diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11348 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999380306128772,
+  "eval_steps": 500,
+  "global_step": 8068,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001239387742455227,
+      "grad_norm": 0.3613625061479819,
+      "learning_rate": 2.4783147459727386e-07,
+      "loss": 2.5352,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006196938712276135,
+      "grad_norm": 0.36095038255815987,
+      "learning_rate": 1.2391573729863693e-06,
+      "loss": 2.3916,
+      "step": 5
+    },
+    {
+      "epoch": 0.001239387742455227,
+      "grad_norm": 0.3254584998211072,
+      "learning_rate": 2.4783147459727386e-06,
+      "loss": 2.4891,
+      "step": 10
+    },
+    {
+      "epoch": 0.0018590816136828407,
+      "grad_norm": 0.38126326552507134,
+      "learning_rate": 3.717472118959108e-06,
+      "loss": 2.4898,
+      "step": 15
+    },
+    {
+      "epoch": 0.002478775484910454,
+      "grad_norm": 0.441689451221483,
+      "learning_rate": 4.956629491945477e-06,
+      "loss": 2.4367,
+      "step": 20
+    },
+    {
+      "epoch": 0.003098469356138068,
+      "grad_norm": 0.2973969565575579,
+      "learning_rate": 6.195786864931847e-06,
+      "loss": 2.4406,
+      "step": 25
+    },
+    {
+      "epoch": 0.0037181632273656814,
+      "grad_norm": 0.40935208612020174,
+      "learning_rate": 7.434944237918216e-06,
+      "loss": 2.382,
+      "step": 30
+    },
+    {
+      "epoch": 0.004337857098593295,
+      "grad_norm": 0.4181134678122744,
+      "learning_rate": 8.674101610904585e-06,
+      "loss": 2.4586,
+      "step": 35
+    },
+    {
+      "epoch": 0.004957550969820908,
+      "grad_norm": 0.3374239900963284,
+      "learning_rate": 9.913258983890955e-06,
+      "loss": 2.3852,
+      "step": 40
+    },
+    {
+      "epoch": 0.005577244841048522,
+      "grad_norm": 0.4319325262792448,
+      "learning_rate": 1.1152416356877324e-05,
+      "loss": 2.4039,
+      "step": 45
+    },
+    {
+      "epoch": 0.006196938712276136,
+      "grad_norm": 0.32665555964629,
+      "learning_rate": 1.2391573729863694e-05,
+      "loss": 2.4352,
+      "step": 50
+    },
+    {
+      "epoch": 0.006816632583503749,
+      "grad_norm": 0.5172119594644248,
+      "learning_rate": 1.3630731102850064e-05,
+      "loss": 2.3609,
+      "step": 55
+    },
+    {
+      "epoch": 0.007436326454731363,
+      "grad_norm": 0.4087931864681625,
+      "learning_rate": 1.4869888475836432e-05,
+      "loss": 2.4031,
+      "step": 60
+    },
+    {
+      "epoch": 0.008056020325958976,
+      "grad_norm": 0.46592670985648305,
+      "learning_rate": 1.61090458488228e-05,
+      "loss": 2.3156,
+      "step": 65
+    },
+    {
+      "epoch": 0.00867571419718659,
+      "grad_norm": 0.49541312013878036,
+      "learning_rate": 1.734820322180917e-05,
+      "loss": 2.3438,
+      "step": 70
+    },
+    {
+      "epoch": 0.009295408068414203,
+      "grad_norm": 0.38324458134659845,
+      "learning_rate": 1.858736059479554e-05,
+      "loss": 2.3406,
+      "step": 75
+    },
+    {
+      "epoch": 0.009915101939641817,
+      "grad_norm": 0.43338092650638166,
+      "learning_rate": 1.982651796778191e-05,
+      "loss": 2.3375,
+      "step": 80
+    },
+    {
+      "epoch": 0.01053479581086943,
+      "grad_norm": 0.40416289269468914,
+      "learning_rate": 2.106567534076828e-05,
+      "loss": 2.3328,
+      "step": 85
+    },
+    {
+      "epoch": 0.011154489682097044,
+      "grad_norm": 0.4473494205404002,
+      "learning_rate": 2.230483271375465e-05,
+      "loss": 2.2656,
+      "step": 90
+    },
+    {
+      "epoch": 0.011774183553324657,
+      "grad_norm": 0.3813410762384386,
+      "learning_rate": 2.3543990086741015e-05,
+      "loss": 2.2234,
+      "step": 95
+    },
+    {
+      "epoch": 0.012393877424552271,
+      "grad_norm": 0.3290497483814934,
+      "learning_rate": 2.4783147459727388e-05,
+      "loss": 2.1754,
+      "step": 100
+    },
+    {
+      "epoch": 0.013013571295779885,
+      "grad_norm": 0.3774917329867058,
+      "learning_rate": 2.6022304832713758e-05,
+      "loss": 2.0992,
+      "step": 105
+    },
+    {
+      "epoch": 0.013633265167007498,
+      "grad_norm": 0.24634234770137786,
+      "learning_rate": 2.7261462205700128e-05,
+      "loss": 2.1066,
+      "step": 110
+    },
+    {
+      "epoch": 0.014252959038235112,
+      "grad_norm": 0.2541916772452527,
+      "learning_rate": 2.8500619578686494e-05,
+      "loss": 2.1039,
+      "step": 115
+    },
+    {
+      "epoch": 0.014872652909462726,
+      "grad_norm": 0.21483622246957815,
+      "learning_rate": 2.9739776951672864e-05,
+      "loss": 2.1547,
+      "step": 120
+    },
+    {
+      "epoch": 0.01549234678069034,
+      "grad_norm": 0.1832960445686121,
+      "learning_rate": 3.0978934324659233e-05,
+      "loss": 2.1211,
+      "step": 125
+    },
+    {
+      "epoch": 0.01611204065191795,
+      "grad_norm": 0.2024480547670622,
+      "learning_rate": 3.22180916976456e-05,
+      "loss": 2.1246,
+      "step": 130
+    },
+    {
+      "epoch": 0.016731734523145567,
+      "grad_norm": 0.1587252934099839,
+      "learning_rate": 3.345724907063197e-05,
+      "loss": 2.1402,
+      "step": 135
+    },
+    {
+      "epoch": 0.01735142839437318,
+      "grad_norm": 0.1573425276705792,
+      "learning_rate": 3.469640644361834e-05,
+      "loss": 2.1453,
+      "step": 140
+    },
+    {
+      "epoch": 0.017971122265600794,
+      "grad_norm": 0.1649139918619751,
+      "learning_rate": 3.593556381660471e-05,
+      "loss": 2.091,
+      "step": 145
+    },
+    {
+      "epoch": 0.018590816136828406,
+      "grad_norm": 0.16906893865479186,
+      "learning_rate": 3.717472118959108e-05,
+      "loss": 2.0336,
+      "step": 150
+    },
+    {
+      "epoch": 0.01921051000805602,
+      "grad_norm": 0.16828989999123184,
+      "learning_rate": 3.841387856257745e-05,
+      "loss": 2.1066,
+      "step": 155
+    },
+    {
+      "epoch": 0.019830203879283633,
+      "grad_norm": 0.15086389397017358,
+      "learning_rate": 3.965303593556382e-05,
+      "loss": 2.0637,
+      "step": 160
+    },
+    {
+      "epoch": 0.02044989775051125,
+      "grad_norm": 0.13741000516222265,
+      "learning_rate": 4.0892193308550185e-05,
+      "loss": 2.0602,
+      "step": 165
+    },
+    {
+      "epoch": 0.02106959162173886,
+      "grad_norm": 0.13109937575764552,
+      "learning_rate": 4.213135068153656e-05,
+      "loss": 2.0883,
+      "step": 170
+    },
+    {
+      "epoch": 0.021689285492966476,
+      "grad_norm": 0.13563532496297764,
+      "learning_rate": 4.337050805452293e-05,
+      "loss": 2.0859,
+      "step": 175
+    },
+    {
+      "epoch": 0.022308979364194088,
+      "grad_norm": 0.13613940691719567,
+      "learning_rate": 4.46096654275093e-05,
+      "loss": 2.05,
+      "step": 180
+    },
+    {
+      "epoch": 0.022928673235421703,
+      "grad_norm": 0.12247823742232716,
+      "learning_rate": 4.5848822800495664e-05,
+      "loss": 2.041,
+      "step": 185
+    },
+    {
+      "epoch": 0.023548367106649315,
+      "grad_norm": 0.11943084710129132,
+      "learning_rate": 4.708798017348203e-05,
+      "loss": 1.9684,
+      "step": 190
+    },
+    {
+      "epoch": 0.02416806097787693,
+      "grad_norm": 0.10996794203028536,
+      "learning_rate": 4.83271375464684e-05,
+      "loss": 2.0195,
+      "step": 195
+    },
+    {
+      "epoch": 0.024787754849104542,
+      "grad_norm": 0.11919593904845416,
+      "learning_rate": 4.9566294919454776e-05,
+      "loss": 2.0355,
+      "step": 200
+    },
+    {
+      "epoch": 0.025407448720332158,
+      "grad_norm": 0.13573935802440174,
+      "learning_rate": 5.080545229244115e-05,
+      "loss": 2.0465,
+      "step": 205
+    },
+    {
+      "epoch": 0.02602714259155977,
+      "grad_norm": 0.11810880112190057,
+      "learning_rate": 5.2044609665427516e-05,
+      "loss": 2.0449,
+      "step": 210
+    },
+    {
+      "epoch": 0.02664683646278738,
+      "grad_norm": 0.10714263949710778,
+      "learning_rate": 5.328376703841388e-05,
+      "loss": 2.0453,
+      "step": 215
+    },
+    {
+      "epoch": 0.027266530334014997,
+      "grad_norm": 0.10808893820532955,
+      "learning_rate": 5.4522924411400255e-05,
+      "loss": 2.0484,
+      "step": 220
+    },
+    {
+      "epoch": 0.02788622420524261,
+      "grad_norm": 0.1079631127104769,
+      "learning_rate": 5.576208178438662e-05,
+      "loss": 2.0219,
+      "step": 225
+    },
+    {
+      "epoch": 0.028505918076470224,
+      "grad_norm": 0.115215707179092,
+      "learning_rate": 5.700123915737299e-05,
+      "loss": 2.0219,
+      "step": 230
+    },
+    {
+      "epoch": 0.029125611947697836,
+      "grad_norm": 0.10429743509556648,
+      "learning_rate": 5.8240396530359354e-05,
+      "loss": 2.0191,
+      "step": 235
+    },
+    {
+      "epoch": 0.02974530581892545,
+      "grad_norm": 0.10512419844589924,
+      "learning_rate": 5.947955390334573e-05,
+      "loss": 1.9578,
+      "step": 240
+    },
+    {
+      "epoch": 0.030364999690153063,
+      "grad_norm": 0.10841051120377791,
+      "learning_rate": 6.0718711276332094e-05,
+      "loss": 2.023,
+      "step": 245
+    },
+    {
+      "epoch": 0.03098469356138068,
+      "grad_norm": 0.1123866377946053,
+      "learning_rate": 6.195786864931847e-05,
+      "loss": 2.0043,
+      "step": 250
+    },
+    {
+      "epoch": 0.031604387432608294,
+      "grad_norm": 0.1071143230575679,
+      "learning_rate": 6.319702602230483e-05,
+      "loss": 2.0059,
+      "step": 255
+    },
+    {
+      "epoch": 0.0322240813038359,
+      "grad_norm": 0.11205273394431785,
+      "learning_rate": 6.44361833952912e-05,
+      "loss": 1.9516,
+      "step": 260
+    },
+    {
+      "epoch": 0.03284377517506352,
+      "grad_norm": 0.11124243248908515,
+      "learning_rate": 6.567534076827757e-05,
+      "loss": 2.0324,
+      "step": 265
+    },
+    {
+      "epoch": 0.03346346904629113,
+      "grad_norm": 0.11056302895417636,
+      "learning_rate": 6.691449814126395e-05,
+      "loss": 2.0059,
+      "step": 270
+    },
+    {
+      "epoch": 0.03408316291751875,
+      "grad_norm": 0.1150685621638358,
+      "learning_rate": 6.815365551425031e-05,
+      "loss": 1.8949,
+      "step": 275
+    },
+    {
+      "epoch": 0.03470285678874636,
+      "grad_norm": 0.11053590058509483,
+      "learning_rate": 6.939281288723668e-05,
+      "loss": 1.9539,
+      "step": 280
+    },
+    {
+      "epoch": 0.03532255065997397,
+      "grad_norm": 0.11097452419745213,
+      "learning_rate": 7.063197026022306e-05,
+      "loss": 2.0055,
+      "step": 285
+    },
+    {
+      "epoch": 0.03594224453120159,
+      "grad_norm": 0.10981036645663586,
+      "learning_rate": 7.187112763320942e-05,
+      "loss": 1.9684,
+      "step": 290
+    },
+    {
+      "epoch": 0.0365619384024292,
+      "grad_norm": 0.10851974717472565,
+      "learning_rate": 7.311028500619579e-05,
+      "loss": 1.9937,
+      "step": 295
+    },
+    {
+      "epoch": 0.03718163227365681,
+      "grad_norm": 0.1097063039665451,
+      "learning_rate": 7.434944237918216e-05,
+      "loss": 1.9258,
+      "step": 300
+    },
+    {
+      "epoch": 0.03780132614488443,
+      "grad_norm": 0.11108822223408563,
+      "learning_rate": 7.558859975216854e-05,
+      "loss": 1.9703,
+      "step": 305
+    },
+    {
+      "epoch": 0.03842102001611204,
+      "grad_norm": 0.10760540355469742,
+      "learning_rate": 7.68277571251549e-05,
+      "loss": 1.9465,
+      "step": 310
+    },
+    {
+      "epoch": 0.03904071388733965,
+      "grad_norm": 0.11516090750067655,
+      "learning_rate": 7.806691449814127e-05,
+      "loss": 1.952,
+      "step": 315
+    },
+    {
+      "epoch": 0.039660407758567266,
+      "grad_norm": 0.1043408406511001,
+      "learning_rate": 7.930607187112764e-05,
+      "loss": 1.9531,
+      "step": 320
+    },
+    {
+      "epoch": 0.04028010162979488,
+      "grad_norm": 0.1039519846752159,
+      "learning_rate": 8.0545229244114e-05,
+      "loss": 1.9426,
+      "step": 325
+    },
+    {
+      "epoch": 0.0408997955010225,
+      "grad_norm": 0.10746843855868768,
+      "learning_rate": 8.178438661710037e-05,
+      "loss": 1.9855,
+      "step": 330
+    },
+    {
+      "epoch": 0.041519489372250105,
+      "grad_norm": 0.10777591543631691,
+      "learning_rate": 8.302354399008675e-05,
+      "loss": 1.9445,
+      "step": 335
+    },
+    {
+      "epoch": 0.04213918324347772,
+      "grad_norm": 0.10806752734339195,
+      "learning_rate": 8.426270136307312e-05,
+      "loss": 1.8906,
+      "step": 340
+    },
+    {
+      "epoch": 0.042758877114705336,
+      "grad_norm": 0.11502306811099913,
+      "learning_rate": 8.550185873605948e-05,
+      "loss": 1.9727,
+      "step": 345
+    },
+    {
+      "epoch": 0.04337857098593295,
+      "grad_norm": 0.11448974921102456,
+      "learning_rate": 8.674101610904586e-05,
+      "loss": 1.8648,
+      "step": 350
+    },
+    {
+      "epoch": 0.04399826485716056,
+      "grad_norm": 0.1078095422643582,
+      "learning_rate": 8.798017348203223e-05,
+      "loss": 1.9875,
+      "step": 355
+    },
+    {
+      "epoch": 0.044617958728388175,
+      "grad_norm": 0.11060380654587663,
+      "learning_rate": 8.92193308550186e-05,
+      "loss": 2.0211,
+      "step": 360
+    },
+    {
+      "epoch": 0.04523765259961579,
+      "grad_norm": 0.11613817321861132,
+      "learning_rate": 9.045848822800496e-05,
+      "loss": 1.9891,
+      "step": 365
+    },
+    {
+      "epoch": 0.045857346470843406,
+      "grad_norm": 0.11368583100057289,
+      "learning_rate": 9.169764560099133e-05,
+      "loss": 1.9773,
+      "step": 370
+    },
+    {
+      "epoch": 0.046477040342071015,
+      "grad_norm": 0.10894141405623521,
+      "learning_rate": 9.29368029739777e-05,
+      "loss": 1.9566,
+      "step": 375
+    },
+    {
+      "epoch": 0.04709673421329863,
+      "grad_norm": 0.11421341665297284,
+      "learning_rate": 9.417596034696406e-05,
+      "loss": 1.9379,
+      "step": 380
+    },
+    {
+      "epoch": 0.047716428084526245,
+      "grad_norm": 0.11239189457188502,
+      "learning_rate": 9.541511771995044e-05,
+      "loss": 1.9359,
+      "step": 385
+    },
+    {
+      "epoch": 0.04833612195575386,
+      "grad_norm": 0.10853175700299506,
+      "learning_rate": 9.66542750929368e-05,
+      "loss": 1.9539,
+      "step": 390
+    },
+    {
+      "epoch": 0.04895581582698147,
+      "grad_norm": 0.11310381389332551,
+      "learning_rate": 9.789343246592317e-05,
+      "loss": 1.9348,
+      "step": 395
+    },
+    {
+      "epoch": 0.049575509698209085,
+      "grad_norm": 0.11560816087723923,
+      "learning_rate": 9.913258983890955e-05,
+      "loss": 1.9453,
+      "step": 400
+    },
+    {
+      "epoch": 0.0501952035694367,
+      "grad_norm": 0.12016209733805872,
+      "learning_rate": 0.0001003717472118959,
+      "loss": 1.9293,
+      "step": 405
+    },
+    {
+      "epoch": 0.050814897440664315,
+      "grad_norm": 0.11841169962945568,
+      "learning_rate": 0.0001016109045848823,
+      "loss": 1.9348,
+      "step": 410
+    },
+    {
+      "epoch": 0.051434591311891924,
+      "grad_norm": 0.11168273067829344,
+      "learning_rate": 0.00010285006195786867,
+      "loss": 1.932,
+      "step": 415
+    },
+    {
+      "epoch": 0.05205428518311954,
+      "grad_norm": 0.10983446111870222,
+      "learning_rate": 0.00010408921933085503,
+      "loss": 1.9398,
+      "step": 420
+    },
+    {
+      "epoch": 0.052673979054347154,
+      "grad_norm": 0.11246006030379667,
+      "learning_rate": 0.0001053283767038414,
+      "loss": 1.8598,
+      "step": 425
+    },
+    {
+      "epoch": 0.05329367292557476,
+      "grad_norm": 0.11621584827465425,
+      "learning_rate": 0.00010656753407682776,
+      "loss": 1.9211,
+      "step": 430
+    },
+    {
+      "epoch": 0.05391336679680238,
+      "grad_norm": 0.12034248030098742,
+      "learning_rate": 0.00010780669144981412,
+      "loss": 1.966,
+      "step": 435
+    },
+    {
+      "epoch": 0.054533060668029994,
