989e-06,
-      "loss": 3.6883,
+      "loss": 1.8289,
       "step": 7380
     },
     {
       "epoch": 0.9152878478031852,
-      "grad_norm": 0.2706709319540534,
+      "grad_norm": 0.13977123482192788,
       "learning_rate": 4.334664737456539e-06,
-      "loss": 3.6016,
+      "loss": 1.8039,
       "step": 7385
     },
     {
       "epoch": 0.9159075416744128,
-      "grad_norm": 0.27981114336986884,
+      "grad_norm": 0.1385005141537386,
       "learning_rate": 4.271886091794719e-06,
-      "loss": 3.6891,
+      "loss": 1.8824,
       "step": 7390
     },
     {
       "epoch": 0.9165272355456404,
-      "grad_norm": 0.28334376858050014,
+      "grad_norm": 0.1338052818221739,
       "learning_rate": 4.209555454591197e-06,
-      "loss": 3.7117,
+      "loss": 1.7785,
       "step": 7395
     },
     {
       "epoch": 0.917146929416868,
-      "grad_norm": 0.28061498567931165,
+      "grad_norm": 0.13855858216006417,
       "learning_rate": 4.147673117553896e-06,
-      "loss": 3.5844,
+      "loss": 1.8516,
       "step": 7400
     },
     {
       "epoch": 0.9177666232880957,
-      "grad_norm": 0.27056248644458036,
+      "grad_norm": 0.13709575865497378,
       "learning_rate": 4.086239370292755e-06,
-      "loss": 3.675,
+      "loss": 1.8473,
       "step": 7405
     },
     {
       "epoch": 0.9183863171593233,
-      "grad_norm": 0.2623174731789405,
+      "grad_norm": 0.14296857282320086,
       "learning_rate": 4.025254500318265e-06,
-      "loss": 3.7977,
+      "loss": 1.8418,
       "step": 7410
     },
     {
       "epoch": 0.9190060110305509,
-      "grad_norm": 0.2668902359387414,
+      "grad_norm": 0.13897241781853462,
       "learning_rate": 3.964718793040178e-06,
-      "loss": 3.7039,
+      "loss": 1.816,
       "step": 7415
     },
     {
       "epoch": 0.9196257049017785,
-      "grad_norm": 0.2682284186560273,
+      "grad_norm": 0.1412624511626614,
       "learning_rate": 3.904632531766195e-06,
-      "loss": 3.6523,
+      "loss": 1.7785,
       "step": 7420
     },
     {
       "epoch": 0.9202453987730062,
-      "grad_norm": 0.2817539903708584,
+      "grad_norm": 0.14569448140912444,
       "learning_rate": 3.84499599770054e-06,
-      "loss": 3.7437,
+      "loss": 1.8477,
       "step": 7425
     },
     {
       "epoch": 0.9208650926442338,
-      "grad_norm": 0.28262725368468417,
+      "grad_norm": 0.13269332412014898,
       "learning_rate": 3.785809469942758e-06,
-      "loss": 3.6078,
+      "loss": 1.8219,
       "step": 7430
     },
     {
       "epoch": 0.9214847865154614,
-      "grad_norm": 0.2872841560550761,
+      "grad_norm": 0.14219480383775535,
       "learning_rate": 3.727073225486344e-06,
-      "loss": 3.5781,
+      "loss": 1.8141,
       "step": 7435
     },
     {
       "epoch": 0.922104480386689,
-      "grad_norm": 0.26171137152403856,
+      "grad_norm": 0.13581718669588042,
       "learning_rate": 3.6687875392174665e-06,
-      "loss": 3.6469,
+      "loss": 1.8223,
       "step": 7440
     },
     {
       "epoch": 0.9227241742579166,
-      "grad_norm": 0.2721412298030815,
+      "grad_norm": 0.1441209126264499,
       "learning_rate": 3.61095268391366e-06,
-      "loss": 3.6672,
+      "loss": 1.832,
       "step": 7445
     },
     {
       "epoch": 0.9233438681291442,
-      "grad_norm": 0.2689383341336418,
+      "grad_norm": 0.14114396455281378,
       "learning_rate": 3.5535689302426236e-06,
-      "loss": 3.6063,
+      "loss": 1.8387,
       "step": 7450
     },
     {
       "epoch": 0.9239635620003718,
-      "grad_norm": 0.27165413446195674,
+      "grad_norm": 0.1387853984081508,
       "learning_rate": 3.496636546760812e-06,
-      "loss": 3.6414,
+      "loss": 1.7766,
       "step": 7455
     },
     {
       "epoch": 0.9245832558715994,
-      "grad_norm": 0.29153732834104146,
+      "grad_norm": 0.1395682007656557,
       "learning_rate": 3.4401557999123146e-06,
-      "loss": 3.5938,
+      "loss": 1.8605,
       "step": 7460
     },
     {
       "epoch": 0.925202949742827,
-      "grad_norm": 0.26713819753828977,
+      "grad_norm": 0.13184826376020614,
       "learning_rate": 3.3841269540275553e-06,
-      "loss": 3.6484,
+      "loss": 1.8656,
       "step": 7465
     },
     {
       "epoch": 0.9258226436140546,
-      "grad_norm": 0.2814855237337874,