+      "grad_norm": 0.1160099754241247,
+      "learning_rate": 0.00010904584882280051,
+      "loss": 1.9609,
+      "step": 440
+    },
+    {
+      "epoch": 0.05515275453925761,
+      "grad_norm": 0.11125026431874137,
+      "learning_rate": 0.00011028500619578688,
+      "loss": 1.9117,
+      "step": 445
+    },
+    {
+      "epoch": 0.05577244841048522,
+      "grad_norm": 0.11612202738414985,
+      "learning_rate": 0.00011152416356877324,
+      "loss": 1.9352,
+      "step": 450
+    },
+    {
+      "epoch": 0.05639214228171283,
+      "grad_norm": 0.11452837316949507,
+      "learning_rate": 0.00011276332094175961,
+      "loss": 1.9422,
+      "step": 455
+    },
+    {
+      "epoch": 0.05701183615294045,
+      "grad_norm": 0.12866517250152734,
+      "learning_rate": 0.00011400247831474598,
+      "loss": 1.8824,
+      "step": 460
+    },
+    {
+      "epoch": 0.057631530024168064,
+      "grad_norm": 0.11047960782931952,
+      "learning_rate": 0.00011524163568773234,
+      "loss": 1.9336,
+      "step": 465
+    },
+    {
+      "epoch": 0.05825122389539567,
+      "grad_norm": 0.12040909536957894,
+      "learning_rate": 0.00011648079306071871,
+      "loss": 1.9598,
+      "step": 470
+    },
+    {
+      "epoch": 0.05887091776662329,
+      "grad_norm": 0.11354827146813799,
+      "learning_rate": 0.00011771995043370509,
+      "loss": 1.9406,
+      "step": 475
+    },
+    {
+      "epoch": 0.0594906116378509,
+      "grad_norm": 0.11018041152409878,
+      "learning_rate": 0.00011895910780669145,
+      "loss": 1.909,
+      "step": 480
+    },
+    {
+      "epoch": 0.06011030550907852,
+      "grad_norm": 0.11350276690749768,
+      "learning_rate": 0.00012019826517967782,
+      "loss": 1.9453,
+      "step": 485
+    },
+    {
+      "epoch": 0.06072999938030613,
+      "grad_norm": 0.12128379272584458,
+      "learning_rate": 0.00012143742255266419,
+      "loss": 1.9266,
+      "step": 490
+    },
+    {
+      "epoch": 0.06134969325153374,
+      "grad_norm": 0.11759828192541434,
+      "learning_rate": 0.00012267657992565055,
+      "loss": 1.8941,
+      "step": 495
+    },
+    {
+      "epoch": 0.06196938712276136,
+      "grad_norm": 0.12384786337021533,
+      "learning_rate": 0.00012391573729863693,
+      "loss": 1.8883,
+      "step": 500
+    },
+    {
+      "epoch": 0.06258908099398897,
+      "grad_norm": 0.11752293374203299,
+      "learning_rate": 0.00012515489467162331,
+      "loss": 1.8762,
+      "step": 505
+    },
+    {
+      "epoch": 0.06320877486521659,
+      "grad_norm": 0.11824701347044138,
+      "learning_rate": 0.00012639405204460967,
+      "loss": 1.8578,
+      "step": 510
+    },
+    {
+      "epoch": 0.0638284687364442,
+      "grad_norm": 0.11789835565142269,
+      "learning_rate": 0.00012763320941759605,
+      "loss": 1.9703,
+      "step": 515
+    },
+    {
+      "epoch": 0.0644481626076718,
+      "grad_norm": 0.11561056451604006,
+      "learning_rate": 0.0001288723667905824,
+      "loss": 1.9176,
+      "step": 520
+    },
+    {
+      "epoch": 0.06506785647889943,
+      "grad_norm": 0.12047906524830992,
+      "learning_rate": 0.00013011152416356878,
+      "loss": 1.9141,
+      "step": 525
+    },
+    {
+      "epoch": 0.06568755035012704,
+      "grad_norm": 0.11738452165789834,
+      "learning_rate": 0.00013135068153655513,
+      "loss": 1.9887,
+      "step": 530
+    },
+    {
+      "epoch": 0.06630724422135464,
+      "grad_norm": 0.12704500717051184,
+      "learning_rate": 0.0001325898389095415,
+      "loss": 1.9461,
+      "step": 535
+    },
+    {
+      "epoch": 0.06692693809258227,
+      "grad_norm": 0.12232806111597486,
+      "learning_rate": 0.0001338289962825279,
+      "loss": 1.9965,
+      "step": 540
+    },
+    {
+      "epoch": 0.06754663196380987,
+      "grad_norm": 0.11511330645040316,
+      "learning_rate": 0.00013506815365551427,
+      "loss": 1.9387,
+      "step": 545
+    },
+    {
+      "epoch": 0.0681663258350375,
+      "grad_norm": 0.12623271700622418,
+      "learning_rate": 0.00013630731102850062,
+      "loss": 1.8922,
+      "step": 550
+    },
+    {
+      "epoch": 0.0687860197062651,
+      "grad_norm": 0.1233130933433468,
+      "learning_rate": 0.000137546468401487,
+      "loss": 1.9695,
+      "step": 555
+    },
+    {
+      "epoch": 0.06940571357749271,
+      "grad_norm": 0.12482945990926926,
+      "learning_rate": 0.00013878562577447336,
+      "loss": 1.973,
+      "step": 560
+    },
+    {
+      "epoch": 0.07002540744872034,
+      "grad_norm": 0.12412994649354765,
+      "learning_rate": 0.00014002478314745974,
+      "loss": 1.9359,
+      "step": 565
+    },
+    {
+      "epoch": 0.07064510131994794,
+      "grad_norm": 0.12960183005831585,
+      "learning_rate": 0.00014126394052044612,
+      "loss": 1.8832,
+      "step": 570
+    },
+    {
+      "epoch": 0.07126479519117555,
+      "grad_norm": 0.13101220872323543,
+      "learning_rate": 0.00014250309789343247,
+      "loss": 1.9379,
+      "step": 575
+    },
+    {
+      "epoch": 0.07188448906240318,
+      "grad_norm": 0.11679270737421997,
+      "learning_rate": 0.00014374225526641885,
+      "loss": 1.9242,
+      "step": 580
+    },
+    {
+      "epoch": 0.07250418293363078,
+      "grad_norm": 0.11469156246728562,
+      "learning_rate": 0.0001449814126394052,
+      "loss": 1.9559,
+      "step": 585
+    },
+    {
+      "epoch": 0.0731238768048584,
+      "grad_norm": 0.1190243467694481,
+      "learning_rate": 0.00014622057001239158,
+      "loss": 1.8816,
+      "step": 590
+    },
+    {
+      "epoch": 0.07374357067608601,
+      "grad_norm": 0.12510722707465416,
+      "learning_rate": 0.00014745972738537794,
+      "loss": 1.9332,
+      "step": 595
+    },
+    {
+      "epoch": 0.07436326454731362,
+      "grad_norm": 0.14223237579204956,
+      "learning_rate": 0.00014869888475836432,
+      "loss": 1.9219,
+      "step": 600
+    },
+    {
+      "epoch": 0.07498295841854125,
+      "grad_norm": 0.12710564296924554,
+      "learning_rate": 0.0001499380421313507,
+      "loss": 1.9555,
+      "step": 605
+    },
+    {
+      "epoch": 0.07560265228976885,
+      "grad_norm": 0.11559551956238681,
+      "learning_rate": 0.00015117719950433707,
+      "loss": 1.9074,
+      "step": 610
+    },
+    {
+      "epoch": 0.07622234616099646,
+      "grad_norm": 0.11995055927268913,
+      "learning_rate": 0.00015241635687732343,
+      "loss": 1.8871,
+      "step": 615
+    },
+    {
+      "epoch": 0.07684204003222408,
+      "grad_norm": 0.13173160493796293,
+      "learning_rate": 0.0001536555142503098,
+      "loss": 1.9035,
+      "step": 620
+    },
+    {
+      "epoch": 0.0774617339034517,
+      "grad_norm": 0.1254908751927584,
+      "learning_rate": 0.00015489467162329616,
+      "loss": 1.9352,
+      "step": 625
+    },
+    {
+      "epoch": 0.0780814277746793,
+      "grad_norm": 0.12357179246830041,
+      "learning_rate": 0.00015613382899628254,
+      "loss": 1.932,
+      "step": 630
+    },
+    {
+      "epoch": 0.07870112164590692,
+      "grad_norm": 0.12344265317790121,
+      "learning_rate": 0.00015737298636926892,
+      "loss": 1.9211,
+      "step": 635
+    },
+    {
+      "epoch": 0.07932081551713453,
+      "grad_norm": 0.12852213663167542,
+      "learning_rate": 0.00015861214374225527,
+      "loss": 1.9,
+      "step": 640
+    },
+    {
+      "epoch": 0.07994050938836215,
+      "grad_norm": 0.131328675524144,
+      "learning_rate": 0.00015985130111524165,
+      "loss": 1.9531,
+      "step": 645
+    },
+    {
+      "epoch": 0.08056020325958976,
+      "grad_norm": 0.12494108251782414,
+      "learning_rate": 0.000161090458488228,
+      "loss": 1.8996,
+      "step": 650
+    },
+    {
+      "epoch": 0.08117989713081737,
+      "grad_norm": 0.128320637236905,
+      "learning_rate": 0.00016232961586121439,
+      "loss": 1.9301,
+      "step": 655
+    },
+    {
+      "epoch": 0.081799591002045,
+      "grad_norm": 0.12771710393625474,
+      "learning_rate": 0.00016356877323420074,
+      "loss": 1.9137,
+      "step": 660
+    },
+    {
+      "epoch": 0.0824192848732726,
+      "grad_norm": 0.12694987323527102,
+      "learning_rate": 0.00016480793060718712,
+      "loss": 1.9598,
+      "step": 665
+    },
+    {
+      "epoch": 0.08303897874450021,
+      "grad_norm": 0.13588091244879139,
+      "learning_rate": 0.0001660470879801735,
+      "loss": 1.8539,
+      "step": 670
+    },
+    {
+      "epoch": 0.08365867261572783,
+      "grad_norm": 0.12779362690623974,
+      "learning_rate": 0.00016728624535315988,
+      "loss": 1.9277,
+      "step": 675
+    },
+    {
+      "epoch": 0.08427836648695544,
+      "grad_norm": 0.1325286483243818,
+      "learning_rate": 0.00016852540272614623,
+      "loss": 1.8766,
+      "step": 680
+    },
+    {
+      "epoch": 0.08489806035818306,
+      "grad_norm": 0.127439252199406,
+      "learning_rate": 0.00016976456009913258,
+      "loss": 1.8848,
+      "step": 685
+    },
+    {
+      "epoch": 0.08551775422941067,
+      "grad_norm": 0.1241157391254646,
+      "learning_rate": 0.00017100371747211896,
+      "loss": 1.9816,
+      "step": 690
+    },
+    {
+      "epoch": 0.08613744810063828,
+      "grad_norm": 0.13332515296422381,
+      "learning_rate": 0.00017224287484510532,
+      "loss": 1.9262,
+      "step": 695
+    },
+    {
+      "epoch": 0.0867571419718659,
+      "grad_norm": 0.12252637760350603,
+      "learning_rate": 0.00017348203221809172,
+      "loss": 1.9293,
+      "step": 700
+    },
+    {
+      "epoch": 0.08737683584309351,
+      "grad_norm": 0.12496069484354921,
+      "learning_rate": 0.00017472118959107808,
+      "loss": 1.8855,
+      "step": 705
+    },
+    {
+      "epoch": 0.08799652971432112,
+      "grad_norm": 0.13098416875774188,
+      "learning_rate": 0.00017596034696406446,
+      "loss": 1.9121,
+      "step": 710
+    },
+    {
+      "epoch": 0.08861622358554874,
+      "grad_norm": 0.12342851589611775,
+      "learning_rate": 0.0001771995043370508,
+      "loss": 1.9309,
+      "step": 715
+    },
+    {
+      "epoch": 0.08923591745677635,
+      "grad_norm": 0.12375138351033128,
+      "learning_rate": 0.0001784386617100372,
+      "loss": 1.8801,
+      "step": 720
+    },
+    {
+      "epoch": 0.08985561132800396,
+      "grad_norm": 0.12516445329278314,
+      "learning_rate": 0.00017967781908302354,
+      "loss": 1.902,
+      "step": 725
+    },
+    {
+      "epoch": 0.09047530519923158,
+      "grad_norm": 0.12939315510028193,
+      "learning_rate": 0.00018091697645600992,
+      "loss": 1.9633,
+      "step": 730
+    },
+    {
+      "epoch": 0.09109499907045919,
+      "grad_norm": 0.12962320171696218,
+      "learning_rate": 0.0001821561338289963,
+      "loss": 1.9008,
+      "step": 735
+    },
+    {
+      "epoch": 0.09171469294168681,
+      "grad_norm": 0.12286398386603685,
+      "learning_rate": 0.00018339529120198265,
+      "loss": 1.8996,
+      "step": 740
+    },
+    {
+      "epoch": 0.09233438681291442,
+      "grad_norm": 0.1304419130957006,
+      "learning_rate": 0.00018463444857496903,
+      "loss": 1.9105,
+      "step": 745
+    },
+    {
+      "epoch": 0.09295408068414203,
+      "grad_norm": 0.1283679639528834,
+      "learning_rate": 0.0001858736059479554,
+      "loss": 1.8859,
+      "step": 750
+    },
+    {
+      "epoch": 0.09357377455536965,
+      "grad_norm": 0.12212618023022137,
+      "learning_rate": 0.00018711276332094177,
+      "loss": 1.882,
+      "step": 755
+    },
+    {
+      "epoch": 0.09419346842659726,
+      "grad_norm": 0.12769862702936974,
+      "learning_rate": 0.00018835192069392812,
+      "loss": 1.9723,
+      "step": 760
+    },
+    {
+      "epoch": 0.09481316229782487,
+      "grad_norm": 0.12391205743839302,
+      "learning_rate": 0.00018959107806691453,
+      "loss": 1.907,
+      "step": 765
+    },
+    {
+      "epoch": 0.09543285616905249,
+      "grad_norm": 0.12369280757608463,
+      "learning_rate": 0.00019083023543990088,
+      "loss": 1.9145,
+      "step": 770
+    },
+    {
+      "epoch": 0.0960525500402801,
+      "grad_norm": 0.12396791138067718,
+      "learning_rate": 0.00019206939281288726,
+      "loss": 1.8906,
+      "step": 775
+    },
+    {
+      "epoch": 0.09667224391150772,
+      "grad_norm": 0.12724885128353655,
+      "learning_rate": 0.0001933085501858736,
+      "loss": 1.907,
+      "step": 780
+    },
+    {
+      "epoch": 0.09729193778273533,
+      "grad_norm": 0.12534998156680616,
+      "learning_rate": 0.00019454770755886,
+      "loss": 1.8695,
+      "step": 785
+    },
+    {
+      "epoch": 0.09791163165396294,
+      "grad_norm": 0.13128368147328945,
+      "learning_rate": 0.00019578686493184635,
+      "loss": 1.9289,
+      "step": 790
+    },
+    {
+      "epoch": 0.09853132552519056,
+      "grad_norm": 0.13142114693955176,
+      "learning_rate": 0.00019702602230483272,
+      "loss": 1.8863,
+      "step": 795
+    },
+    {
+      "epoch": 0.09915101939641817,
+      "grad_norm": 0.12804903283927577,
+      "learning_rate": 0.0001982651796778191,
+      "loss": 1.9574,
+      "step": 800
+    },
+    {
+      "epoch": 0.09977071326764578,
+      "grad_norm": 0.12560616934907057,
+      "learning_rate": 0.00019950433705080546,
+      "loss": 1.9027,
+      "step": 805
+    },
+    {
+      "epoch": 0.1003904071388734,
+      "grad_norm": 0.13603145482080747,
+      "learning_rate": 0.00019999991575981254,
+      "loss": 1.8602,
+      "step": 810
+    },
+    {
+      "epoch": 0.10101010101010101,
+      "grad_norm": 0.1283260358018925,
+      "learning_rate": 0.00019999940095918086,
+      "loss": 1.9051,
+      "step": 815
+    },
+    {
+      "epoch": 0.10162979488132863,
+      "grad_norm": 0.1293026088162003,
+      "learning_rate": 0.000199998418160428,
+      "loss": 1.9285,
+      "step": 820
+    },
+    {
+      "epoch": 0.10224948875255624,
+      "grad_norm": 0.13075097487753526,
+      "learning_rate": 0.00019999696736815346,
+      "loss": 1.9273,
+      "step": 825
+    },
+    {
+      "epoch": 0.10286918262378385,
+      "grad_norm": 0.12662617461298337,
+      "learning_rate": 0.000199995048589147,
+      "loss": 1.8934,
+      "step": 830
+    },
+    {
+      "epoch": 0.10348887649501147,
+      "grad_norm": 0.13113444016935866,
+      "learning_rate": 0.00019999266183238847,
+      "loss": 1.8984,
+      "step": 835
+    },
+    {
+      "epoch": 0.10410857036623908,
+      "grad_norm": 0.12944550629626753,
+      "learning_rate": 0.00019998980710904794,
+      "loss": 1.9191,
+      "step": 840
+    },
+    {
+      "epoch": 0.10472826423746669,
+      "grad_norm": 0.13398656236337397,
+      "learning_rate": 0.00019998648443248556,
+      "loss": 1.9109,
+      "step": 845
+    },
+    {
+      "epoch": 0.10534795810869431,
+      "grad_norm": 0.1267682688223831,
+      "learning_rate": 0.00019998269381825147,
+      "loss": 1.875,
+      "step": 850
+    },
+    {
+      "epoch": 0.10596765197992192,
+      "grad_norm": 0.12746917546254172,
+      "learning_rate": 0.00019997843528408576,
+      "loss": 1.8754,
+      "step": 855
+    },
+    {
+      "epoch": 0.10658734585114953,
+      "grad_norm": 0.1254555174289125,
+      "learning_rate": 0.00019997370884991842,
+      "loss": 1.9078,
+      "step": 860
+    },
+    {
+      "epoch": 0.10720703972237715,
+      "grad_norm": 0.12777549717459696,
+      "learning_rate": 0.0001999685145378692,
+      "loss": 1.8938,
+      "step": 865
+    },
+    {
+      "epoch": 0.10782673359360476,
+      "grad_norm": 0.14795135942631768,
+      "learning_rate": 0.00019996285237224758,
+      "loss": 1.8574,
+      "step": 870
+    },
+    {
+      "epoch": 0.10844642746483238,
+      "grad_norm": 0.13079486561769416,
+      "learning_rate": 0.00019995672237955246,
+      "loss": 1.9207,
+      "step": 875
+    },
+    {
+      "epoch": 0.10906612133605999,
+      "grad_norm": 0.12773479517789088,
+      "learning_rate": 0.00019995012458847233,
+      "loss": 1.9152,
+      "step": 880
+    },
+    {
+      "epoch": 0.1096858152072876,
+      "grad_norm": 0.14019466780605774,