+      "grad_norm": 0.13132758063469174,
       "learning_rate": 3.3285502713220617e-06,
-      "loss": 3.6789,
+      "loss": 1.9137,
       "step": 7470
     },
     {
       "epoch": 0.9264423374852823,
-      "grad_norm": 0.2743082315201234,
+      "grad_norm": 0.13838010265846035,
       "learning_rate": 3.2734260118952307e-06,
-      "loss": 3.5234,
+      "loss": 1.8414,
       "step": 7475
     },
     {
       "epoch": 0.9270620313565099,
-      "grad_norm": 0.26820958735170797,
+      "grad_norm": 0.13791135376377014,
       "learning_rate": 3.218754433729065e-06,
-      "loss": 3.6195,
+      "loss": 1.8473,
       "step": 7480
     },
     {
       "epoch": 0.9276817252277375,
-      "grad_norm": 0.28914257172709573,
+      "grad_norm": 0.13341582163949117,
       "learning_rate": 3.1645357926870955e-06,
-      "loss": 3.6086,
+      "loss": 1.8469,
       "step": 7485
     },
     {
       "epoch": 0.9283014190989651,
-      "grad_norm": 0.2756007960053005,
+      "grad_norm": 0.12360153082389551,
       "learning_rate": 3.110770342513036e-06,
-      "loss": 3.6508,
+      "loss": 1.8477,
       "step": 7490
     },
     {
       "epoch": 0.9289211129701928,
-      "grad_norm": 0.27707405209101954,
+      "grad_norm": 0.14160081548498316,
       "learning_rate": 3.057458334829699e-06,
-      "loss": 3.6102,
+      "loss": 1.8289,
       "step": 7495
     },
     {
       "epoch": 0.9295408068414204,
-      "grad_norm": 0.288026781525324,
+      "grad_norm": 0.13906615538183184,
       "learning_rate": 3.0046000191377934e-06,
-      "loss": 3.7766,
+      "loss": 1.8344,
       "step": 7500
     },
     {
       "epoch": 0.930160500712648,
-      "grad_norm": 0.2683433788201112,
+      "grad_norm": 0.13640236310813858,
       "learning_rate": 2.9521956428146923e-06,
-      "loss": 3.7789,
+      "loss": 1.7816,
       "step": 7505
     },
     {
       "epoch": 0.9307801945838756,
-      "grad_norm": 0.29027153699942254,
+      "grad_norm": 0.13041948410128965,
       "learning_rate": 2.9002454511133923e-06,
-      "loss": 3.6172,
+      "loss": 1.8438,
       "step": 7510
     },
     {
       "epoch": 0.9313998884551031,
-      "grad_norm": 0.2667413134023255,
+      "grad_norm": 0.13181252651115552,
       "learning_rate": 2.8487496871612453e-06,
-      "loss": 3.5625,
+      "loss": 1.8445,
       "step": 7515
     },
     {
       "epoch": 0.9320195823263308,
-      "grad_norm": 0.2704211917414998,
+      "grad_norm": 0.13523830165567224,
       "learning_rate": 2.7977085919589254e-06,
-      "loss": 3.6969,
+      "loss": 1.8453,
       "step": 7520
     },
     {
       "epoch": 0.9326392761975584,
-      "grad_norm": 0.2709804732784805,
+      "grad_norm": 0.13314991745111548,
       "learning_rate": 2.7471224043792098e-06,
-      "loss": 3.6844,
+      "loss": 1.8387,
       "step": 7525
     },
     {
       "epoch": 0.933258970068786,
-      "grad_norm": 0.27017634645825134,
+      "grad_norm": 0.13734870109010086,
       "learning_rate": 2.6969913611659457e-06,
-      "loss": 3.6047,
+      "loss": 1.907,
       "step": 7530
     },
     {
       "epoch": 0.9338786639400136,
-      "grad_norm": 0.28977647686803293,
+      "grad_norm": 0.1353394145014779,
       "learning_rate": 2.6473156969328503e-06,
-      "loss": 3.6891,
+      "loss": 1.7922,
       "step": 7535
     },
     {
       "epoch": 0.9344983578112412,
-      "grad_norm": 0.26221324731263035,
+      "grad_norm": 0.13982214073105317,
       "learning_rate": 2.5980956441625236e-06,
-      "loss": 3.7344,
+      "loss": 1.8031,
       "step": 7540
     },
     {
       "epoch": 0.9351180516824689,
-      "grad_norm": 0.27043105541626966,
+      "grad_norm": 0.13754359803353158,
       "learning_rate": 2.5493314332052377e-06,
-      "loss": 3.7297,
+      "loss": 1.8371,
       "step": 7545
     },
     {
       "epoch": 0.9357377455536965,
-      "grad_norm": 0.2645303314795488,
+      "grad_norm": 0.13846264686921053,
       "learning_rate": 2.501023292277971e-06,
-      "loss": 3.6734,
+      "loss": 1.85,
       "step": 7550
     },
     {
       "epoch": 0.9363574394249241,
-      "grad_norm": 0.2872448404216296,
+      "grad_norm": 0.13537888334342718,
       "learning_rate": 2.453171447463265e-06,
-      "loss": 3.5742,
+      "loss": 1.8641,