+      "learning_rate": 0.00019994305902988488,
+      "loss": 1.8484,
+      "step": 885
+    },
+    {
+      "epoch": 0.11030550907851522,
+      "grad_norm": 0.12914830902783628,
+      "learning_rate": 0.00019993552573685703,
+      "loss": 1.9531,
+      "step": 890
+    },
+    {
+      "epoch": 0.11092520294974283,
+      "grad_norm": 0.12289429503961614,
+      "learning_rate": 0.00019992752474464463,
+      "loss": 1.8996,
+      "step": 895
+    },
+    {
+      "epoch": 0.11154489682097044,
+      "grad_norm": 0.13824146618127445,
+      "learning_rate": 0.00019991905609069237,
+      "loss": 1.9203,
+      "step": 900
+    },
+    {
+      "epoch": 0.11216459069219806,
+      "grad_norm": 0.12867807538864207,
+      "learning_rate": 0.0001999101198146337,
+      "loss": 1.8961,
+      "step": 905
+    },
+    {
+      "epoch": 0.11278428456342567,
+      "grad_norm": 0.1284327023064257,
+      "learning_rate": 0.0001999007159582904,
+      "loss": 1.8984,
+      "step": 910
+    },
+    {
+      "epoch": 0.11340397843465329,
+      "grad_norm": 0.13149370957180515,
+      "learning_rate": 0.00019989084456567267,
+      "loss": 1.8809,
+      "step": 915
+    },
+    {
+      "epoch": 0.1140236723058809,
+      "grad_norm": 0.14535516131324824,
+      "learning_rate": 0.00019988050568297866,
+      "loss": 1.9066,
+      "step": 920
+    },
+    {
+      "epoch": 0.1146433661771085,
+      "grad_norm": 0.13159281057665195,
+      "learning_rate": 0.0001998696993585945,
+      "loss": 1.9094,
+      "step": 925
+    },
+    {
+      "epoch": 0.11526306004833613,
+      "grad_norm": 0.1361806065321,
+      "learning_rate": 0.00019985842564309382,
+      "loss": 1.9602,
+      "step": 930
+    },
+    {
+      "epoch": 0.11588275391956374,
+      "grad_norm": 0.12097934827295764,
+      "learning_rate": 0.00019984668458923775,
+      "loss": 1.8895,
+      "step": 935
+    },
+    {
+      "epoch": 0.11650244779079134,
+      "grad_norm": 0.12566168126158242,
+      "learning_rate": 0.00019983447625197457,
+      "loss": 1.8875,
+      "step": 940
+    },
+    {
+      "epoch": 0.11712214166201897,
+      "grad_norm": 0.12430870531788435,
+      "learning_rate": 0.0001998218006884393,
+      "loss": 1.9559,
+      "step": 945
+    },
+    {
+      "epoch": 0.11774183553324657,
+      "grad_norm": 0.12903502537406508,
+      "learning_rate": 0.0001998086579579538,
+      "loss": 1.8797,
+      "step": 950
+    },
+    {
+      "epoch": 0.11836152940447418,
+      "grad_norm": 0.12526103131226338,
+      "learning_rate": 0.00019979504812202612,
+      "loss": 1.907,
+      "step": 955
+    },
+    {
+      "epoch": 0.1189812232757018,
+      "grad_norm": 0.1291422540842116,
+      "learning_rate": 0.00019978097124435042,
+      "loss": 1.9887,
+      "step": 960
+    },
+    {
+      "epoch": 0.11960091714692941,
+      "grad_norm": 0.13690123211378433,
+      "learning_rate": 0.00019976642739080665,
+      "loss": 1.8879,
+      "step": 965
+    },
+    {
+      "epoch": 0.12022061101815704,
+      "grad_norm": 0.13042250114631507,
+      "learning_rate": 0.00019975141662946014,
+      "loss": 1.8988,
+      "step": 970
+    },
+    {
+      "epoch": 0.12084030488938464,
+      "grad_norm": 0.1450939461723542,
+      "learning_rate": 0.0001997359390305614,
+      "loss": 1.8363,
+      "step": 975
+    },
+    {
+      "epoch": 0.12145999876061225,
+      "grad_norm": 0.12494449135760256,
+      "learning_rate": 0.00019971999466654577,
+      "loss": 1.8691,
+      "step": 980
+    },
+    {
+      "epoch": 0.12207969263183988,
+      "grad_norm": 0.13013968478788646,
+      "learning_rate": 0.000199703583612033,
+      "loss": 1.8727,
+      "step": 985
+    },
+    {
+      "epoch": 0.12269938650306748,
+      "grad_norm": 0.12026110255667402,
+      "learning_rate": 0.00019968670594382694,
+      "loss": 1.9199,
+      "step": 990
+    },
+    {
+      "epoch": 0.12331908037429509,
+      "grad_norm": 0.13639861474847226,
+      "learning_rate": 0.00019966936174091527,
+      "loss": 1.875,
+      "step": 995
+    },
+    {
+      "epoch": 0.12393877424552271,
+      "grad_norm": 0.1311732320208071,
+      "learning_rate": 0.00019965155108446906,
+      "loss": 1.9246,
+      "step": 1000
+    },
+    {
+      "epoch": 0.12455846811675032,
+      "grad_norm": 0.13525131264110185,
+      "learning_rate": 0.00019963327405784226,
+      "loss": 1.9121,
+      "step": 1005
+    },
+    {
+      "epoch": 0.12517816198797793,
+      "grad_norm": 0.1377173253291631,
+      "learning_rate": 0.0001996145307465716,
+      "loss": 1.9074,
+      "step": 1010
+    },
+    {
+      "epoch": 0.12579785585920555,
+      "grad_norm": 0.12714986650171514,
+      "learning_rate": 0.00019959532123837588,
+      "loss": 1.9238,
+      "step": 1015
+    },
+    {
+      "epoch": 0.12641754973043318,
+      "grad_norm": 0.1316460319074934,
+      "learning_rate": 0.00019957564562315583,
+      "loss": 1.8453,
+      "step": 1020
+    },
+    {
+      "epoch": 0.12703724360166077,
+      "grad_norm": 0.12924564279125103,
+      "learning_rate": 0.0001995555039929935,
+      "loss": 1.9152,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1276569374728884,
+      "grad_norm": 0.12389037014071365,
+      "learning_rate": 0.00019953489644215186,
+      "loss": 1.9145,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12827663134411602,
+      "grad_norm": 0.12731994833926935,
+      "learning_rate": 0.00019951382306707452,
+      "loss": 1.8539,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1288963252153436,
+      "grad_norm": 0.13145951711925763,
+      "learning_rate": 0.00019949228396638503,
+      "loss": 1.8926,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12951601908657123,
+      "grad_norm": 0.1345163191426047,
+      "learning_rate": 0.00019947027924088656,
+      "loss": 1.8746,
+      "step": 1045
+    },
+    {
+      "epoch": 0.13013571295779885,
+      "grad_norm": 0.13149828806285122,
+      "learning_rate": 0.00019944780899356146,
+      "loss": 1.9363,
+      "step": 1050
+    },
+    {
+      "epoch": 0.13075540682902645,
+      "grad_norm": 0.131642891757479,
+      "learning_rate": 0.00019942487332957066,
+      "loss": 1.8734,
+      "step": 1055
+    },
+    {
+      "epoch": 0.13137510070025407,
+      "grad_norm": 0.1357864698335193,
+      "learning_rate": 0.0001994014723562533,
+      "loss": 1.8805,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1319947945714817,
+      "grad_norm": 0.13726338790841353,
+      "learning_rate": 0.00019937760618312617,
+      "loss": 1.9238,
+      "step": 1065
+    },
+    {
+      "epoch": 0.1326144884427093,
+      "grad_norm": 0.13519938007390137,
+      "learning_rate": 0.00019935327492188315,
+      "loss": 1.9059,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1332341823139369,
+      "grad_norm": 0.1324182900169985,
+      "learning_rate": 0.0001993284786863948,
+      "loss": 1.8719,
+      "step": 1075
+    },
+    {
+      "epoch": 0.13385387618516453,
+      "grad_norm": 0.13711559928347855,
+      "learning_rate": 0.0001993032175927077,
+      "loss": 1.9367,
+      "step": 1080
+    },
+    {
+      "epoch": 0.13447357005639216,
+      "grad_norm": 0.1334759916213516,
+      "learning_rate": 0.00019927749175904403,
+      "loss": 1.8883,
+      "step": 1085
+    },
+    {
+      "epoch": 0.13509326392761975,
+      "grad_norm": 0.12943730877342607,
+      "learning_rate": 0.0001992513013058009,
+      "loss": 1.8887,
+      "step": 1090
+    },
+    {
+      "epoch": 0.13571295779884737,
+      "grad_norm": 0.12672939466440764,
+      "learning_rate": 0.00019922464635554988,
+      "loss": 1.8449,
+      "step": 1095
+    },
+    {
+      "epoch": 0.136332651670075,
+      "grad_norm": 0.12225320336434065,
+      "learning_rate": 0.0001991975270330364,
+      "loss": 1.8719,
+      "step": 1100
+    },
+    {
+      "epoch": 0.1369523455413026,
+      "grad_norm": 0.12850014269871904,
+      "learning_rate": 0.00019916994346517915,
+      "loss": 1.8727,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1375720394125302,
+      "grad_norm": 0.13328718338585124,
+      "learning_rate": 0.00019914189578106945,
+      "loss": 1.9055,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13819173328375783,
+      "grad_norm": 0.12984485764601156,
+      "learning_rate": 0.00019911338411197075,
+      "loss": 1.9129,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13881142715498543,
+      "grad_norm": 0.12879734723470498,
+      "learning_rate": 0.0001990844085913179,
+      "loss": 1.8289,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13943112102621305,
+      "grad_norm": 0.12045308358521092,
+      "learning_rate": 0.00019905496935471658,
+      "loss": 1.8953,
+      "step": 1125
+    },
+    {
+      "epoch": 0.14005081489744067,
+      "grad_norm": 0.13681652513593023,
+      "learning_rate": 0.00019902506653994277,
+      "loss": 1.8285,
+      "step": 1130
+    },
+    {
+      "epoch": 0.14067050876866827,
+      "grad_norm": 0.12537351067620614,
+      "learning_rate": 0.00019899470028694185,
+      "loss": 1.8625,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1412902026398959,
+      "grad_norm": 0.13310144010439956,
+      "learning_rate": 0.0001989638707378282,
+      "loss": 1.9676,
+      "step": 1140
+    },
+    {
+      "epoch": 0.1419098965111235,
+      "grad_norm": 0.12651220372995958,
+      "learning_rate": 0.0001989325780368844,
+      "loss": 1.8848,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1425295903823511,
+      "grad_norm": 0.13498557406467906,
+      "learning_rate": 0.00019890082233056054,
+      "loss": 1.9281,
+      "step": 1150
+    },
+    {
+      "epoch": 0.14314928425357873,
+      "grad_norm": 0.13324683156087155,
+      "learning_rate": 0.00019886860376747362,
+      "loss": 1.8738,
+      "step": 1155
+    },
+    {
+      "epoch": 0.14376897812480635,
+      "grad_norm": 0.12727156852136712,
+      "learning_rate": 0.0001988359224984069,
+      "loss": 1.9187,
+      "step": 1160
+    },
+    {
+      "epoch": 0.14438867199603395,
+      "grad_norm": 0.13322593653239903,
+      "learning_rate": 0.0001988027786763089,
+      "loss": 1.8191,
+      "step": 1165
+    },
+    {
+      "epoch": 0.14500836586726157,
+      "grad_norm": 0.13430900015454927,
+      "learning_rate": 0.0001987691724562931,
+      "loss": 1.8746,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1456280597384892,
+      "grad_norm": 0.1335734771928228,
+      "learning_rate": 0.00019873510399563688,
+      "loss": 1.9012,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1462477536097168,
+      "grad_norm": 0.12408030014942514,
+      "learning_rate": 0.00019870057345378097,
+      "loss": 1.8648,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1468674474809444,
+      "grad_norm": 0.11956244949759506,
+      "learning_rate": 0.00019866558099232862,
+      "loss": 1.8637,
+      "step": 1185
+    },
+    {
+      "epoch": 0.14748714135217203,
+      "grad_norm": 0.138357841858996,
+      "learning_rate": 0.00019863012677504485,
+      "loss": 1.8734,
+      "step": 1190
+    },
+    {
+      "epoch": 0.14810683522339965,
+      "grad_norm": 0.12851504413100462,
+      "learning_rate": 0.00019859421096785575,
+      "loss": 1.8891,
+      "step": 1195
+    },
+    {
+      "epoch": 0.14872652909462725,
+      "grad_norm": 0.13319470956489035,
+      "learning_rate": 0.00019855783373884763,
+      "loss": 1.8418,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14934622296585487,
+      "grad_norm": 0.1251908795342198,
+      "learning_rate": 0.00019852099525826628,
+      "loss": 1.9004,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1499659168370825,
+      "grad_norm": 0.12994851392351664,
+      "learning_rate": 0.00019848369569851608,
+      "loss": 1.8406,
+      "step": 1210
+    },
+    {
+      "epoch": 0.15058561070831009,
+      "grad_norm": 0.12421201591064086,
+      "learning_rate": 0.00019844593523415935,
+      "loss": 1.927,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1512053045795377,
+      "grad_norm": 0.1348561729670815,
+      "learning_rate": 0.00019840771404191538,
+      "loss": 1.8785,
+      "step": 1220
+    },
+    {
+      "epoch": 0.15182499845076533,
+      "grad_norm": 0.12902823419358234,
+      "learning_rate": 0.00019836903230065973,
+      "loss": 1.9129,
+      "step": 1225
+    },
+    {
+      "epoch": 0.15244469232199293,
+      "grad_norm": 0.1303696455593782,
+      "learning_rate": 0.0001983298901914233,
+      "loss": 1.8965,
+      "step": 1230
+    },
+    {
+      "epoch": 0.15306438619322055,
+      "grad_norm": 0.12750859082981958,
+      "learning_rate": 0.00019829028789739156,
+      "loss": 1.8863,
+      "step": 1235
+    },
+    {
+      "epoch": 0.15368408006444817,
+      "grad_norm": 0.13301195737140625,
+      "learning_rate": 0.00019825022560390353,
+      "loss": 1.8848,
+      "step": 1240
+    },
+    {
+      "epoch": 0.15430377393567576,
+      "grad_norm": 0.12458777864772885,
+      "learning_rate": 0.00019820970349845117,
+      "loss": 1.918,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1549234678069034,
+      "grad_norm": 0.12157252838516859,
+      "learning_rate": 0.0001981687217706783,
+      "loss": 1.9004,
+      "step": 1250
+    },
+    {
+      "epoch": 0.155543161678131,
+      "grad_norm": 0.12992471411020662,
+      "learning_rate": 0.0001981272806123798,
+      "loss": 1.8449,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1561628555493586,
+      "grad_norm": 0.13490267199807432,
+      "learning_rate": 0.00019808538021750063,
+      "loss": 1.8594,
+      "step": 1260
+    },
+    {
+      "epoch": 0.15678254942058623,
+      "grad_norm": 0.13254926124911628,
+      "learning_rate": 0.00019804302078213506,
+      "loss": 1.9066,
+      "step": 1265
+    },
+    {
+      "epoch": 0.15740224329181385,
+      "grad_norm": 0.13655436155848294,
+      "learning_rate": 0.00019800020250452564,
+      "loss": 1.9094,
+      "step": 1270
+    },
+    {
+      "epoch": 0.15802193716304147,
+      "grad_norm": 0.12605752395912317,
+      "learning_rate": 0.00019795692558506232,
+      "loss": 1.8914,
+      "step": 1275
+    },
+    {
+      "epoch": 0.15864163103426906,
+      "grad_norm": 0.12417347152492184,
+      "learning_rate": 0.0001979131902262814,
+      "loss": 1.891,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1592613249054967,
+      "grad_norm": 0.1230993318638121,
+      "learning_rate": 0.00019786899663286486,
+      "loss": 1.9051,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1598810187767243,
+      "grad_norm": 0.13002111203651795,
+      "learning_rate": 0.00019782434501163909,
+      "loss": 1.9371,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1605007126479519,
+      "grad_norm": 0.12883861862990198,
+      "learning_rate": 0.0001977792355715741,
+      "loss": 1.9074,
+      "step": 1295
+    },
+    {
+      "epoch": 0.16112040651917953,
+      "grad_norm": 0.12844516943687437,
+      "learning_rate": 0.00019773366852378246,
+      "loss": 1.8348,
+      "step": 1300
+    },
+    {
+      "epoch": 0.16174010039040715,
+      "grad_norm": 0.12861625452896877,
+      "learning_rate": 0.0001976876440815184,
+      "loss": 1.8395,
+      "step": 1305
+    },
+    {
+      "epoch": 0.16235979426163474,
+      "grad_norm": 0.12058020178260095,
+      "learning_rate": 0.0001976411624601767,
+      "loss": 1.8328,
+      "step": 1310
+    },
+    {
+      "epoch": 0.16297948813286237,
+      "grad_norm": 0.1266307036564728,
+      "learning_rate": 0.00019759422387729183,
+      "loss": 1.8406,
+      "step": 1315
+    },
+    {
+      "epoch": 0.16359918200409,
+      "grad_norm": 0.1320325104662958,
+      "learning_rate": 0.00019754682855253674,
+      "loss": 1.8789,
+      "step": 1320
+    },
+    {
+      "epoch": 0.16421887587531758,
+      "grad_norm": 0.13663051531666534,
+      "learning_rate": 0.00019749897670772205,
+      "loss": 1.8375,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1648385697465452,
+      "grad_norm": 0.1264957686459983,
+      "learning_rate": 0.00019745066856679478,
+      "loss": 1.8758,
+      "step": 1330
+    },
+    {
+      "epoch": 0.16545826361777283,
+      "grad_norm": 0.13992651333444445,
+      "learning_rate": 0.0001974019043558375,
+      "loss": 1.8523,
+      "step": 1335
+    },
+    {
+      "epoch": 0.16607795748900042,
+      "grad_norm": 0.12649107132742746,
+      "learning_rate": 0.00019735268430306718,
+      "loss": 1.8438,
+      "step": 1340
+    },
+    {
+      "epoch": 0.16669765136022804,
+      "grad_norm": 0.12995820389298687,
+      "learning_rate": 0.00019730300863883405,
+      "loss": 1.8723,
+      "step": 1345
+    },
+    {
+      "epoch": 0.16731734523145567,
+      "grad_norm": 0.12383168873813469,
+      "learning_rate": 0.0001972528775956208,
+      "loss": 1.8973,
+      "step": 1350
+    },
+    {
+      "epoch": 0.16793703910268326,
+      "grad_norm": 0.12991019310770477,
+      "learning_rate": 0.0001972022914080411,
+      "loss": 1.8313,
+      "step": 1355
+    },
+    {
+      "epoch": 0.16855673297391088,
+      "grad_norm": 0.12295152055363918,
+      "learning_rate": 0.00019715125031283877,
+      "loss": 1.8652,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1691764268451385,
+      "grad_norm": 0.13895756409315704,
+      "learning_rate": 0.00019709975454888662,
+      "loss": 1.8391,
+      "step": 1365
+    },
+    {
+      "epoch": 0.16979612071636613,
+      "grad_norm": 0.12962858797129284,
+      "learning_rate": 0.00019704780435718532,
+      "loss": 1.8965,
+      "step": 1370
+    },
+    {
+      "epoch": 0.17041581458759372,
+      "grad_norm": 0.13434107430628525,
+      "learning_rate": 0.00019699539998086223,
+      "loss": 1.8488,
+      "step": 1375
+    },
+    {
+      "epoch": 0.17103550845882134,
+      "grad_norm": 0.12227223666385428,
+      "learning_rate": 0.00019694254166517032,
+      "loss": 1.8836,
+      "step": 1380
+    },
+    {
+      "epoch": 0.17165520233004897,
+      "grad_norm": 0.13008761760930657,
+      "learning_rate": 0.00019688922965748696,
+      "loss": 1.8973,
+      "step": 1385
+    },
+    {
+      "epoch": 0.17227489620127656,
+      "grad_norm": 0.12801298155182156,
+      "learning_rate": 0.0001968354642073129,
+      "loss": 1.8383,
+      "step": 1390
+    },
+    {
+      "epoch": 0.17289459007250418,
+      "grad_norm": 0.1292755368394873,
+      "learning_rate": 0.00019678124556627094,
+      "loss": 1.8434,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1735142839437318,
+      "grad_norm": 0.13946593061603393,
+      "learning_rate": 0.00019672657398810478,
+      "loss": 1.8535,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1741339778149594,
+      "grad_norm": 0.14090247834119649,
+      "learning_rate": 0.00019667144972867795,
+      "loss": 1.8801,
+      "step": 1405
+    },
+    {
+      "epoch": 0.17475367168618702,
+      "grad_norm": 0.13668444514745795,
+      "learning_rate": 0.00019661587304597243,
+      "loss": 1.8801,
+      "step": 1410
+    },
+    {
+      "epoch": 0.17537336555741465,
+      "grad_norm": 0.13208454467125907,
+      "learning_rate": 0.0001965598442000877,
+      "loss": 1.9043,
+      "step": 1415
+    },
+    {
+      "epoch": 0.17599305942864224,
+      "grad_norm": 0.1392934854468196,
+      "learning_rate": 0.0001965033634532392,
+      "loss": 1.8613,
+      "step": 1420
+    },
+    {
+      "epoch": 0.17661275329986986,
+      "grad_norm": 0.12790012697738787,
+      "learning_rate": 0.00019644643106975739,
+      "loss": 1.8238,
+      "step": 1425
+    },
+    {
+      "epoch": 0.17723244717109748,
+      "grad_norm": 0.13787833838331934,
+      "learning_rate": 0.00019638904731608637,
+      "loss": 1.9125,
+      "step": 1430
+    },
+    {
+      "epoch": 0.17785214104232508,
+      "grad_norm": 0.1243448758230821,
+      "learning_rate": 0.00019633121246078256,
+      "loss": 1.8512,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1784718349135527,
+      "grad_norm": 0.12359501590686292,
+      "learning_rate": 0.00019627292677451368,
+      "loss": 1.8918,
+      "step": 1440
+    },
+    {
+      "epoch": 0.17909152878478032,
+      "grad_norm": 0.14192690308588027,
+      "learning_rate": 0.00019621419053005726,
+      "loss": 1.8824,
+      "step": 1445
+    },
+    {
+      "epoch": 0.17971122265600792,
+      "grad_norm": 0.12650409825125716,
+      "learning_rate": 0.00019615500400229946,
+      "loss": 1.8875,
+      "step": 1450
+    },
+    {
+      "epoch": 0.18033091652723554,
+      "grad_norm": 0.130110107018088,
+      "learning_rate": 0.0001960953674682338,
+      "loss": 1.8426,
+      "step": 1455
+    },
+    {
+      "epoch": 0.18095061039846316,
+      "grad_norm": 0.12396235695123667,
+      "learning_rate": 0.00019603528120695982,
+      "loss": 1.9012,
+      "step": 1460
+    },
+    {
+      "epoch": 0.18157030426969079,
+      "grad_norm": 0.12716887462079618,
+      "learning_rate": 0.00019597474549968173,
+      "loss": 1.8551,
+      "step": 1465
+    },
+    {
+      "epoch": 0.18218999814091838,
+      "grad_norm": 0.11791740501594407,
+      "learning_rate": 0.00019591376062970728,
+      "loss": 1.9535,
+      "step": 1470
+    },
+    {
+      "epoch": 0.182809692012146,
+      "grad_norm": 0.1274909732144608,
+      "learning_rate": 0.00019585232688244613,
+      "loss": 1.8738,
+      "step": 1475
+    },
+    {
+      "epoch": 0.18342938588337362,
+      "grad_norm": 0.12950727276315843,
+      "learning_rate": 0.00019579044454540883,
+      "loss": 1.8336,
+      "step": 1480
+    },
+    {
+      "epoch": 0.18404907975460122,
+      "grad_norm": 0.12786571322089985,
+      "learning_rate": 0.0001957281139082053,
+      "loss": 1.9125,
+      "step": 1485
+    },
+    {
+      "epoch": 0.18466877362582884,
+      "grad_norm": 0.1238604246672542,
+      "learning_rate": 0.00019566533526254348,
+      "loss": 1.8684,
+      "step": 1490
+    },
+    {
+      "epoch": 0.18528846749705646,
+      "grad_norm": 0.11970443128542427,
+      "learning_rate": 0.00019560210890222802,
+      "loss": 1.8883,
+      "step": 1495
+    },
+    {
+      "epoch": 0.18590816136828406,
+      "grad_norm": 0.12314025496676621,
+      "learning_rate": 0.00019553843512315887,
+      "loss": 1.9094,
+      "step": 1500
+    },
+    {
+      "epoch": 0.18652785523951168,
+      "grad_norm": 0.12798261977540473,
+      "learning_rate": 0.00019547431422332992,
+      "loss": 1.8488,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1871475491107393,
+      "grad_norm": 0.1286897352380571,
+      "learning_rate": 0.00019540974650282756,
+      "loss": 1.9156,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1877672429819669,
+      "grad_norm": 0.1359355825058481,
+      "learning_rate": 0.0001953447322638293,
+      "loss": 1.8629,
+      "step": 1515
+    },
+    {
+      "epoch": 0.18838693685319452,
+      "grad_norm": 0.13358157574590807,
+      "learning_rate": 0.0001952792718106024,
+      "loss": 1.8797,
+      "step": 1520
+    },
+    {
+      "epoch": 0.18900663072442214,
+      "grad_norm": 0.1284059481658899,
+      "learning_rate": 0.00019521336544950238,
+      "loss": 1.8664,
+      "step": 1525
+    },
+    {
+      "epoch": 0.18962632459564974,
+      "grad_norm": 0.14299111302674886,
+      "learning_rate": 0.00019514701348897164,
+      "loss": 1.8043,
+      "step": 1530
+    },
+    {
+      "epoch": 0.19024601846687736,
+      "grad_norm": 0.13170613006037593,
+      "learning_rate": 0.00019508021623953795,
+      "loss": 1.918,
+      "step": 1535
+    },
+    {
+      "epoch": 0.19086571233810498,
+      "grad_norm": 0.1322137048503741,
+      "learning_rate": 0.00019501297401381304,
+      "loss": 1.8687,
+      "step": 1540
+    },
+    {
+      "epoch": 0.19148540620933258,
+      "grad_norm": 0.1372246664399263,
+      "learning_rate": 0.00019494528712649117,
+      "loss": 1.9109,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1921051000805602,
+      "grad_norm": 0.13020416605079768,
+      "learning_rate": 0.0001948771558943476,
+      "loss": 1.907,
+      "step": 1550
+    },
+    {
+      "epoch": 0.19272479395178782,
+      "grad_norm": 0.12282250558731672,
+      "learning_rate": 0.00019480858063623715,
+      "loss": 1.8738,
+      "step": 1555
+    },
+    {
+      "epoch": 0.19334448782301544,
+      "grad_norm": 0.1347150806147481,
+      "learning_rate": 0.0001947395616730926,
+      "loss": 1.907,
+      "step": 1560
+    },
+    {
+      "epoch": 0.19396418169424304,
+      "grad_norm": 0.12434032734923321,
+      "learning_rate": 0.00019467009932792336,
+      "loss": 1.8703,
+      "step": 1565
+    },
+    {
+      "epoch": 0.19458387556547066,
+      "grad_norm": 0.13012917758896814,
+      "learning_rate": 0.00019460019392581387,
+      "loss": 1.9035,
+      "step": 1570
+    },
+    {
+      "epoch": 0.19520356943669828,
+      "grad_norm": 0.12768898254941669,
+      "learning_rate": 0.00019452984579392205,
+      "loss": 1.8516,
+      "step": 1575
+    },
+    {
+      "epoch": 0.19582326330792588,
+      "grad_norm": 0.12366914376030068,
+      "learning_rate": 0.0001944590552614778,
+      "loss": 1.857,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1964429571791535,
+      "grad_norm": 0.15841193439353185,
+      "learning_rate": 0.0001943878226597815,
+      "loss": 1.8621,
+      "step": 1585
+    },
+    {
+      "epoch": 0.19706265105038112,
+      "grad_norm": 0.13946946613234204,
+      "learning_rate": 0.0001943161483222023,
+      "loss": 1.8508,
+      "step": 1590
+    },
+    {
+      "epoch": 0.19768234492160872,
+      "grad_norm": 0.12879914509204915,
+      "learning_rate": 0.00019424403258417683,
+      "loss": 1.875,
+      "step": 1595
+    },
+    {
+      "epoch": 0.19830203879283634,
+      "grad_norm": 0.13449677568515495,
+      "learning_rate": 0.00019417147578320744,
+      "loss": 1.8648,
+      "step": 1600
+    },
+    {
+      "epoch": 0.19892173266406396,
+      "grad_norm": 0.1251274556275624,
+      "learning_rate": 0.00019409847825886054,
+      "loss": 1.923,
+      "step": 1605
+    },
+    {
+      "epoch": 0.19954142653529156,
+      "grad_norm": 0.13841429870418875,
+      "learning_rate": 0.00019402504035276525,
+      "loss": 1.8086,
+      "step": 1610
+    },
+    {
+      "epoch": 0.20016112040651918,
+      "grad_norm": 0.12356779695871337,
+      "learning_rate": 0.00019395116240861172,
+      "loss": 1.8313,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2007808142777468,
+      "grad_norm": 0.1192001269564105,
+      "learning_rate": 0.0001938768447721493,
+      "loss": 1.8734,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2014005081489744,
+      "grad_norm": 0.12435187316100045,
+      "learning_rate": 0.00019380208779118532,
+      "loss": 1.8449,
+      "step": 1625
+    },
+    {
+      "epoch": 0.20202020202020202,
+      "grad_norm": 0.12735167929794539,
+      "learning_rate": 0.00019372689181558307,
+      "loss": 1.8262,
+      "step": 1630
+    },
+    {
+      "epoch": 0.20263989589142964,
+      "grad_norm": 0.12643283485876602,
+      "learning_rate": 0.00019365125719726046,
+      "loss": 1.8355,
+      "step": 1635
+    },
+    {
+      "epoch": 0.20325958976265726,
+      "grad_norm": 0.13092666514756482,
+      "learning_rate": 0.00019357518429018815,
+      "loss": 1.8996,
+      "step": 1640
+    },
+    {
+      "epoch": 0.20387928363388486,
+      "grad_norm": 0.1277631476280917,
+      "learning_rate": 0.00019349867345038808,
+      "loss": 1.8914,
+      "step": 1645
+    },
+    {
+      "epoch": 0.20449897750511248,
+      "grad_norm": 0.13075728650959278,
+      "learning_rate": 0.0001934217250359317,
+      "loss": 1.8426,
+      "step": 1650
+    },
+    {
+      "epoch": 0.2051186713763401,
+      "grad_norm": 0.12669704558351194,
+      "learning_rate": 0.0001933443394069383,
+      "loss": 1.8613,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2057383652475677,
+      "grad_norm": 0.12867521676946295,
+      "learning_rate": 0.0001932665169255733,
+      "loss": 1.8926,
+      "step": 1660
+    },
+    {
+      "epoch": 0.20635805911879532,
+      "grad_norm": 0.1404983490484146,
+      "learning_rate": 0.00019318825795604667,
+      "loss": 1.8676,
+      "step": 1665
+    },
+    {
+      "epoch": 0.20697775299002294,
+      "grad_norm": 0.12750956916082043,
+      "learning_rate": 0.00019310956286461108,
+      "loss": 1.8461,
+      "step": 1670
+    },
+    {
+      "epoch": 0.20759744686125053,
+      "grad_norm": 0.1272989702864413,
+      "learning_rate": 0.00019303043201956033,
+      "loss": 1.8441,
+      "step": 1675
+    },
+    {
+      "epoch": 0.20821714073247816,
+      "grad_norm": 0.13245347386183784,
+      "learning_rate": 0.00019295086579122748,
+      "loss": 1.8059,
+      "step": 1680
+    },
+    {
+      "epoch": 0.20883683460370578,
+      "grad_norm": 0.11974342684601433,
+      "learning_rate": 0.0001928708645519832,
+      "loss": 1.8805,
+      "step": 1685
+    },
+    {
+      "epoch": 0.20945652847493337,
+      "grad_norm": 0.1373585266936289,
+      "learning_rate": 0.00019279042867623405,
+      "loss": 1.875,
+      "step": 1690
+    },
+    {
+      "epoch": 0.210076222346161,
+      "grad_norm": 0.126610203931597,
+      "learning_rate": 0.00019270955854042065,
+      "loss": 1.898,
+      "step": 1695
+    },
+    {
+      "epoch": 0.21069591621738862,
+      "grad_norm": 0.13468171577740737,
+      "learning_rate": 0.00019262825452301603,
+      "loss": 1.8695,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2113156100886162,
+      "grad_norm": 0.13874720674647448,
+      "learning_rate": 0.0001925465170045237,
+      "loss": 1.8281,
+      "step": 1705
+    },
+    {
+      "epoch": 0.21193530395984383,
+      "grad_norm": 0.12340652395591858,
+      "learning_rate": 0.00019246434636747603,
+      "loss": 1.8727,
+      "step": 1710
+    },
+    {
+      "epoch": 0.21255499783107146,
+      "grad_norm": 0.13391821183419367,
+      "learning_rate": 0.00019238174299643235,
+      "loss": 1.8535,
+      "step": 1715
+    },
+    {
+      "epoch": 0.21317469170229905,
+      "grad_norm": 0.1238891855327304,
+      "learning_rate": 0.00019229870727797716,
+      "loss": 1.9117,
+      "step": 1720
+    },
+    {
+      "epoch": 0.21379438557352667,
+      "grad_norm": 0.13921826355338024,
+      "learning_rate": 0.00019221523960071847,
+      "loss": 1.8496,
+      "step": 1725
+    },
+    {
+      "epoch": 0.2144140794447543,
+      "grad_norm": 0.1296981763357881,
+      "learning_rate": 0.00019213134035528574,
+      "loss": 1.9004,
+      "step": 1730
+    },
+    {
+      "epoch": 0.21503377331598192,
+      "grad_norm": 0.1341944973320657,
+      "learning_rate": 0.0001920470099343282,
+      "loss": 1.8727,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2156534671872095,
+      "grad_norm": 0.12518986519019676,
+      "learning_rate": 0.000191962248732513,
+      "loss": 1.9031,
+      "step": 1740
+    },
+    {
+      "epoch": 0.21627316105843714,
+      "grad_norm": 0.12632835935474943,
+      "learning_rate": 0.00019187705714652337,
+      "loss": 1.8535,
+      "step": 1745
+    },
+    {
+      "epoch": 0.21689285492966476,
+      "grad_norm": 0.1263494950807743,
+      "learning_rate": 0.00019179143557505676,
+      "loss": 1.8516,
+      "step": 1750
+    },
+    {
+      "epoch": 0.21751254880089235,
+      "grad_norm": 0.12443856243939533,
+      "learning_rate": 0.0001917053844188228,
+      "loss": 1.8672,
+      "step": 1755
+    },
+    {
+      "epoch": 0.21813224267211997,
+      "grad_norm": 0.1320392752907634,
+      "learning_rate": 0.0001916189040805418,
+      "loss": 1.834,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2187519365433476,
+      "grad_norm": 0.12493530880754049,
+      "learning_rate": 0.0001915319949649425,
+      "loss": 1.8453,
+      "step": 1765
+    },
+    {
+      "epoch": 0.2193716304145752,
+      "grad_norm": 0.1322646318302995,
+      "learning_rate": 0.00019144465747876038,
+      "loss": 1.866,
+      "step": 1770
+    },
+    {
+      "epoch": 0.21999132428580281,
+      "grad_norm": 0.1349930390610452,