       "step": 7555
     },
     {
       "epoch": 0.9369771332961517,
-      "grad_norm": 0.28659408753141385,
+      "grad_norm": 0.14209971212182437,
       "learning_rate": 2.4057761227081923e-06,
-      "loss": 3.6891,
+      "loss": 1.7953,
       "step": 7560
     },
     {
       "epoch": 0.9375968271673794,
-      "grad_norm": 0.2878587690908836,
+      "grad_norm": 0.13889335296014865,
       "learning_rate": 2.358837539823311e-06,
-      "loss": 3.7313,
+      "loss": 1.8383,
       "step": 7565
     },
     {
       "epoch": 0.938216521038607,
-      "grad_norm": 0.2653905160082446,
+      "grad_norm": 0.13458314070605037,
       "learning_rate": 2.3123559184816344e-06,
-      "loss": 3.6633,
+      "loss": 1.7836,
       "step": 7570
     },
     {
       "epoch": 0.9388362149098345,
-      "grad_norm": 0.2735739394398778,
+      "grad_norm": 0.13737327467603194,
       "learning_rate": 2.2663314762175647e-06,
-      "loss": 3.8047,
+      "loss": 1.875,
       "step": 7575
     },
     {
       "epoch": 0.9394559087810621,
-      "grad_norm": 0.28433844946547354,
+      "grad_norm": 0.13624035692346814,
       "learning_rate": 2.2207644284259256e-06,
-      "loss": 3.6414,
+      "loss": 1.8383,
       "step": 7580
     },
     {
       "epoch": 0.9400756026522897,
-      "grad_norm": 0.27475684476595646,
+      "grad_norm": 0.13373715255768495,
       "learning_rate": 2.1756549883609313e-06,
-      "loss": 3.7437,
+      "loss": 1.8426,
       "step": 7585
     },
     {
       "epoch": 0.9406952965235174,
-      "grad_norm": 0.2780996306984348,
+      "grad_norm": 0.1328307541501429,
       "learning_rate": 2.131003367135154e-06,
-      "loss": 3.6781,
+      "loss": 1.8473,
       "step": 7590
     },
     {
       "epoch": 0.941314990394745,
-      "grad_norm": 0.2583439651780764,
+      "grad_norm": 0.1365831528128905,
       "learning_rate": 2.086809773718601e-06,
-      "loss": 3.675,
+      "loss": 1.8145,
       "step": 7595
     },
     {
       "epoch": 0.9419346842659726,
-      "grad_norm": 0.27909927819630376,
+      "grad_norm": 0.13862135902790484,
       "learning_rate": 2.0430744149377177e-06,
-      "loss": 3.6602,
+      "loss": 1.8508,
       "step": 7600
     },
     {
       "epoch": 0.9425543781372002,
-      "grad_norm": 0.2840758366639068,
+      "grad_norm": 0.14064173658635798,
       "learning_rate": 1.999797495474365e-06,
-      "loss": 3.768,
+      "loss": 1.8242,
       "step": 7605
     },
     {
       "epoch": 0.9431740720084278,
-      "grad_norm": 0.28264929875319833,
+      "grad_norm": 0.13911440336191405,
       "learning_rate": 1.9569792178649405e-06,
-      "loss": 3.6391,
+      "loss": 1.8699,
       "step": 7610
     },
     {
       "epoch": 0.9437937658796555,
-      "grad_norm": 0.2699490069013214,
+      "grad_norm": 0.13490137157085297,
       "learning_rate": 1.914619782499383e-06,
-      "loss": 3.7398,
+      "loss": 1.891,
       "step": 7615
     },
     {
       "epoch": 0.9444134597508831,
-      "grad_norm": 0.28918121210572356,
+      "grad_norm": 0.14290819661650486,
       "learning_rate": 1.8727193876202143e-06,
-      "loss": 3.5922,
+      "loss": 1.8387,
       "step": 7620
     },
     {
       "epoch": 0.9450331536221107,
-      "grad_norm": 0.26447781129623193,
+      "grad_norm": 0.13470735057128577,
       "learning_rate": 1.8312782293216979e-06,
-      "loss": 3.6164,
+      "loss": 1.834,
       "step": 7625
     },
     {
       "epoch": 0.9456528474933383,
-      "grad_norm": 0.2655529192807083,
+      "grad_norm": 0.13573866020248607,
       "learning_rate": 1.7902965015488381e-06,
-      "loss": 3.7133,
+      "loss": 1.8918,
       "step": 7630
     },
     {
       "epoch": 0.946272541364566,
-      "grad_norm": 0.26970685815626805,
+      "grad_norm": 0.14082077621818678,
       "learning_rate": 1.749774396096482e-06,
-      "loss": 3.7609,
+      "loss": 1.8762,
       "step": 7635
     },
     {
       "epoch": 0.9468922352357935,
-      "grad_norm": 0.27989040137666626,
+      "grad_norm": 0.13993020823079827,
       "learning_rate": 1.709712102608463e-06,
-      "loss": 3.6578,
+      "loss": 1.8129,
       "step": 7640
     },
     {
       "epoch": 0.9475119291070211,
-      "grad_norm": 0.2768738052873664,
+      "grad_norm": 0.13932041357054537,
       "learning_rate": 1.6701098085767031e-06,
-      "loss": 3.5641,
+      "loss": 1.8062,
       "step": 7645
     },
     {
       "epoch": 0.9481316229782487,
-      "grad_norm": 0.26954554955538307,
+      "grad_norm": 0.13250968165271637,
       "learning_rate": 1.630967699340269e-06,
-      "loss": 3.5883,
+      "loss": 1.8555,
       "step": 7650
     },
     {
       "epoch": 0.9487513168494763,
-      "grad_norm": 0.2672662342631292,
+      "grad_norm": 0.1322252771091484,
       "learning_rate": 1.5922859580846271e-06,
-      "loss": 3.7078,
+      "loss": 1.8555,
       "step": 7655
     },
     {
       "epoch": 0.949371010720704,
-      "grad_norm": 0.2703231564176733,
+      "grad_norm": 0.14227157372874508,
       "learning_rate": 1.5540647658406682e-06,
-      "loss": 3.6016,
+      "loss": 1.8102,
       "step": 7660
     },
     {
       "epoch": 0.9499907045919316,
-      "grad_norm": 0.2783617624945699,
+      "grad_norm": 0.14627459669348514,
       "learning_rate": 1.5163043014839284e-06,
-      "loss": 3.5859,
+      "loss": 1.7781,
       "step": 7665
     },
     {
       "epoch": 0.9506103984631592,
-      "grad_norm": 0.27960467541396195,
+      "grad_norm": 0.1365770378516687,
       "learning_rate": 1.479004741733736e-06,
-      "loss": 3.7055,
+      "loss": 1.859,
       "step": 7670
     },
     {
       "epoch": 0.9512300923343868,
-      "grad_norm": 0.27926562477154315,
+      "grad_norm": 0.1351641368471888,
       "learning_rate": 1.4421662611523667e-06,
-      "loss": 3.675,
+      "loss": 1.748,
       "step": 7675
     },
     {
       "epoch": 0.9518497862056144,
-      "grad_norm": 0.2717062431835573,
+      "grad_norm": 0.14013923132334566,
       "learning_rate": 1.4057890321442558e-06,
-      "loss": 3.6898,
+      "loss": 1.8434,
       "step": 7680
     },
     {
       "epoch": 0.9524694800768421,
-      "grad_norm": 0.28087453505860577,
+      "grad_norm": 0.14025363267615307,
       "learning_rate": 1.3698732249551648e-06,
-      "loss": 3.7852,
+      "loss": 1.7656,
       "step": 7685
     },
     {
       "epoch": 0.9530891739480697,
-      "grad_norm": 0.28704266677304857,
+      "grad_norm": 0.13360928888669812,
       "learning_rate": 1.3344190076714059e-06,
-      "loss": 3.675,
+      "loss": 1.9168,
       "step": 7690
     },
     {
       "epoch": 0.9537088678192973,
-      "grad_norm": 0.25837975350935316,
+      "grad_norm": 0.13388017546109943,
       "learning_rate": 1.2994265462190513e-06,
-      "loss": 3.7273,
+      "loss": 1.7984,
       "step": 7695
     },
     {
       "epoch": 0.9543285616905249,
-      "grad_norm": 0.2673603885607993,
+      "grad_norm": 0.1355624389649471,
       "learning_rate": 1.2648960043631474e-06,
-      "loss": 3.7102,
+      "loss": 1.8902,
       "step": 7700
     },
     {
       "epoch": 0.9549482555617524,
-      "grad_norm": 0.27804443443459514,
+      "grad_norm": 0.14142442625050944,
       "learning_rate": 1.230827543706925e-06,
-      "loss": 3.6562,
+      "loss": 1.791,
       "step": 7705
     },
     {
       "epoch": 0.9555679494329801,
-      "grad_norm": 0.26707459118833776,
+      "grad_norm": 0.13651526693745009,
       "learning_rate": 1.1972213236911112e-06,
-      "loss": 3.6453,
+      "loss": 1.8707,
       "step": 7710
     },
     {
       "epoch": 0.9561876433042077,
-      "grad_norm": 0.2724309585648748,
+      "grad_norm": 0.1330434243692908,
       "learning_rate": 1.1640775015931304e-06,
-      "loss": 3.5938,
+      "loss": 1.8598,
       "step": 7715
     },
     {
       "epoch": 0.9568073371754353,
-      "grad_norm": 0.2717087150472206,
+      "grad_norm": 0.13280703904278598,
       "learning_rate": 1.1313962325263717e-06,
-      "loss": 3.5758,
+      "loss": 1.8473,
       "step": 7720
     },
     {
       "epoch": 0.9574270310466629,
-      "grad_norm": 0.27390839096167285,
+      "grad_norm": 0.1371503651919113,
       "learning_rate": 1.0991776694394883e-06,
-      "loss": 3.6859,
+      "loss": 1.8715,
       "step": 7725
     },
     {
       "epoch": 0.9580467249178906,
-      "grad_norm": 0.2664273842532675,
+      "grad_norm": 0.13562632762998766,
       "learning_rate": 1.0674219631156334e-06,
-      "loss": 3.6953,
+      "loss": 1.8465,
       "step": 7730
     },
     {