+      "learning_rate": 0.00019135689203073563,
+      "loss": 1.857,
+      "step": 1775
+    },
+    {
+      "epoch": 0.22061101815703044,
+      "grad_norm": 0.12533754824423338,
+      "learning_rate": 0.00019126869903161146,
+      "loss": 1.9152,
+      "step": 1780
+    },
+    {
+      "epoch": 0.22123071202825803,
+      "grad_norm": 0.13663631158953193,
+      "learning_rate": 0.00019118007889413186,
+      "loss": 1.8102,
+      "step": 1785
+    },
+    {
+      "epoch": 0.22185040589948565,
+      "grad_norm": 0.12533002892577624,
+      "learning_rate": 0.00019109103203303988,
+      "loss": 1.8387,
+      "step": 1790
+    },
+    {
+      "epoch": 0.22247009977071328,
+      "grad_norm": 0.1359058002472198,
+      "learning_rate": 0.00019100155886507566,
+      "loss": 1.9219,
+      "step": 1795
+    },
+    {
+      "epoch": 0.22308979364194087,
+      "grad_norm": 0.12945958578973213,
+      "learning_rate": 0.0001909116598089745,
+      "loss": 1.8602,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2237094875131685,
+      "grad_norm": 0.1264797795918434,
+      "learning_rate": 0.00019082133528546476,
+      "loss": 1.8727,
+      "step": 1805
+    },
+    {
+      "epoch": 0.22432918138439611,
+      "grad_norm": 0.12673897565989584,
+      "learning_rate": 0.0001907305857172661,
+      "loss": 1.8844,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2249488752556237,
+      "grad_norm": 0.11996164040647495,
+      "learning_rate": 0.00019063941152908727,
+      "loss": 1.8664,
+      "step": 1815
+    },
+    {
+      "epoch": 0.22556856912685133,
+      "grad_norm": 0.12717207452871931,
+      "learning_rate": 0.00019054781314762433,
+      "loss": 1.8543,
+      "step": 1820
+    },
+    {
+      "epoch": 0.22618826299807895,
+      "grad_norm": 0.12807052491062512,
+      "learning_rate": 0.0001904557910015586,
+      "loss": 1.8781,
+      "step": 1825
+    },
+    {
+      "epoch": 0.22680795686930658,
+      "grad_norm": 0.1310959903315099,
+      "learning_rate": 0.00019036334552155452,
+      "loss": 1.8336,
+      "step": 1830
+    },
+    {
+      "epoch": 0.22742765074053417,
+      "grad_norm": 0.13530374164558082,
+      "learning_rate": 0.00019027047714025784,
+      "loss": 1.8531,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2280473446117618,
+      "grad_norm": 0.13283518160109148,
+      "learning_rate": 0.0001901771862922934,
+      "loss": 1.8594,
+      "step": 1840
+    },
+    {
+      "epoch": 0.22866703848298942,
+      "grad_norm": 0.14381832693100272,
+      "learning_rate": 0.00019008347341426324,
+      "loss": 1.8473,
+      "step": 1845
+    },
+    {
+      "epoch": 0.229286732354217,
+      "grad_norm": 0.12561347944988807,
+      "learning_rate": 0.0001899893389447445,
+      "loss": 1.8813,
+      "step": 1850
+    },
+    {
+      "epoch": 0.22990642622544463,
+      "grad_norm": 0.12885427981187214,
+      "learning_rate": 0.0001898947833242874,
+      "loss": 1.8465,
+      "step": 1855
+    },
+    {
+      "epoch": 0.23052612009667225,
+      "grad_norm": 0.1398342796676745,
+      "learning_rate": 0.00018979980699541308,
+      "loss": 1.8602,
+      "step": 1860
+    },
+    {
+      "epoch": 0.23114581396789985,
+      "grad_norm": 0.12603617420354743,
+      "learning_rate": 0.00018970441040261165,
+      "loss": 1.8055,
+      "step": 1865
+    },
+    {
+      "epoch": 0.23176550783912747,
+      "grad_norm": 0.12826056653704176,
+      "learning_rate": 0.00018960859399234006,
+      "loss": 1.8164,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2323852017103551,
+      "grad_norm": 0.12843392111688107,
+      "learning_rate": 0.00018951235821301995,
+      "loss": 1.8934,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2330048955815827,
+      "grad_norm": 0.12781068150685332,
+      "learning_rate": 0.0001894157035150357,
+      "loss": 1.8414,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2336245894528103,
+      "grad_norm": 0.12383040948770438,
+      "learning_rate": 0.00018931863035073217,
+      "loss": 1.9203,
+      "step": 1885
+    },
+    {
+      "epoch": 0.23424428332403793,
+      "grad_norm": 0.12366748753721642,
+      "learning_rate": 0.00018922113917441269,
+      "loss": 1.8516,
+      "step": 1890
+    },
+    {
+      "epoch": 0.23486397719526553,
+      "grad_norm": 0.1325347292455204,
+      "learning_rate": 0.00018912323044233684,
+      "loss": 1.8555,
+      "step": 1895
+    },
+    {
+      "epoch": 0.23548367106649315,
+      "grad_norm": 0.13177269918492968,
+      "learning_rate": 0.00018902490461271843,
+      "loss": 1.8777,
+      "step": 1900
+    },
+    {
+      "epoch": 0.23610336493772077,
+      "grad_norm": 0.13170172393384472,
+      "learning_rate": 0.00018892616214572319,
+      "loss": 1.8934,
+      "step": 1905
+    },
+    {
+      "epoch": 0.23672305880894837,
+      "grad_norm": 0.14106704771321893,
+      "learning_rate": 0.0001888270035034668,
+      "loss": 1.8449,
+      "step": 1910
+    },
+    {
+      "epoch": 0.237342752680176,
+      "grad_norm": 0.12621612288999542,
+      "learning_rate": 0.00018872742915001267,
+      "loss": 1.8184,
+      "step": 1915
+    },
+    {
+      "epoch": 0.2379624465514036,
+      "grad_norm": 0.13102027385319423,
+      "learning_rate": 0.00018862743955136966,
+      "loss": 1.8348,
+      "step": 1920
+    },
+    {
+      "epoch": 0.23858214042263123,
+      "grad_norm": 0.12924449635578353,
+      "learning_rate": 0.00018852703517549,
+      "loss": 1.8789,
+      "step": 1925
+    },
+    {
+      "epoch": 0.23920183429385883,
+      "grad_norm": 0.137053445230958,
+      "learning_rate": 0.00018842621649226712,
+      "loss": 1.8961,
+      "step": 1930
+    },
+    {
+      "epoch": 0.23982152816508645,
+      "grad_norm": 0.13676238584015243,
+      "learning_rate": 0.00018832498397353337,
+      "loss": 1.9055,
+      "step": 1935
+    },
+    {
+      "epoch": 0.24044122203631407,
+      "grad_norm": 0.12473045166890744,
+      "learning_rate": 0.0001882233380930579,
+      "loss": 1.9051,
+      "step": 1940
+    },
+    {
+      "epoch": 0.24106091590754167,
+      "grad_norm": 0.13737094149660017,
+      "learning_rate": 0.00018812127932654437,
+      "loss": 1.8383,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2416806097787693,
+      "grad_norm": 0.13541763846499583,
+      "learning_rate": 0.00018801880815162873,
+      "loss": 1.8684,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2423003036499969,
+      "grad_norm": 0.12777856593407638,
+      "learning_rate": 0.00018791592504787704,
+      "loss": 1.8383,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2429199975212245,
+      "grad_norm": 0.13129622175885042,
+      "learning_rate": 0.00018781263049678318,
+      "loss": 1.8508,
+      "step": 1960
+    },
+    {
+      "epoch": 0.24353969139245213,
+      "grad_norm": 0.12557864503351404,
+      "learning_rate": 0.00018770892498176658,
+      "loss": 1.9406,
+      "step": 1965
+    },
+    {
+      "epoch": 0.24415938526367975,
+      "grad_norm": 0.13946020278404306,
+      "learning_rate": 0.00018760480898817003,
+      "loss": 1.893,
+      "step": 1970
+    },
+    {
+      "epoch": 0.24477907913490735,
+      "grad_norm": 0.13062544520558794,
+      "learning_rate": 0.00018750028300325733,
+      "loss": 1.7953,
+      "step": 1975
+    },
+    {
+      "epoch": 0.24539877300613497,
+      "grad_norm": 0.1335190355390475,
+      "learning_rate": 0.0001873953475162111,
+      "loss": 1.8555,
+      "step": 1980
+    },
+    {
+      "epoch": 0.2460184668773626,
+      "grad_norm": 0.13163131758724117,
+      "learning_rate": 0.00018729000301813032,
+      "loss": 1.9035,
+      "step": 1985
+    },
+    {
+      "epoch": 0.24663816074859019,
+      "grad_norm": 0.1339228643234811,
+      "learning_rate": 0.00018718425000202826,
+      "loss": 1.8313,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2472578546198178,
+      "grad_norm": 0.1306228391266261,
+      "learning_rate": 0.00018707808896283,
+      "loss": 1.8941,
+      "step": 1995
+    },
+    {
+      "epoch": 0.24787754849104543,
+      "grad_norm": 0.12487620123598475,
+      "learning_rate": 0.00018697152039737018,
+      "loss": 1.7969,
+      "step": 2000
+    },
+    {
+      "epoch": 0.24849724236227302,
+      "grad_norm": 0.12511623039704756,
+      "learning_rate": 0.0001868645448043907,
+      "loss": 1.8188,
+      "step": 2005
+    },
+    {
+      "epoch": 0.24911693623350065,
+      "grad_norm": 0.1312732190979278,
+      "learning_rate": 0.00018675716268453827,
+      "loss": 1.832,
+      "step": 2010
+    },
+    {
+      "epoch": 0.24973663010472827,
+      "grad_norm": 0.1297564730601911,
+      "learning_rate": 0.00018664937454036226,
+      "loss": 1.8414,
+      "step": 2015
+    },
+    {
+      "epoch": 0.25035632397595586,
+      "grad_norm": 0.13336748960832828,
+      "learning_rate": 0.0001865411808763122,
+      "loss": 1.866,
+      "step": 2020
+    },
+    {
+      "epoch": 0.2509760178471835,
+      "grad_norm": 0.13371049555198566,
+      "learning_rate": 0.00018643258219873534,
+      "loss": 1.8648,
+      "step": 2025
+    },
+    {
+      "epoch": 0.2515957117184111,
+      "grad_norm": 0.12145110937533733,
+      "learning_rate": 0.00018632357901587456,
+      "loss": 1.8977,
+      "step": 2030
+    },
+    {
+      "epoch": 0.25221540558963873,
+      "grad_norm": 0.1273408283653277,
+      "learning_rate": 0.00018621417183786577,
+      "loss": 1.8953,
+      "step": 2035
+    },
+    {
+      "epoch": 0.25283509946086635,
+      "grad_norm": 0.1290600654953522,
+      "learning_rate": 0.00018610436117673555,
+      "loss": 1.8383,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2534547933320939,
+      "grad_norm": 0.1253131282820343,
+      "learning_rate": 0.00018599414754639883,
+      "loss": 1.8941,
+      "step": 2045
+    },
+    {
+      "epoch": 0.25407448720332154,
+      "grad_norm": 0.1300726181872789,
+      "learning_rate": 0.00018588353146265643,
+      "loss": 1.8547,
+      "step": 2050
+    },
+    {
+      "epoch": 0.25469418107454916,
+      "grad_norm": 0.12572572001230634,
+      "learning_rate": 0.0001857725134431926,
+      "loss": 1.8625,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2553138749457768,
+      "grad_norm": 0.12568185275742066,
+      "learning_rate": 0.00018566109400757272,
+      "loss": 1.8734,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2559335688170044,
+      "grad_norm": 0.13533271397949245,
+      "learning_rate": 0.0001855492736772408,
+      "loss": 1.8902,
+      "step": 2065
+    },
+    {
+      "epoch": 0.25655326268823203,
+      "grad_norm": 0.1327835673505901,
+      "learning_rate": 0.00018543705297551698,
+      "loss": 1.9043,
+      "step": 2070
+    },
+    {
+      "epoch": 0.25717295655945965,
+      "grad_norm": 0.1332233655568078,
+      "learning_rate": 0.0001853244324275952,
+      "loss": 1.8496,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2577926504306872,
+      "grad_norm": 0.12647398734735402,
+      "learning_rate": 0.00018521141256054067,
+      "loss": 1.8711,
+      "step": 2080
+    },
+    {
+      "epoch": 0.25841234430191484,
+      "grad_norm": 0.12370914252810508,
+      "learning_rate": 0.00018509799390328742,
+      "loss": 1.8402,
+      "step": 2085
+    },
+    {
+      "epoch": 0.25903203817314246,
+      "grad_norm": 0.1342541241421312,
+      "learning_rate": 0.00018498417698663584,
+      "loss": 1.809,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2596517320443701,
+      "grad_norm": 0.12945415515878775,
+      "learning_rate": 0.00018486996234325009,
+      "loss": 1.8363,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2602714259155977,
+      "grad_norm": 0.1208865775592005,
+      "learning_rate": 0.00018475535050765577,
+      "loss": 1.848,
+      "step": 2100
+    },
+    {
+      "epoch": 0.26089111978682533,
+      "grad_norm": 0.1227715928980261,
+      "learning_rate": 0.00018464034201623737,
+      "loss": 1.8215,
+      "step": 2105
+    },
+    {
+      "epoch": 0.2615108136580529,
+      "grad_norm": 0.12505462857949878,
+      "learning_rate": 0.00018452493740723567,
+      "loss": 1.8801,
+      "step": 2110
+    },
+    {
+      "epoch": 0.2621305075292805,
+      "grad_norm": 0.12835962262802786,
+      "learning_rate": 0.0001844091372207453,
+      "loss": 1.7945,
+      "step": 2115
+    },
+    {
+      "epoch": 0.26275020140050814,
+      "grad_norm": 0.12508804411904637,
+      "learning_rate": 0.00018429294199871218,
+      "loss": 1.8551,
+      "step": 2120
+    },
+    {
+      "epoch": 0.26336989527173577,
+      "grad_norm": 0.12702387355591513,
+      "learning_rate": 0.0001841763522849311,
+      "loss": 1.8238,
+      "step": 2125
+    },
+    {
+      "epoch": 0.2639895891429634,
+      "grad_norm": 0.1348850938489859,
+      "learning_rate": 0.00018405936862504293,
+      "loss": 1.8121,
+      "step": 2130
+    },
+    {
+      "epoch": 0.264609283014191,
+      "grad_norm": 0.1278684357058116,
+      "learning_rate": 0.00018394199156653233,
+      "loss": 1.8574,
+      "step": 2135
+    },
+    {
+      "epoch": 0.2652289768854186,
+      "grad_norm": 0.12483388812413534,
+      "learning_rate": 0.00018382422165872498,
+      "loss": 1.8504,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2658486707566462,
+      "grad_norm": 0.13515557229656505,
+      "learning_rate": 0.00018370605945278512,
+      "loss": 1.8609,
+      "step": 2145
+    },
+    {
+      "epoch": 0.2664683646278738,
+      "grad_norm": 0.13397417059997252,
+      "learning_rate": 0.00018358750550171303,
+      "loss": 1.9426,
+      "step": 2150
+    },
+    {
+      "epoch": 0.26708805849910144,
+      "grad_norm": 0.12551446472885072,
+      "learning_rate": 0.00018346856036034225,
+      "loss": 1.8531,
+      "step": 2155
+    },
+    {
+      "epoch": 0.26770775237032907,
+      "grad_norm": 0.12574015924446805,
+      "learning_rate": 0.0001833492245853371,
+      "loss": 1.857,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2683274462415567,
+      "grad_norm": 0.1301344380849854,
+      "learning_rate": 0.00018322949873519028,
+      "loss": 1.8902,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2689471401127843,
+      "grad_norm": 0.134107216212603,
+      "learning_rate": 0.00018310938337021967,
+      "loss": 1.8844,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2695668339840119,
+      "grad_norm": 0.13544984849232908,
+      "learning_rate": 0.00018298887905256642,
+      "loss": 1.8387,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2701865278552395,
+      "grad_norm": 0.13740107014915645,
+      "learning_rate": 0.00018286798634619178,
+      "loss": 1.8867,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2708062217264671,
+      "grad_norm": 0.13047734738410044,
+      "learning_rate": 0.0001827467058168748,
+      "loss": 1.8621,
+      "step": 2185
+    },
+    {
+      "epoch": 0.27142591559769474,
+      "grad_norm": 0.1275936167540542,
+      "learning_rate": 0.00018262503803220941,
+      "loss": 1.9242,
+      "step": 2190
+    },
+    {
+      "epoch": 0.27204560946892237,
+      "grad_norm": 0.12739893626922644,
+      "learning_rate": 0.00018250298356160203,
+      "loss": 1.8133,
+      "step": 2195
+    },
+    {
+      "epoch": 0.27266530334015,
+      "grad_norm": 0.13208684150469857,
+      "learning_rate": 0.00018238054297626868,
+      "loss": 1.8328,
+      "step": 2200
+    },
+    {
+      "epoch": 0.27328499721137756,
+      "grad_norm": 0.12642860309275045,
+      "learning_rate": 0.0001822577168492324,
+      "loss": 1.8848,
+      "step": 2205
+    },
+    {
+      "epoch": 0.2739046910826052,
+      "grad_norm": 0.1207837863995249,
+      "learning_rate": 0.00018213450575532068,
+      "loss": 1.898,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2745243849538328,
+      "grad_norm": 0.134513799642859,
+      "learning_rate": 0.0001820109102711625,
+      "loss": 1.8652,
+      "step": 2215
+    },
+    {
+      "epoch": 0.2751440788250604,
+      "grad_norm": 0.12522620018177075,
+      "learning_rate": 0.00018188693097518589,
+      "loss": 1.8902,
+      "step": 2220
+    },
+    {
+      "epoch": 0.27576377269628805,
+      "grad_norm": 0.12565243572374343,
+      "learning_rate": 0.00018176256844761515,
+      "loss": 1.8578,
+      "step": 2225
+    },
+    {
+      "epoch": 0.27638346656751567,
+      "grad_norm": 0.12338799943023959,
+      "learning_rate": 0.000181637823270468,
+      "loss": 1.9074,
+      "step": 2230
+    },
+    {
+      "epoch": 0.27700316043874323,
+      "grad_norm": 0.1313348559311083,
+      "learning_rate": 0.00018151269602755305,
+      "loss": 1.8352,
+      "step": 2235
+    },
+    {
+      "epoch": 0.27762285430997086,
+      "grad_norm": 0.1262516619013633,
+      "learning_rate": 0.00018138718730446694,
+      "loss": 1.8164,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2782425481811985,
+      "grad_norm": 0.13220071119654697,
+      "learning_rate": 0.00018126129768859166,
+      "loss": 1.8285,
+      "step": 2245
+    },
+    {
+      "epoch": 0.2788622420524261,
+      "grad_norm": 0.12746131878131237,
+      "learning_rate": 0.0001811350277690918,
+      "loss": 1.8668,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2794819359236537,
+      "grad_norm": 0.12708251481689406,
+      "learning_rate": 0.00018100837813691173,
+      "loss": 1.8996,
+      "step": 2255
+    },
+    {
+      "epoch": 0.28010162979488135,
+      "grad_norm": 0.13164815119644507,
+      "learning_rate": 0.00018088134938477285,
+      "loss": 1.9145,
+      "step": 2260
+    },
+    {
+      "epoch": 0.28072132366610897,
+      "grad_norm": 0.12382229396871405,
+      "learning_rate": 0.00018075394210717097,
+      "loss": 1.9527,
+      "step": 2265
+    },
+    {
+      "epoch": 0.28134101753733654,
+      "grad_norm": 0.13047982998936075,
+      "learning_rate": 0.0001806261569003733,
+      "loss": 1.8805,
+      "step": 2270
+    },
+    {
+      "epoch": 0.28196071140856416,
+      "grad_norm": 0.12580182752485483,
+      "learning_rate": 0.00018049799436241584,
+      "loss": 1.8184,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2825804052797918,
+      "grad_norm": 0.13179353346365458,
+      "learning_rate": 0.00018036945509310035,
+      "loss": 1.8332,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2832000991510194,
+      "grad_norm": 0.1490150630853317,
+      "learning_rate": 0.00018024053969399186,
+      "loss": 1.8203,
+      "step": 2285
+    },
+    {
+      "epoch": 0.283819793022247,
+      "grad_norm": 0.13210374436560907,
+      "learning_rate": 0.00018011124876841564,
+      "loss": 1.8578,
+      "step": 2290
+    },
+    {
+      "epoch": 0.28443948689347465,
+      "grad_norm": 0.13422338407254142,
+      "learning_rate": 0.0001799815829214544,
+      "loss": 1.8871,
+      "step": 2295
+    },
+    {
+      "epoch": 0.2850591807647022,
+      "grad_norm": 0.1297246090756236,
+      "learning_rate": 0.00017985154275994546,
+      "loss": 1.841,
+      "step": 2300
+    },
+    {
+      "epoch": 0.28567887463592984,
+      "grad_norm": 0.12536111156667865,
+      "learning_rate": 0.00017972112889247808,
+      "loss": 1.859,
+      "step": 2305
+    },
+    {
+      "epoch": 0.28629856850715746,
+      "grad_norm": 0.138340567721797,
+      "learning_rate": 0.00017959034192939027,
+      "loss": 1.7711,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2869182623783851,
+      "grad_norm": 0.13158620923757655,
+      "learning_rate": 0.0001794591824827663,
+      "loss": 1.8227,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2875379562496127,
+      "grad_norm": 0.1292868891126794,
+      "learning_rate": 0.0001793276511664335,
+      "loss": 1.8871,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2881576501208403,
+      "grad_norm": 0.12811495044307233,
+      "learning_rate": 0.00017919574859595977,
+      "loss": 1.8543,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2887773439920679,
+      "grad_norm": 0.12149407958829071,
+      "learning_rate": 0.00017906347538865021,
+      "loss": 1.8535,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2893970378632955,
+      "grad_norm": 0.1317042941594247,
+      "learning_rate": 0.00017893083216354477,
+      "loss": 1.8824,
+      "step": 2335
+    },
+    {
+      "epoch": 0.29001673173452314,
+      "grad_norm": 0.12569760233357527,
+      "learning_rate": 0.00017879781954141497,
+      "loss": 1.8234,
+      "step": 2340
+    },
+    {
+      "epoch": 0.29063642560575076,
+      "grad_norm": 0.13084498370382033,
+      "learning_rate": 0.00017866443814476107,
+      "loss": 1.852,
+      "step": 2345
+    },
+    {
+      "epoch": 0.2912561194769784,
+      "grad_norm": 0.13205282361051077,
+      "learning_rate": 0.0001785306885978092,
+      "loss": 1.8832,
+      "step": 2350
+    },
+    {
+      "epoch": 0.291875813348206,
+      "grad_norm": 0.12919651417069067,
+      "learning_rate": 0.00017839657152650856,
+      "loss": 1.8453,
+      "step": 2355
+    },
+    {
+      "epoch": 0.2924955072194336,
+      "grad_norm": 0.12973043338594262,
+      "learning_rate": 0.00017826208755852827,
+      "loss": 1.8797,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2931152010906612,
+      "grad_norm": 0.13301208195385272,
+      "learning_rate": 0.00017812723732325446,
+      "loss": 1.8641,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2937348949618888,
+      "grad_norm": 0.12949835549240182,
+      "learning_rate": 0.00017799202145178758,
+      "loss": 1.8406,
+      "step": 2370
+    },
+    {
+      "epoch": 0.29435458883311644,
+      "grad_norm": 0.1293697097422646,
+      "learning_rate": 0.00017785644057693913,
+      "loss": 1.8559,
+      "step": 2375
+    },
+    {
+      "epoch": 0.29497428270434406,
+      "grad_norm": 0.1520616458448482,
+      "learning_rate": 0.0001777204953332288,
+      "loss": 1.8887,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2955939765755717,
+      "grad_norm": 0.13380400311187673,
+      "learning_rate": 0.00017758418635688167,
+      "loss": 1.9051,
+      "step": 2385
+    },
+    {
+      "epoch": 0.2962136704467993,
+      "grad_norm": 0.12836449608199532,
+      "learning_rate": 0.00017744751428582496,
+      "loss": 1.8859,
+      "step": 2390
+    },
+    {
+      "epoch": 0.29683336431802687,
+      "grad_norm": 0.13080261256263226,
+      "learning_rate": 0.00017731047975968523,
+      "loss": 1.8539,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2974530581892545,
+      "grad_norm": 0.12988299510465,
+      "learning_rate": 0.00017717308341978538,
+      "loss": 1.8566,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2980727520604821,
+      "grad_norm": 0.125683224518266,
+      "learning_rate": 0.00017703532590914147,
+      "loss": 1.8676,
+      "step": 2405
+    },
+    {
+      "epoch": 0.29869244593170974,
+      "grad_norm": 0.13352496446514742,
+      "learning_rate": 0.00017689720787245997,
+      "loss": 1.923,
+      "step": 2410
+    },
+    {
+      "epoch": 0.29931213980293736,
+      "grad_norm": 0.1293774171094288,
+      "learning_rate": 0.00017675872995613458,
+      "loss": 1.868,
+      "step": 2415
+    },
+    {
+      "epoch": 0.299931833674165,
+      "grad_norm": 0.12527135631615616,
+      "learning_rate": 0.0001766198928082432,
+      "loss": 1.8621,
+      "step": 2420
+    },
+    {
+      "epoch": 0.30055152754539255,
+      "grad_norm": 0.12613717510254913,
+      "learning_rate": 0.00017648069707854497,
+      "loss": 1.9,
+      "step": 2425
+    },
+    {
+      "epoch": 0.30117122141662017,
+      "grad_norm": 0.12391342907752452,
+      "learning_rate": 0.0001763411434184772,
+      "loss": 1.8531,
+      "step": 2430
+    },
+    {
+      "epoch": 0.3017909152878478,
+      "grad_norm": 0.13123817620567785,
+      "learning_rate": 0.00017620123248115235,
+      "loss": 1.8664,
+      "step": 2435
+    },
+    {
+      "epoch": 0.3024106091590754,
+      "grad_norm": 0.12780408489494227,
+      "learning_rate": 0.0001760609649213548,
+      "loss": 1.8922,
+      "step": 2440
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 0.12077509717420204,
+      "learning_rate": 0.00017592034139553812,
+      "loss": 1.9,
+      "step": 2445
+    },
+    {
+      "epoch": 0.30364999690153066,
+      "grad_norm": 0.12597036196005937,
+      "learning_rate": 0.00017577936256182167,
+      "loss": 1.8551,
+      "step": 2450
+    },
+    {
+      "epoch": 0.3042696907727583,
+      "grad_norm": 0.13100450560697094,
+      "learning_rate": 0.00017563802907998773,
+      "loss": 1.8453,
+      "step": 2455
+    },
+    {
+      "epoch": 0.30488938464398585,
+      "grad_norm": 0.128806698790364,
+      "learning_rate": 0.00017549634161147823,
+      "loss": 1.8484,
+      "step": 2460
+    },
+    {
+      "epoch": 0.3055090785152135,
+      "grad_norm": 0.12628216846498763,
+      "learning_rate": 0.0001753543008193919,
+      "loss": 1.882,
+      "step": 2465
+    },
+    {
+      "epoch": 0.3061287723864411,
+      "grad_norm": 0.12839071253465598,
+      "learning_rate": 0.00017521190736848096,
+      "loss": 1.8957,
+      "step": 2470
+    },
+    {
+      "epoch": 0.3067484662576687,
+      "grad_norm": 0.12646846306076212,
+      "learning_rate": 0.00017506916192514801,
+      "loss": 1.8148,
+      "step": 2475
+    },
+    {
+      "epoch": 0.30736816012889634,
+      "grad_norm": 0.1300463966004593,
+      "learning_rate": 0.0001749260651574431,
+      "loss": 1.8625,
+      "step": 2480
+    },
+    {
+      "epoch": 0.30798785400012396,
+      "grad_norm": 0.13684852530638011,
+      "learning_rate": 0.00017478261773506043,
+      "loss": 1.8324,
+      "step": 2485
+    },
+    {
+      "epoch": 0.30860754787135153,
+      "grad_norm": 0.1333149939890643,
+      "learning_rate": 0.00017463882032933524,
+      "loss": 1.8238,
+      "step": 2490
+    },
+    {
+      "epoch": 0.30922724174257915,
+      "grad_norm": 0.13250928441732007,
+      "learning_rate": 0.00017449467361324076,
+      "loss": 1.8562,
+      "step": 2495
+    },
+    {
+      "epoch": 0.3098469356138068,
+      "grad_norm": 0.12650581345903503,
+      "learning_rate": 0.0001743501782613849,
+      "loss": 1.918,
+      "step": 2500
+    },
+    {
+      "epoch": 0.3104666294850344,
+      "grad_norm": 0.13012679795096624,
+      "learning_rate": 0.00017420533495000727,
+      "loss": 1.8469,
+      "step": 2505
+    },
+    {
+      "epoch": 0.311086323356262,
+      "grad_norm": 0.1297240234416192,
+      "learning_rate": 0.0001740601443569759,
+      "loss": 1.8047,
+      "step": 2510
+    },
+    {
+      "epoch": 0.31170601722748964,
+      "grad_norm": 0.1309817865465321,
+      "learning_rate": 0.0001739146071617841,
+      "loss": 1.8641,
+      "step": 2515
+    },
+    {
+      "epoch": 0.3123257110987172,
+      "grad_norm": 0.12922856903591073,
+      "learning_rate": 0.0001737687240455473,
+      "loss": 1.834,
+      "step": 2520
+    },
+    {
+      "epoch": 0.31294540496994483,
+      "grad_norm": 0.12557065472012063,
+      "learning_rate": 0.00017362249569099982,
+      "loss": 1.8438,
+      "step": 2525
+    },
+    {
+      "epoch": 0.31356509884117245,
+      "grad_norm": 0.13353066027905686,
+      "learning_rate": 0.00017347592278249175,
+      "loss": 1.8586,
+      "step": 2530
+    },
+    {
+      "epoch": 0.3141847927124001,
+      "grad_norm": 0.13316269753092513,
+      "learning_rate": 0.00017332900600598562,
+      "loss": 1.8707,
+      "step": 2535
+    },
+    {
+      "epoch": 0.3148044865836277,
+      "grad_norm": 0.13353296735977976,
+      "learning_rate": 0.00017318174604905327,
+      "loss": 1.8836,
+      "step": 2540
+    },
+    {
+      "epoch": 0.3154241804548553,
+      "grad_norm": 0.12858939196532748,
+      "learning_rate": 0.00017303414360087278,
+      "loss": 1.825,
+      "step": 2545
+    },
+    {
+      "epoch": 0.31604387432608294,
+      "grad_norm": 0.13179975498557261,
+      "learning_rate": 0.00017288619935222486,
+      "loss": 1.8516,
+      "step": 2550
+    },
+    {
+      "epoch": 0.3166635681973105,
+      "grad_norm": 0.12482617396470475,
+      "learning_rate": 0.00017273791399548998,
+      "loss": 1.8391,
+      "step": 2555
+    },
+    {
+      "epoch": 0.31728326206853813,
+      "grad_norm": 0.13139623735524517,
+      "learning_rate": 0.000172589288224645,
+      "loss": 1.8605,
+      "step": 2560
+    },
+    {
+      "epoch": 0.31790295593976575,
+      "grad_norm": 0.1310695530113606,
+      "learning_rate": 0.00017244032273525995,
+      "loss": 1.8414,
+      "step": 2565
+    },
+    {
+      "epoch": 0.3185226498109934,
+      "grad_norm": 0.12710157216789886,
+      "learning_rate": 0.0001722910182244946,
+      "loss": 1.8543,
+      "step": 2570
+    },
+    {
+      "epoch": 0.319142343682221,
+      "grad_norm": 0.12590627027340592,
+      "learning_rate": 0.00017214137539109552,
+      "loss": 1.8793,
+      "step": 2575
+    },
+    {
+      "epoch": 0.3197620375534486,
+      "grad_norm": 0.13436407452579877,
+      "learning_rate": 0.0001719913949353925,
+      "loss": 1.8652,
+      "step": 2580
+    },
+    {
+      "epoch": 0.3203817314246762,
+      "grad_norm": 0.12643462648025297,
+      "learning_rate": 0.0001718410775592955,
+      "loss": 1.8992,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3210014252959038,
+      "grad_norm": 0.1305952405164263,
+      "learning_rate": 0.00017169042396629117,
+      "loss": 1.848,
+      "step": 2590
+    },
+    {
+      "epoch": 0.32162111916713143,
+      "grad_norm": 0.150091829244743,
+      "learning_rate": 0.00017153943486143978,
+      "loss": 1.8566,
+      "step": 2595
+    },
+    {
+      "epoch": 0.32224081303835905,
+      "grad_norm": 0.1254009301014799,
+      "learning_rate": 0.00017138811095137175,
+      "loss": 1.8629,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3228605069095867,
+      "grad_norm": 0.13664056730838736,
+      "learning_rate": 0.0001712364529442843,
+      "loss": 1.8066,
+      "step": 2605
+    },
+    {
+      "epoch": 0.3234802007808143,
+      "grad_norm": 0.1335954880543391,
+      "learning_rate": 0.00017108446154993838,
+      "loss": 1.868,
+      "step": 2610
+    },
+    {
+      "epoch": 0.32409989465204186,
+      "grad_norm": 0.1306781290863923,
+      "learning_rate": 0.0001709321374796551,
+      "loss": 1.8758,
+      "step": 2615
+    },
+    {
+      "epoch": 0.3247195885232695,
+      "grad_norm": 0.14742493532611176,
+      "learning_rate": 0.00017077948144631248,
+      "loss": 1.8562,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3253392823944971,
+      "grad_norm": 0.12846260878216087,
+      "learning_rate": 0.00017062649416434223,
+      "loss": 1.8352,
+      "step": 2625
+    },
+    {
+      "epoch": 0.32595897626572473,
+      "grad_norm": 0.13165013316462518,
+      "learning_rate": 0.00017047317634972617,
+      "loss": 1.8586,
+      "step": 2630
+    },
+    {
+      "epoch": 0.32657867013695235,
+      "grad_norm": 0.13366541465113413,
+      "learning_rate": 0.00017031952871999315,
+      "loss": 1.9242,
+      "step": 2635
+    },
+    {
+      "epoch": 0.32719836400818,
+      "grad_norm": 0.13286362925121087,
+      "learning_rate": 0.0001701655519942155,
+      "loss": 1.8793,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3278180578794076,
+      "grad_norm": 0.13827798198391242,
+      "learning_rate": 0.00017001124689300568,
+      "loss": 1.8211,
+      "step": 2645
+    },
+    {
+      "epoch": 0.32843775175063517,
+      "grad_norm": 0.13443963143632734,
+      "learning_rate": 0.00016985661413851304,
+      "loss": 1.8512,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3290574456218628,
+      "grad_norm": 0.12931297062470157,
+      "learning_rate": 0.00016970165445442023,
+      "loss": 1.8586,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3296771394930904,
+      "grad_norm": 0.12783732546786497,
+      "learning_rate": 0.00016954636856594005,
+      "loss": 1.8367,
+      "step": 2660
+    },
+    {
+      "epoch": 0.33029683336431803,
+      "grad_norm": 0.12212361016917726,
+      "learning_rate": 0.00016939075719981194,
+      "loss": 1.8582,
+      "step": 2665