       "epoch": 0.9586664187891182,
-      "grad_norm": 0.27668769079201644,
+      "grad_norm": 0.1359086084771363,
       "learning_rate": 1.0361292621718145e-06,
-      "loss": 3.7211,
+      "loss": 1.9387,
       "step": 7735
     },
     {
       "epoch": 0.9592861126603458,
-      "grad_norm": 0.29312759026667695,
+      "grad_norm": 0.13701787879508073,
       "learning_rate": 1.005299713058161e-06,
-      "loss": 3.6727,
+      "loss": 1.8426,
       "step": 7740
     },
     {
       "epoch": 0.9599058065315734,
-      "grad_norm": 0.2896106625011773,
+      "grad_norm": 0.13839945104830967,
       "learning_rate": 9.74933460057248e-07,
-      "loss": 3.7094,
+      "loss": 1.8203,
       "step": 7745
     },
     {
       "epoch": 0.960525500402801,
-      "grad_norm": 0.27674410647842534,
+      "grad_norm": 0.13677112829600235,
       "learning_rate": 9.450306452834179e-07,
-      "loss": 3.6437,
+      "loss": 1.8578,
       "step": 7750
     },
     {
       "epoch": 0.9611451942740287,
-      "grad_norm": 0.2911730586333492,
+      "grad_norm": 0.13877686312026696,
       "learning_rate": 9.15591408682126e-07,
-      "loss": 3.5891,
+      "loss": 1.8062,
       "step": 7755
     },
     {
       "epoch": 0.9617648881452563,
-      "grad_norm": 0.2793676329228869,
+      "grad_norm": 0.14210169193275773,
       "learning_rate": 8.866158880292741e-07,
-      "loss": 3.6141,
+      "loss": 1.7824,
       "step": 7760
     },
     {
       "epoch": 0.9623845820164838,
-      "grad_norm": 0.274996874794654,
+      "grad_norm": 0.1379664945215972,
       "learning_rate": 8.581042189305555e-07,
-      "loss": 3.7687,
+      "loss": 1.8281,
       "step": 7765
     },
     {
       "epoch": 0.9630042758877114,
-      "grad_norm": 0.2807429875667943,
+      "grad_norm": 0.13641761386203363,
       "learning_rate": 8.300565348208556e-07,
-      "loss": 3.6562,
+      "loss": 1.8309,
       "step": 7770
     },
     {
       "epoch": 0.963623969758939,
-      "grad_norm": 0.2649603137716782,
+      "grad_norm": 0.13365261844070073,
       "learning_rate": 8.024729669635967e-07,
-      "loss": 3.643,
+      "loss": 1.868,
       "step": 7775
     },
     {
       "epoch": 0.9642436636301667,
-      "grad_norm": 0.2810592469822913,
+      "grad_norm": 0.1378426462575117,
       "learning_rate": 7.753536444501164e-07,
-      "loss": 3.625,
+      "loss": 1.7684,
       "step": 7780
     },
     {
       "epoch": 0.9648633575013943,
-      "grad_norm": 0.27763596036414384,
+      "grad_norm": 0.1379659044326487,
       "learning_rate": 7.486986941991125e-07,
-      "loss": 3.7219,
+      "loss": 1.7703,
       "step": 7785
     },
     {
       "epoch": 0.9654830513726219,
-      "grad_norm": 0.2735361381171989,
+      "grad_norm": 0.13320897754847472,
       "learning_rate": 7.225082409559881e-07,
-      "loss": 3.7336,
+      "loss": 1.8211,
       "step": 7790
     },
     {
       "epoch": 0.9661027452438495,
-      "grad_norm": 0.2860714882582262,
+      "grad_norm": 0.13014133781314957,
       "learning_rate": 6.967824072923068e-07,
-      "loss": 3.6891,
+      "loss": 1.7672,
       "step": 7795
     },
     {
       "epoch": 0.9667224391150772,
-      "grad_norm": 0.25802780361746747,
+      "grad_norm": 0.13168145205742476,
       "learning_rate": 6.715213136052056e-07,
-      "loss": 3.7352,
+      "loss": 1.8258,
       "step": 7800
     },
     {
       "epoch": 0.9673421329863048,
-      "grad_norm": 0.2721644334655904,
+      "grad_norm": 0.1469376174744886,
       "learning_rate": 6.467250781168499e-07,
-      "loss": 3.7266,
+      "loss": 1.8527,
       "step": 7805
     },
     {
       "epoch": 0.9679618268575324,
-      "grad_norm": 0.2616403656619833,
+      "grad_norm": 0.13977784865229265,
       "learning_rate": 6.223938168738341e-07,
-      "loss": 3.6125,
+      "loss": 1.8059,
       "step": 7810
     },
     {
       "epoch": 0.96858152072876,
-      "grad_norm": 0.2837728188131973,
+      "grad_norm": 0.13532853632359126,
       "learning_rate": 5.985276437467046e-07,
-      "loss": 3.7289,
+      "loss": 1.8426,
       "step": 7815
     },
     {
       "epoch": 0.9692012145999876,
-      "grad_norm": 0.2674268745522877,
+      "grad_norm": 0.14519841690310878,
       "learning_rate": 5.751266704293601e-07,
-      "loss": 3.6313,