+    },
+    {
+      "epoch": 0.33091652723554565,
+      "grad_norm": 0.1294327845651695,
+      "learning_rate": 0.00016923482108429844,
+      "loss": 1.8465,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3315362211067733,
+      "grad_norm": 0.13189238283540644,
+      "learning_rate": 0.00016907856094918207,
+      "loss": 1.8051,
+      "step": 2675
+    },
+    {
+      "epoch": 0.33215591497800084,
+      "grad_norm": 0.1276135033870951,
+      "learning_rate": 0.0001689219775257617,
+      "loss": 1.8199,
+      "step": 2680
+    },
+    {
+      "epoch": 0.33277560884922847,
+      "grad_norm": 0.12275325328072717,
+      "learning_rate": 0.00016876507154684918,
+      "loss": 1.8047,
+      "step": 2685
+    },
+    {
+      "epoch": 0.3333953027204561,
+      "grad_norm": 0.13012589500787683,
+      "learning_rate": 0.00016860784374676593,
+      "loss": 1.8621,
+      "step": 2690
+    },
+    {
+      "epoch": 0.3340149965916837,
+      "grad_norm": 0.13152129389683137,
+      "learning_rate": 0.00016845029486133956,
+      "loss": 1.8516,
+      "step": 2695
+    },
+    {
+      "epoch": 0.33463469046291133,
+      "grad_norm": 0.12631329142384912,
+      "learning_rate": 0.00016829242562790026,
+      "loss": 1.8387,
+      "step": 2700
+    },
+    {
+      "epoch": 0.33525438433413896,
+      "grad_norm": 0.13018878722373028,
+      "learning_rate": 0.00016813423678527754,
+      "loss": 1.923,
+      "step": 2705
+    },
+    {
+      "epoch": 0.3358740782053665,
+      "grad_norm": 0.14109498005999074,
+      "learning_rate": 0.00016797572907379667,
+      "loss": 1.8469,
+      "step": 2710
+    },
+    {
+      "epoch": 0.33649377207659414,
+      "grad_norm": 0.13811693482252332,
+      "learning_rate": 0.00016781690323527511,
+      "loss": 1.8676,
+      "step": 2715
+    },
+    {
+      "epoch": 0.33711346594782177,
+      "grad_norm": 0.1291292771360962,
+      "learning_rate": 0.00016765776001301933,
+      "loss": 1.8566,
+      "step": 2720
+    },
+    {
+      "epoch": 0.3377331598190494,
+      "grad_norm": 0.13823691210457728,
+      "learning_rate": 0.00016749830015182107,
+      "loss": 1.9406,
+      "step": 2725
+    },
+    {
+      "epoch": 0.338352853690277,
+      "grad_norm": 0.13089413884817042,
+      "learning_rate": 0.00016733852439795394,
+      "loss": 1.8734,
+      "step": 2730
+    },
+    {
+      "epoch": 0.33897254756150463,
+      "grad_norm": 0.14018900588379418,
+      "learning_rate": 0.00016717843349916994,
+      "loss": 1.8359,
+      "step": 2735
+    },
+    {
+      "epoch": 0.33959224143273226,
+      "grad_norm": 0.12736041535496453,
+      "learning_rate": 0.000167018028204696,
+      "loss": 1.8742,
+      "step": 2740
+    },
+    {
+      "epoch": 0.3402119353039598,
+      "grad_norm": 0.12533414989711167,
+      "learning_rate": 0.0001668573092652303,
+      "loss": 1.8648,
+      "step": 2745
+    },
+    {
+      "epoch": 0.34083162917518744,
+      "grad_norm": 0.13957662916108884,
+      "learning_rate": 0.00016669627743293907,
+      "loss": 1.9223,
+      "step": 2750
+    },
+    {
+      "epoch": 0.34145132304641507,
+      "grad_norm": 0.13354411893886944,
+      "learning_rate": 0.00016653493346145267,
+      "loss": 1.8301,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3420710169176427,
+      "grad_norm": 0.13557566100086996,
+      "learning_rate": 0.00016637327810586246,
+      "loss": 1.8387,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3426907107888703,
+      "grad_norm": 0.12737311407019075,
+      "learning_rate": 0.00016621131212271695,
+      "loss": 1.832,
+      "step": 2765
+    },
+    {
+      "epoch": 0.34331040466009793,
+      "grad_norm": 0.13905113414944018,
+      "learning_rate": 0.00016604903627001844,
+      "loss": 1.8586,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3439300985313255,
+      "grad_norm": 0.13141740905832047,
+      "learning_rate": 0.00016588645130721948,
+      "loss": 1.9059,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3445497924025531,
+      "grad_norm": 0.1292375170284331,
+      "learning_rate": 0.00016572355799521912,
+      "loss": 1.8629,
+      "step": 2780
+    },
+    {
+      "epoch": 0.34516948627378075,
+      "grad_norm": 0.13097804366675025,
+      "learning_rate": 0.0001655603570963596,
+      "loss": 1.8402,
+      "step": 2785
+    },
+    {
+      "epoch": 0.34578918014500837,
+      "grad_norm": 0.12857874757255375,
+      "learning_rate": 0.00016539684937442263,
+      "loss": 1.8871,
+      "step": 2790
+    },
+    {
+      "epoch": 0.346408874016236,
+      "grad_norm": 0.12786886162667369,
+      "learning_rate": 0.0001652330355946259,
+      "loss": 1.8047,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3470285678874636,
+      "grad_norm": 0.143616962351737,
+      "learning_rate": 0.00016506891652361933,
+      "loss": 1.8383,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3476482617586912,
+      "grad_norm": 0.13453451476859302,
+      "learning_rate": 0.0001649044929294818,
+      "loss": 1.8516,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3482679556299188,
+      "grad_norm": 0.13515198281229954,
+      "learning_rate": 0.00016473976558171714,
+      "loss": 1.9059,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3488876495011464,
+      "grad_norm": 0.13811663765869722,
+      "learning_rate": 0.00016457473525125093,
+      "loss": 1.8078,
+      "step": 2815
+    },
+    {
+      "epoch": 0.34950734337237405,
+      "grad_norm": 0.12993511623884146,
+      "learning_rate": 0.00016440940271042663,
+      "loss": 1.8887,
+      "step": 2820
+    },
+    {
+      "epoch": 0.35012703724360167,
+      "grad_norm": 0.1373771927813058,
+      "learning_rate": 0.00016424376873300207,
+      "loss": 1.8465,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3507467311148293,
+      "grad_norm": 0.1330128701703067,
+      "learning_rate": 0.00016407783409414577,
+      "loss": 1.8609,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3513664249860569,
+      "grad_norm": 0.11946071667616472,
+      "learning_rate": 0.00016391159957043335,
+      "loss": 1.8535,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3519861188572845,
+      "grad_norm": 0.1358363297173592,
+      "learning_rate": 0.000163745065939844,
+      "loss": 1.8699,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3526058127285121,
+      "grad_norm": 0.128552153238772,
+      "learning_rate": 0.0001635782339817566,
+      "loss": 1.9055,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3532255065997397,
+      "grad_norm": 0.13106197892945834,
+      "learning_rate": 0.00016341110447694624,
+      "loss": 1.827,
+      "step": 2850
+    },
+    {
+      "epoch": 0.35384520047096735,
+      "grad_norm": 0.14036033306901916,
+      "learning_rate": 0.00016324367820758057,
+      "loss": 1.8266,
+      "step": 2855
+    },
+    {
+      "epoch": 0.35446489434219497,
+      "grad_norm": 0.12890303821028082,
+      "learning_rate": 0.0001630759559572161,
+      "loss": 1.8477,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3550845882134226,
+      "grad_norm": 0.13118051072317582,
+      "learning_rate": 0.00016290793851079447,
+      "loss": 1.8879,
+      "step": 2865
+    },
+    {
+      "epoch": 0.35570428208465016,
+      "grad_norm": 0.12771050153008334,
+      "learning_rate": 0.00016273962665463892,
+      "loss": 1.8883,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3563239759558778,
+      "grad_norm": 0.12557326568723504,
+      "learning_rate": 0.00016257102117645048,
+      "loss": 1.8121,
+      "step": 2875
+    },
+    {
+      "epoch": 0.3569436698271054,
+      "grad_norm": 0.1317338184540393,
+      "learning_rate": 0.00016240212286530432,
+      "loss": 1.8645,
+      "step": 2880
+    },
+    {
+      "epoch": 0.357563363698333,
+      "grad_norm": 0.12301737660160941,
+      "learning_rate": 0.00016223293251164616,
+      "loss": 1.8715,
+      "step": 2885
+    },
+    {
+      "epoch": 0.35818305756956065,
+      "grad_norm": 0.127922605349782,
+      "learning_rate": 0.00016206345090728834,
+      "loss": 1.8387,
+      "step": 2890
+    },
+    {
+      "epoch": 0.35880275144078827,
+      "grad_norm": 0.1294192610668883,
+      "learning_rate": 0.00016189367884540638,
+      "loss": 1.848,
+      "step": 2895
+    },
+    {
+      "epoch": 0.35942244531201584,
+      "grad_norm": 0.13398944968217794,
+      "learning_rate": 0.00016172361712053513,
+      "loss": 1.8586,
+      "step": 2900
+    },
+    {
+      "epoch": 0.36004213918324346,
+      "grad_norm": 0.12899074210444184,
+      "learning_rate": 0.00016155326652856497,
+      "loss": 1.8805,
+      "step": 2905
+    },
+    {
+      "epoch": 0.3606618330544711,
+      "grad_norm": 0.12795186731054056,
+      "learning_rate": 0.0001613826278667383,
+      "loss": 1.7922,
+      "step": 2910
+    },
+    {
+      "epoch": 0.3612815269256987,
+      "grad_norm": 0.1292195667193673,
+      "learning_rate": 0.00016121170193364557,
+      "loss": 1.816,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3619012207969263,
+      "grad_norm": 0.1242942240620901,
+      "learning_rate": 0.0001610404895292218,
+      "loss": 1.9125,
+      "step": 2920
+    },
+    {
+      "epoch": 0.36252091466815395,
+      "grad_norm": 0.13101097061222353,
+      "learning_rate": 0.0001608689914547426,
+      "loss": 1.8641,
+      "step": 2925
+    },
+    {
+      "epoch": 0.36314060853938157,
+      "grad_norm": 0.12617646446429898,
+      "learning_rate": 0.00016069720851282052,
+      "loss": 1.8461,
+      "step": 2930
+    },
+    {
+      "epoch": 0.36376030241060914,
+      "grad_norm": 0.12484378328833615,
+      "learning_rate": 0.00016052514150740135,
+      "loss": 1.8918,
+      "step": 2935
+    },
+    {
+      "epoch": 0.36437999628183676,
+      "grad_norm": 0.128015270056116,
+      "learning_rate": 0.00016035279124376026,
+      "loss": 1.8203,
+      "step": 2940
+    },
+    {
+      "epoch": 0.3649996901530644,
+      "grad_norm": 0.12690248111261712,
+      "learning_rate": 0.00016018015852849806,
+      "loss": 1.8211,
+      "step": 2945
+    },
+    {
+      "epoch": 0.365619384024292,
+      "grad_norm": 0.13279636029030822,
+      "learning_rate": 0.00016000724416953744,
+      "loss": 1.8551,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3662390778955196,
+      "grad_norm": 0.13480615224006273,
+      "learning_rate": 0.00015983404897611928,
+      "loss": 1.8484,
+      "step": 2955
+    },
+    {
+      "epoch": 0.36685877176674725,
+      "grad_norm": 0.12954165767486733,
+      "learning_rate": 0.00015966057375879858,
+      "loss": 1.8434,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3674784656379748,
+      "grad_norm": 0.13031965865973283,
+      "learning_rate": 0.00015948681932944104,
+      "loss": 1.8582,
+      "step": 2965
+    },
+    {
+      "epoch": 0.36809815950920244,
+      "grad_norm": 0.13110591784796458,
+      "learning_rate": 0.00015931278650121897,
+      "loss": 1.866,
+      "step": 2970
+    },
+    {
+      "epoch": 0.36871785338043006,
+      "grad_norm": 0.13182197561526388,
+      "learning_rate": 0.0001591384760886076,
+      "loss": 1.8301,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3693375472516577,
+      "grad_norm": 0.13143677143578264,
+      "learning_rate": 0.00015896388890738127,
+      "loss": 1.8531,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3699572411228853,
+      "grad_norm": 0.13715925277059812,
+      "learning_rate": 0.00015878902577460963,
+      "loss": 1.8344,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3705769349941129,
+      "grad_norm": 0.13007938351554157,
+      "learning_rate": 0.00015861388750865375,
+      "loss": 1.8637,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3711966288653405,
+      "grad_norm": 0.13870936181867594,
+      "learning_rate": 0.00015843847492916235,
+      "loss": 1.8594,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3718163227365681,
+      "grad_norm": 0.1338434765895186,
+      "learning_rate": 0.00015826278885706788,
+      "loss": 1.8293,
+      "step": 3000
+    },
+    {
+      "epoch": 0.37243601660779574,
+      "grad_norm": 0.13035091608618987,
+      "learning_rate": 0.0001580868301145828,
+      "loss": 1.8566,
+      "step": 3005
+    },
+    {
+      "epoch": 0.37305571047902336,
+      "grad_norm": 0.14555147420902945,
+      "learning_rate": 0.00015791059952519567,
+      "loss": 1.8672,
+      "step": 3010
+    },
+    {
+      "epoch": 0.373675404350251,
+      "grad_norm": 0.1295782115936669,
+      "learning_rate": 0.00015773409791366728,
+      "loss": 1.8598,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3742950982214786,
+      "grad_norm": 0.13140008433844813,
+      "learning_rate": 0.00015755732610602677,
+      "loss": 1.8836,
+      "step": 3020
+    },
+    {
+      "epoch": 0.37491479209270623,
+      "grad_norm": 0.12659202187281082,
+      "learning_rate": 0.00015738028492956786,
+      "loss": 1.8406,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3755344859639338,
+      "grad_norm": 0.13023530855829013,
+      "learning_rate": 0.0001572029752128449,
+      "loss": 1.8645,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3761541798351614,
+      "grad_norm": 0.13142420798148008,
+      "learning_rate": 0.00015702539778566897,
+      "loss": 1.8742,
+      "step": 3035
+    },
+    {
+      "epoch": 0.37677387370638904,
+      "grad_norm": 0.13011574441351573,
+      "learning_rate": 0.0001568475534791041,
+      "loss": 1.8105,
+      "step": 3040
+    },
+    {
+      "epoch": 0.37739356757761666,
+      "grad_norm": 0.1385944738745723,
+      "learning_rate": 0.00015666944312546328,
+      "loss": 1.8242,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3780132614488443,
+      "grad_norm": 0.13206170515775578,
+      "learning_rate": 0.0001564910675583046,
+      "loss": 1.8781,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3786329553200719,
+      "grad_norm": 0.1272155538752064,
+      "learning_rate": 0.0001563124276124274,
+      "loss": 1.8844,
+      "step": 3055
+    },
+    {
+      "epoch": 0.3792526491912995,
+      "grad_norm": 0.12879360712674345,
+      "learning_rate": 0.00015613352412386825,
+      "loss": 1.8016,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3798723430625271,
+      "grad_norm": 0.13141621693892783,
+      "learning_rate": 0.00015595435792989718,
+      "loss": 1.866,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3804920369337547,
+      "grad_norm": 0.13148734211884924,
+      "learning_rate": 0.0001557749298690135,
+      "loss": 1.8859,
+      "step": 3070
+    },
+    {
+      "epoch": 0.38111173080498234,
+      "grad_norm": 0.1233370862820081,
+      "learning_rate": 0.00015559524078094235,
+      "loss": 1.8398,
+      "step": 3075
+    },
+    {
+      "epoch": 0.38173142467620996,
+      "grad_norm": 0.1343298831070218,
+      "learning_rate": 0.00015541529150663022,
+      "loss": 1.8273,
+      "step": 3080
+    },
+    {
+      "epoch": 0.3823511185474376,
+      "grad_norm": 0.1350754368311016,
+      "learning_rate": 0.00015523508288824145,
+      "loss": 1.8598,
+      "step": 3085
+    },
+    {
+      "epoch": 0.38297081241866515,
+      "grad_norm": 0.13186218005747855,
+      "learning_rate": 0.00015505461576915402,
+      "loss": 1.8434,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3835905062898928,
+      "grad_norm": 0.1256587528049024,
+      "learning_rate": 0.00015487389099395565,
+      "loss": 1.868,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3842102001611204,
+      "grad_norm": 0.13460533409016884,
+      "learning_rate": 0.00015469290940844005,
+      "loss": 1.8883,
+      "step": 3100
+    },
+    {
+      "epoch": 0.384829894032348,
+      "grad_norm": 0.13262872301234474,
+      "learning_rate": 0.00015451167185960267,
+      "loss": 1.8703,
+      "step": 3105
+    },
+    {
+      "epoch": 0.38544958790357564,
+      "grad_norm": 0.1307836877374553,
+      "learning_rate": 0.00015433017919563692,
+      "loss": 1.8898,
+      "step": 3110
+    },
+    {