+      "loss": 1.8184,
       "step": 7820
     },
     {
       "epoch": 0.9698209084712153,
-      "grad_norm": 0.2746401213632551,
+      "grad_norm": 0.13321098148161775,
       "learning_rate": 5.521910064385627e-07,
-      "loss": 3.6617,
+      "loss": 1.8414,
       "step": 7825
     },
     {
       "epoch": 0.9704406023424428,
-      "grad_norm": 0.26258980431856366,
+      "grad_norm": 0.1353961588033009,
       "learning_rate": 5.297207591134612e-07,
-      "loss": 3.693,
+      "loss": 1.7965,
       "step": 7830
     },
     {
       "epoch": 0.9710602962136704,
-      "grad_norm": 0.2736638794617298,
+      "grad_norm": 0.13453437613463365,
       "learning_rate": 5.077160336149911e-07,
-      "loss": 3.6656,
+      "loss": 1.8438,
       "step": 7835
     },
     {
       "epoch": 0.971679990084898,
-      "grad_norm": 0.2693661535971264,
+      "grad_norm": 0.14592528106657068,
       "learning_rate": 4.861769329254862e-07,
-      "loss": 3.6242,
+      "loss": 1.8098,
       "step": 7840
     },
     {
       "epoch": 0.9722996839561256,
-      "grad_norm": 0.2645066499365123,
+      "grad_norm": 0.14121740990989298,
       "learning_rate": 4.651035578481344e-07,
-      "loss": 3.682,
+      "loss": 1.8746,
       "step": 7845
     },
     {
       "epoch": 0.9729193778273533,
-      "grad_norm": 0.27512468136313495,
+      "grad_norm": 0.13805818067090103,
       "learning_rate": 4.4449600700652296e-07,
-      "loss": 3.643,
+      "loss": 1.8492,
       "step": 7850
     },
     {
       "epoch": 0.9735390716985809,
-      "grad_norm": 0.27388444761756614,
+      "grad_norm": 0.14604169034139383,
       "learning_rate": 4.243543768441827e-07,
-      "loss": 3.7188,
+      "loss": 1.8508,
       "step": 7855
     },
     {
       "epoch": 0.9741587655698085,
-      "grad_norm": 0.28910817736216143,
+      "grad_norm": 0.13383491518372784,
       "learning_rate": 4.046787616241221e-07,
-      "loss": 3.6852,
+      "loss": 1.8828,
       "step": 7860
     },
     {
       "epoch": 0.9747784594410361,
-      "grad_norm": 0.2926126161949655,
+      "grad_norm": 0.13449740986423936,
       "learning_rate": 3.8546925342842764e-07,
-      "loss": 3.543,
+      "loss": 1.7391,
       "step": 7865
     },
     {
       "epoch": 0.9753981533122638,
-      "grad_norm": 0.26509568145736156,
+      "grad_norm": 0.13967533122339457,
       "learning_rate": 3.6672594215774183e-07,
-      "loss": 3.5805,
+      "loss": 1.8344,
       "step": 7870
     },
     {
       "epoch": 0.9760178471834914,
-      "grad_norm": 0.27175913423112874,
+      "grad_norm": 0.14033499838206215,
       "learning_rate": 3.484489155309523e-07,
-      "loss": 3.6203,
+      "loss": 1.8375,
       "step": 7875
     },
     {
       "epoch": 0.976637541054719,
-      "grad_norm": 0.27950277725425826,
+      "grad_norm": 0.14348107282049077,
       "learning_rate": 3.3063825908471456e-07,
-      "loss": 3.6234,
+      "loss": 1.8531,
       "step": 7880
     },
     {
       "epoch": 0.9772572349259466,
-      "grad_norm": 0.2849384937701663,
+      "grad_norm": 0.13855132244869428,
       "learning_rate": 3.132940561730524e-07,
-      "loss": 3.6492,
+      "loss": 1.7863,
       "step": 7885
     },
     {
       "epoch": 0.9778769287971742,
-      "grad_norm": 0.27675344143637354,
+      "grad_norm": 0.13157405016785448,
       "learning_rate": 2.9641638796701344e-07,
-      "loss": 3.7031,
+      "loss": 1.8426,
       "step": 7890
     },
     {
       "epoch": 0.9784966226684018,
-      "grad_norm": 0.26836533772485915,
+      "grad_norm": 0.1321435703732912,
       "learning_rate": 2.800053334542363e-07,
-      "loss": 3.6266,
+      "loss": 1.8637,
       "step": 7895
     },
     {
       "epoch": 0.9791163165396294,
-      "grad_norm": 0.2784555897429475,
+      "grad_norm": 0.13158992278105788,
       "learning_rate": 2.6406096943859537e-07,
-      "loss": 3.7789,
+      "loss": 1.8746,
       "step": 7900
     },
     {
       "epoch": 0.979736010410857,
-      "grad_norm": 0.2739776407729332,
+      "grad_norm": 0.13933541006297465,
       "learning_rate": 2.485833705398677e-07,
-      "loss": 3.7125,
+      "loss": 1.8199,
       "step": 7905
     },
     {
       "epoch": 0.9803557042820846,
-      "grad_norm": 0.2801988570193097,
+      "grad_norm": 0.13437527753897568,
       "learning_rate": 2.3357260919336654e-07,
-      "loss": 3.6797,
+      "loss": 1.8555,
       "step": 7910
     },
     {
       "epoch": 0.9809753981533123,
-      "grad_norm": 0.2817935004020373,
+      "grad_norm": 0.14403000978113806,
       "learning_rate": 2.1902875564958624e-07,
-      "loss": 3.6828,
+      "loss": 1.8395,
       "step": 7915
     },
     {
       "epoch": 0.9815950920245399,
-      "grad_norm": 0.26306490781018277,
+      "grad_norm": 0.13600747653132192,
       "learning_rate": 2.0495187797390236e-07,
-      "loss": 3.7039,
+      "loss": 1.868,
       "step": 7920
     },
     {
       "epoch": 0.9822147858957675,
-      "grad_norm": 0.2666398441401795,
+      "grad_norm": 0.1471070831087313,
       "learning_rate": 1.913420420462164e-07,
-      "loss": 3.668,
+      "loss": 1.818,
       "step": 7925
     },
     {
       "epoch": 0.9828344797669951,
-      "grad_norm": 0.2822222017860193,
+      "grad_norm": 0.1335371161396645,
       "learning_rate": 1.7819931156071168e-07,
-      "loss": 3.6266,
+      "loss": 1.8746,
       "step": 7930
     },
     {
       "epoch": 0.9834541736382227,
-      "grad_norm": 0.27776801734518003,
+      "grad_norm": 0.1368536776343064,
       "learning_rate": 1.6552374802546454e-07,
-      "loss": 3.6852,
+      "loss": 1.8391,
       "step": 7935
     },
     {
       "epoch": 0.9840738675094504,
-      "grad_norm": 0.26060817961696686,
+      "grad_norm": 0.1330749589293093,
       "learning_rate": 1.5331541076225585e-07,
-      "loss": 3.7227,
+      "loss": 1.875,
       "step": 7940
     },
     {
       "epoch": 0.984693561380678,
-      "grad_norm": 0.2675642716698818,
+      "grad_norm": 0.13468183805269288,
       "learning_rate": 1.4157435690619337e-07,
-      "loss": 3.6625,
+      "loss": 1.7766,
       "step": 7945
     },
     {
       "epoch": 0.9853132552519056,
-      "grad_norm": 0.27291219918221404,
+      "grad_norm": 0.13485415942676057,
       "learning_rate": 1.30300641405523e-07,
-      "loss": 3.6742,
+      "loss": 1.857,
       "step": 7950
     },
     {
       "epoch": 0.9859329491231331,
-      "grad_norm": 0.26586541429232524,
+      "grad_norm": 0.13908303443528403,
       "learning_rate": 1.194943170213403e-07,
-      "loss": 3.5688,
+      "loss": 1.8359,
       "step": 7955
     },
     {
       "epoch": 0.9865526429943607,
-      "grad_norm": 0.268823372681042,
+      "grad_norm": 0.13274714646219282,
       "learning_rate": 1.0915543432733488e-07,
-      "loss": 3.7336,
+      "loss": 1.8434,
       "step": 7960
     },
     {
       "epoch": 0.9871723368655884,
-      "grad_norm": 0.2681016890128257,
+      "grad_norm": 0.13810326013237345,
       "learning_rate": 9.928404170959082e-08,
-      "loss": 3.5867,
+      "loss": 1.8164,
       "step": 7965
     },
     {
       "epoch": 0.987792030736816,
-      "grad_norm": 0.26956499858488664,
+      "grad_norm": 0.13614189789038067,
       "learning_rate": 8.988018536630893e-08,
-      "loss": 3.725,
+      "loss": 1.8387,
       "step": 7970
     },
     {
       "epoch": 0.9884117246080436,
-      "grad_norm": 0.28229057560627346,
+      "grad_norm": 0.13375315569066618,
       "learning_rate": 8.094390930762919e-08,
-      "loss": 3.7289,
+      "loss": 1.8,
       "step": 7975
     },
     {
       "epoch": 0.9890314184792712,
-      "grad_norm": 0.27614610199741324,
+      "grad_norm": 0.13158730991646866,
       "learning_rate": 7.247525535538647e-08,
-      "loss": 3.6398,
+      "loss": 1.7934,
       "step": 7980
     },
     {
       "epoch": 0.9896511123504989,
-      "grad_norm": 0.2803128283949079,
+      "grad_norm": 0.14477857824803034,
       "learning_rate": 6.447426314297734e-08,
-      "loss": 3.7266,
+      "loss": 1.8332,
       "step": 7985
     },
     {
       "epoch": 0.9902708062217265,
-      "grad_norm": 0.2618509943839932,
+      "grad_norm": 0.1342267918315416,
       "learning_rate": 5.6940970115115785e-08,
-      "loss": 3.6914,
+      "loss": 1.8547,
       "step": 7990
     },
     {
       "epoch": 0.9908905000929541,
-      "grad_norm": 0.2726310143354448,
+      "grad_norm": 0.13785984492136893,
       "learning_rate": 4.9875411527677826e-08,
-      "loss": 3.8883,
+      "loss": 1.7652,
       "step": 7995
     },
     {
       "epoch": 0.9915101939641817,
-      "grad_norm": 0.2766308585212035,
+      "grad_norm": 0.13807203287629677,
       "learning_rate": 4.327762044755712e-08,
-      "loss": 3.6102,
+      "loss": 1.8668,
       "step": 8000
     },
     {
       "epoch": 0.9921298878354093,
-      "grad_norm": 0.2662764431301861,
+      "grad_norm": 0.13646195929256758,
       "learning_rate": 3.714762775245406e-08,
-      "loss": 3.6828,
+      "loss": 1.7598,
       "step": 8005
     },
     {
       "epoch": 0.992749581706637,
-      "grad_norm": 0.2698434445721707,
+      "grad_norm": 0.13971148375976059,
       "learning_rate": 3.148546213080916e-08,
-      "loss": 3.6898,
+      "loss": 1.8727,
       "step": 8010
     },
     {
       "epoch": 0.9933692755778646,
-      "grad_norm": 0.29310884330046,
+      "grad_norm": 0.13334054828935307,
       "learning_rate": 2.6291150081603212e-08,
-      "loss": 3.5945,
+      "loss": 1.8863,
       "step": 8015
     },
     {
       "epoch": 0.9939889694490921,
-      "grad_norm": 0.2770822794995166,
+      "grad_norm": 0.14352383701066074,
       "learning_rate": 2.156471591426845e-08,
-      "loss": 3.718,
+      "loss": 1.8934,
       "step": 8020
     },
     {
       "epoch": 0.9946086633203197,
-      "grad_norm": 0.27590920995954854,
+      "grad_norm": 0.13909857608391912,
       "learning_rate": 1.7306181748566463e-08,
-      "loss": 3.7328,
+      "loss": 1.8133,
       "step": 8025
     },
     {
       "epoch": 0.9952283571915473,
-      "grad_norm": 0.2739410257909428,
+      "grad_norm": 0.13480520184307876,
       "learning_rate": 1.351556751445493e-08,
-      "loss": 3.7203,
+      "loss": 1.8406,
       "step": 8030
     },
     {
       "epoch": 0.995848051062775,
-      "grad_norm": 0.26566933154278977,
+      "grad_norm": 0.14279753608828183,
       "learning_rate": 1.0192890952054334e-08,
-      "loss": 3.5352,
+      "loss": 1.8004,
       "step": 8035
     },
     {
       "epoch": 0.9964677449340026,
-      "grad_norm": 0.2684622940211439,
+      "grad_norm": 0.14041065956395132,
       "learning_rate": 7.338167611536939e-09,
-      "loss": 3.7133,
+      "loss": 1.7984,
       "step": 8040
     },
     {
       "epoch": 0.9970874388052302,
-      "grad_norm": 0.27691748975672703,
+      "grad_norm": 0.1353003051841239,
       "learning_rate": 4.9514108530157585e-09,
-      "loss": 3.575,
+      "loss": 1.7949,
       "step": 8045
     },
     {
       "epoch": 0.9977071326764578,
-      "grad_norm": 0.27023267310378274,
+      "grad_norm": 0.14086128743096502,
       "learning_rate": 3.0326318465334625e-09,
-      "loss": 3.65,
+      "loss": 1.8348,
       "step": 8050
     },
     {
       "epoch": 0.9983268265476855,
-      "grad_norm": 0.27647933459013596,
+      "grad_norm": 0.1353261527492863,
       "learning_rate": 1.5818395720068602e-09,
-      "loss": 3.6539,
+      "loss": 1.757,
       "step": 8055
     },
     {
       "epoch": 0.9989465204189131,
-      "grad_norm": 0.273299155688557,
+      "grad_norm": 0.13528022729742975,
       "learning_rate": 5.99040819149188e-10,
-      "loss": 3.5516,
+      "loss": 1.8582,
       "step": 8060
     },
     {
       "epoch": 0.9995662142901407,
-      "grad_norm": 0.2738691124969836,
+      "grad_norm": 0.13580379005193177,
       "learning_rate": 8.42401874701082e-11,
-      "loss": 3.5602,
+      "loss": 1.8289,
       "step": 8065
     },
+    {
+      "epoch": 0.9999380306128772,
+      "eval_loss": 1.7869144678115845,
+      "eval_runtime": 146.7975,
+      "eval_samples_per_second": 97.27,
+      "eval_steps_per_second": 6.083,
+      "step": 8068
+    },
     {
       "epoch": 0.9999380306128772,
       "step": 8068,
-      "total_flos": 4248659495485440.0,
-      "train_loss": 0.0,
-      "train_runtime": 0.0126,
-      "train_samples_per_second": 10236188.723,
-      "train_steps_per_second": 639727.105
+      "total_flos": 4248917998829568.0,
+      "train_loss": 1.8640377591255577,
+      "train_runtime": 7899.8856,
+      "train_samples_per_second": 16.341,
+      "train_steps_per_second": 1.021
     }
   ],
   "logging_steps": 5,
@@ -11333,7 +11341,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 4248659495485440.0,
+  "total_flos": 4248917998829568.0,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null