+      "epoch": 0.38606928177480326,
+      "grad_norm": 0.13241485557865607,
+      "learning_rate": 0.00015414843226593016,
+      "loss": 1.8715,
+      "step": 3115
+    },
+    {
+      "epoch": 0.3866889756460309,
+      "grad_norm": 0.12308084527810047,
+      "learning_rate": 0.0001539664319210597,
+      "loss": 1.8891,
+      "step": 3120
+    },
+    {
+      "epoch": 0.38730866951725845,
+      "grad_norm": 0.12579495949571556,
+      "learning_rate": 0.0001537841790127888,
+      "loss": 1.891,
+      "step": 3125
+    },
+    {
+      "epoch": 0.3879283633884861,
+      "grad_norm": 0.13544825188327536,
+      "learning_rate": 0.00015360167439406274,
+      "loss": 1.8902,
+      "step": 3130
+    },
+    {
+      "epoch": 0.3885480572597137,
+      "grad_norm": 0.1367382298904419,
+      "learning_rate": 0.00015341891891900494,
+      "loss": 1.8469,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3891677511309413,
+      "grad_norm": 0.1281769119603224,
+      "learning_rate": 0.00015323591344291258,
+      "loss": 1.8297,
+      "step": 3140
+    },
+    {
+      "epoch": 0.38978744500216894,
+      "grad_norm": 0.1295407756633459,
+      "learning_rate": 0.00015305265882225303,
+      "loss": 1.8543,
+      "step": 3145
+    },
+    {
+      "epoch": 0.39040713887339656,
+      "grad_norm": 0.12523656961779545,
+      "learning_rate": 0.00015286915591465969,
+      "loss": 1.8477,
+      "step": 3150
+    },
+    {
+      "epoch": 0.39102683274462413,
+      "grad_norm": 0.14659564859622612,
+      "learning_rate": 0.00015268540557892773,
+      "loss": 1.8566,
+      "step": 3155
+    },
+    {
+      "epoch": 0.39164652661585175,
+      "grad_norm": 0.12755016149566428,
+      "learning_rate": 0.0001525014086750105,
+      "loss": 1.8641,
+      "step": 3160
+    },
+    {
+      "epoch": 0.3922662204870794,
+      "grad_norm": 0.12426211762458562,
+      "learning_rate": 0.0001523171660640152,
+      "loss": 1.8516,
+      "step": 3165
+    },
+    {
+      "epoch": 0.392885914358307,
+      "grad_norm": 0.13100509076688815,
+      "learning_rate": 0.00015213267860819896,
+      "loss": 1.8355,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3935056082295346,
+      "grad_norm": 0.13166122634442548,
+      "learning_rate": 0.00015194794717096475,
+      "loss": 1.8535,
+      "step": 3175
+    },
+    {
+      "epoch": 0.39412530210076224,
+      "grad_norm": 0.12851391806205498,
+      "learning_rate": 0.00015176297261685742,
+      "loss": 1.8531,
+      "step": 3180
+    },
+    {
+      "epoch": 0.3947449959719898,
+      "grad_norm": 0.13014329204831934,
+      "learning_rate": 0.00015157775581155957,
+      "loss": 1.8949,
+      "step": 3185
+    },
+    {
+      "epoch": 0.39536468984321743,
+      "grad_norm": 0.12402746941996257,
+      "learning_rate": 0.00015139229762188761,
+      "loss": 1.8113,
+      "step": 3190
+    },
+    {
+      "epoch": 0.39598438371444505,
+      "grad_norm": 0.1340147610045317,
+      "learning_rate": 0.00015120659891578754,
+      "loss": 1.7973,
+      "step": 3195
+    },
+    {
+      "epoch": 0.3966040775856727,
+      "grad_norm": 0.13126263396327228,
+      "learning_rate": 0.00015102066056233104,
+      "loss": 1.8086,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3972237714569003,
+      "grad_norm": 0.1254442119538948,
+      "learning_rate": 0.00015083448343171138,
+      "loss": 1.8617,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3978434653281279,
+      "grad_norm": 0.13193426597795005,
+      "learning_rate": 0.00015064806839523915,
+      "loss": 1.8191,
+      "step": 3210
+    },
+    {
+      "epoch": 0.39846315919935554,
+      "grad_norm": 0.13785220644107396,
+      "learning_rate": 0.00015046141632533844,
+      "loss": 1.7977,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3990828530705831,
+      "grad_norm": 0.13241964464110698,
+      "learning_rate": 0.0001502745280955428,
+      "loss": 1.882,
+      "step": 3220
+    },
+    {
+      "epoch": 0.39970254694181073,
+      "grad_norm": 0.13524242845383544,
+      "learning_rate": 0.00015008740458049075,
+      "loss": 1.8496,
+      "step": 3225
+    },
+    {
+      "epoch": 0.40032224081303835,
+      "grad_norm": 0.1292830611807527,
+      "learning_rate": 0.0001499000466559221,
+      "loss": 1.8172,
+      "step": 3230
+    },
+    {
+      "epoch": 0.400941934684266,
+      "grad_norm": 0.12650175037356148,
+      "learning_rate": 0.0001497124551986737,
+      "loss": 1.8625,
+      "step": 3235
+    },
+    {
+      "epoch": 0.4015616285554936,
+      "grad_norm": 0.13556994409473824,
+      "learning_rate": 0.00014952463108667527,
+      "loss": 1.8828,
+      "step": 3240
+    },
+    {
+      "epoch": 0.4021813224267212,
+      "grad_norm": 0.13166998058062032,
+      "learning_rate": 0.0001493365751989454,
+      "loss": 1.832,
+      "step": 3245
+    },
+    {
+      "epoch": 0.4028010162979488,
+      "grad_norm": 0.13449713435485267,
+      "learning_rate": 0.0001491482884155874,
+      "loss": 1.8602,
+      "step": 3250
+    },
+    {
+      "epoch": 0.4034207101691764,
+      "grad_norm": 0.13389299126436824,
+      "learning_rate": 0.00014895977161778515,
+      "loss": 1.8105,
+      "step": 3255
+    },
+    {
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.1379880279783267,
+      "learning_rate": 0.00014877102568779896,
+      "loss": 1.8164,
+      "step": 3260
+    },
+    {
+      "epoch": 0.40466009791163166,
+      "grad_norm": 0.13303918044811053,
+      "learning_rate": 0.00014858205150896161,
+      "loss": 1.8984,
+      "step": 3265
+    },
+    {
+      "epoch": 0.4052797917828593,
+      "grad_norm": 0.1321191306570561,
+      "learning_rate": 0.00014839284996567392,
+      "loss": 1.8586,
+      "step": 3270
+    },
+    {
+      "epoch": 0.4058994856540869,
+      "grad_norm": 0.12648828416896452,
+      "learning_rate": 0.00014820342194340097,
+      "loss": 1.8602,
+      "step": 3275
+    },
+    {
+      "epoch": 0.4065191795253145,
+      "grad_norm": 0.1305701561156593,
+      "learning_rate": 0.00014801376832866754,
+      "loss": 1.798,
+      "step": 3280
+    },
+    {
+      "epoch": 0.4071388733965421,
+      "grad_norm": 0.1334941945609644,
+      "learning_rate": 0.00014782389000905435,
+      "loss": 1.8973,
+      "step": 3285
+    },
+    {
+      "epoch": 0.4077585672677697,
+      "grad_norm": 0.13360995175426915,
+      "learning_rate": 0.00014763378787319373,
+      "loss": 1.8,
+      "step": 3290
+    },
+    {
+      "epoch": 0.40837826113899733,
+      "grad_norm": 0.12345991314875522,
+      "learning_rate": 0.00014744346281076536,
+      "loss": 1.8465,
+      "step": 3295
+    },
+    {
+      "epoch": 0.40899795501022496,
+      "grad_norm": 0.13250381935580435,
+      "learning_rate": 0.00014725291571249236,
+      "loss": 1.7988,
+      "step": 3300
+    },
+    {
+      "epoch": 0.4096176488814526,
+      "grad_norm": 0.13402729816174527,
+      "learning_rate": 0.00014706214747013685,
+      "loss": 1.9363,
+      "step": 3305
+    },
+    {
+      "epoch": 0.4102373427526802,
+      "grad_norm": 0.1279554878458354,
+      "learning_rate": 0.00014687115897649603,
+      "loss": 1.8898,
+      "step": 3310
+    },
+    {
+      "epoch": 0.41085703662390777,
+      "grad_norm": 0.1666624811197605,
+      "learning_rate": 0.00014667995112539774,
+      "loss": 1.8234,
+      "step": 3315
+    },
+    {
+      "epoch": 0.4114767304951354,
+      "grad_norm": 0.12776748050466544,
+      "learning_rate": 0.00014648852481169658,
+      "loss": 1.893,
+      "step": 3320
+    },
+    {
+      "epoch": 0.412096424366363,
+      "grad_norm": 0.1286599439516026,
+      "learning_rate": 0.0001462968809312694,
+      "loss": 1.8762,
+      "step": 3325
+    },
+    {
+      "epoch": 0.41271611823759063,
+      "grad_norm": 0.13282507876071478,
+      "learning_rate": 0.00014610502038101138,
+      "loss": 1.8781,
+      "step": 3330
+    },
+    {
+      "epoch": 0.41333581210881826,
+      "grad_norm": 0.1337124937581548,
+      "learning_rate": 0.00014591294405883162,
+      "loss": 1.8426,
+      "step": 3335
+    },
+    {
+      "epoch": 0.4139555059800459,
+      "grad_norm": 0.12243201500863707,
+      "learning_rate": 0.00014572065286364908,
+      "loss": 1.8836,
+      "step": 3340
+    },
+    {
+      "epoch": 0.41457519985127345,
+      "grad_norm": 0.12112561857132004,
+      "learning_rate": 0.00014552814769538844,
+      "loss": 1.866,
+      "step": 3345
+    },
+    {
+      "epoch": 0.41519489372250107,
+      "grad_norm": 0.13221217015109857,
+      "learning_rate": 0.00014533542945497553,
+      "loss": 1.8617,
+      "step": 3350
+    },
+    {
+      "epoch": 0.4158145875937287,
+      "grad_norm": 0.13241339851253814,
+      "learning_rate": 0.00014514249904433362,
+      "loss": 1.8809,
+      "step": 3355
+    },
+    {
+      "epoch": 0.4164342814649563,
+      "grad_norm": 0.13060791556881998,
+      "learning_rate": 0.0001449493573663787,
+      "loss": 1.907,
+      "step": 3360
+    },
+    {
+      "epoch": 0.41705397533618394,
+      "grad_norm": 0.1463492820761344,
+      "learning_rate": 0.00014475600532501566,
+      "loss": 1.825,
+      "step": 3365
+    },
+    {
+      "epoch": 0.41767366920741156,
+      "grad_norm": 0.1267514749396063,
+      "learning_rate": 0.00014456244382513386,
+      "loss": 1.8289,
+      "step": 3370
+    },
+    {
+      "epoch": 0.4182933630786392,
+      "grad_norm": 0.12439164502803565,
+      "learning_rate": 0.0001443686737726029,
+      "loss": 1.8645,
+      "step": 3375
+    },
+    {
+      "epoch": 0.41891305694986675,
+      "grad_norm": 0.13123665423117487,
+      "learning_rate": 0.00014417469607426838,
+      "loss": 1.8602,
+      "step": 3380
+    },
+    {
+      "epoch": 0.41953275082109437,
+      "grad_norm": 0.1986304122646991,
+      "learning_rate": 0.00014398051163794776,
+      "loss": 1.8242,
+      "step": 3385
+    },
+    {
+      "epoch": 0.420152444692322,
+      "grad_norm": 0.14117447083626353,
+      "learning_rate": 0.000143786121372426,
+      "loss": 1.8586,
+      "step": 3390
+    },
+    {
+      "epoch": 0.4207721385635496,
+      "grad_norm": 0.13095017207646675,
+      "learning_rate": 0.00014359152618745132,
+      "loss": 1.8605,
+      "step": 3395
+    },
+    {
+      "epoch": 0.42139183243477724,
+      "grad_norm": 0.13210900899766176,
+      "learning_rate": 0.00014339672699373104,
+      "loss": 1.8926,
+      "step": 3400
+    },
+    {
+      "epoch": 0.42201152630600486,
+      "grad_norm": 0.14521019539611432,
+      "learning_rate": 0.0001432017247029271,
+      "loss": 1.8207,
+      "step": 3405
+    },
+    {
+      "epoch": 0.4226312201772324,
+      "grad_norm": 0.15105206106498664,
+      "learning_rate": 0.00014300652022765207,
+      "loss": 1.8301,
+      "step": 3410
+    },
+    {
+      "epoch": 0.42325091404846005,
+      "grad_norm": 0.12934697800384842,
+      "learning_rate": 0.00014281111448146468,
+      "loss": 1.7816,
+      "step": 3415
+    },
+    {
+      "epoch": 0.42387060791968767,
+      "grad_norm": 0.127108123074777,
+      "learning_rate": 0.00014261550837886566,
+      "loss": 1.8328,
+      "step": 3420
+    },
+    {
+      "epoch": 0.4244903017909153,
+      "grad_norm": 0.13025062170320045,
+      "learning_rate": 0.00014241970283529338,
+      "loss": 1.8387,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4251099956621429,
+      "grad_norm": 0.1365173510804643,
+      "learning_rate": 0.00014222369876711955,
+      "loss": 1.8457,
+      "step": 3430
+    },
+    {
+      "epoch": 0.42572968953337054,
+      "grad_norm": 0.13032412192310092,
+      "learning_rate": 0.00014202749709164506,
+      "loss": 1.8723,
+      "step": 3435
+    },
+    {
+      "epoch": 0.4263493834045981,
+      "grad_norm": 0.13718771041804576,
+      "learning_rate": 0.00014183109872709557,
+      "loss": 1.8332,
+      "step": 3440
+    },
+    {
+      "epoch": 0.4269690772758257,
+      "grad_norm": 0.12990350823599006,
+      "learning_rate": 0.0001416345045926172,
+      "loss": 1.8297,
+      "step": 3445
+    },
+    {
+      "epoch": 0.42758877114705335,
+      "grad_norm": 0.130712651106168,
+      "learning_rate": 0.00014143771560827238,
+      "loss": 1.8789,
+      "step": 3450
+    },
+    {
+      "epoch": 0.42820846501828097,
+      "grad_norm": 0.1300038947077918,
+      "learning_rate": 0.00014124073269503534,
+      "loss": 1.873,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4288281588895086,
+      "grad_norm": 0.12574163306755254,
+      "learning_rate": 0.0001410435567747879,
+      "loss": 1.8453,
+      "step": 3460
+    },
+    {
+      "epoch": 0.4294478527607362,
+      "grad_norm": 0.12626722646867042,
+      "learning_rate": 0.00014084618877031524,
+      "loss": 1.8578,
+      "step": 3465
+    },
+    {
+      "epoch": 0.43006754663196384,
+      "grad_norm": 0.13023525157695576,
+      "learning_rate": 0.00014064862960530143,
+      "loss": 1.8457,
+      "step": 3470
+    },
+    {
+      "epoch": 0.4306872405031914,
+      "grad_norm": 0.12891851891269074,
+      "learning_rate": 0.0001404508802043252,
+      "loss": 1.8496,
+      "step": 3475
+    },
+    {
+      "epoch": 0.431306934374419,
+      "grad_norm": 0.1317883450339982,
+      "learning_rate": 0.0001402529414928556,
+      "loss": 1.8367,
+      "step": 3480
+    },
+    {
+      "epoch": 0.43192662824564665,
+      "grad_norm": 0.1320357092237638,
+      "learning_rate": 0.00014005481439724753,
+      "loss": 1.8512,
+      "step": 3485
+    },
+    {
+      "epoch": 0.43254632211687427,
+      "grad_norm": 0.13106411312502514,
+      "learning_rate": 0.00013985649984473773,
+      "loss": 1.8703,
+      "step": 3490
+    },
+    {
+      "epoch": 0.4331660159881019,
+      "grad_norm": 0.1293849394353448,
+      "learning_rate": 0.0001396579987634401,
+      "loss": 1.7855,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4337857098593295,
+      "grad_norm": 0.13104584752365614,
+      "learning_rate": 0.00013945931208234156,
+      "loss": 1.8879,
+      "step": 3500
+    },
+    {
+      "epoch": 0.4344054037305571,
+      "grad_norm": 0.13185634907081104,
+      "learning_rate": 0.00013926044073129759,
+      "loss": 1.827,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4350250976017847,
+      "grad_norm": 0.13564066779325817,
+      "learning_rate": 0.00013906138564102793,
+      "loss": 1.8605,
+      "step": 3510
+    },
+    {
+      "epoch": 0.4356447914730123,
+      "grad_norm": 0.12676901341611954,
+      "learning_rate": 0.0001388621477431123,
+      "loss": 1.8656,
+      "step": 3515
+    },
+    {
+      "epoch": 0.43626448534423995,
+      "grad_norm": 0.12510362563216565,
+      "learning_rate": 0.00013866272796998587,
+      "loss": 1.8875,
+      "step": 3520
+    },
+    {
+      "epoch": 0.43688417921546757,
+      "grad_norm": 0.13507506019177096,
+      "learning_rate": 0.00013846312725493504,
+      "loss": 1.9164,
+      "step": 3525
+    },
+    {
+      "epoch": 0.4375038730866952,
+      "grad_norm": 0.132093881424524,
+      "learning_rate": 0.00013826334653209297,
+      "loss": 1.843,
+      "step": 3530
+    },
+    {
+      "epoch": 0.43812356695792276,
+      "grad_norm": 0.1421065871061589,
+      "learning_rate": 0.00013806338673643534,
+      "loss": 1.8262,
+      "step": 3535
+    },
+    {
+      "epoch": 0.4387432608291504,
+      "grad_norm": 0.13102399751421243,
+      "learning_rate": 0.00013786324880377576,
+      "loss": 1.8156,
+      "step": 3540
+    },
+    {
+      "epoch": 0.439362954700378,
+      "grad_norm": 0.13358396691833502,
+      "learning_rate": 0.0001376629336707617,
+      "loss": 1.8195,
+      "step": 3545
+    },
+    {
+      "epoch": 0.43998264857160563,
+      "grad_norm": 0.13047350696632068,
+      "learning_rate": 0.0001374624422748698,
+      "loss": 1.8738,
+      "step": 3550
+    },
+    {
+      "epoch": 0.44060234244283325,
+      "grad_norm": 0.13627153690659918,
+      "learning_rate": 0.00013726177555440164,
+      "loss": 1.8785,
+      "step": 3555
+    },
+    {
+      "epoch": 0.4412220363140609,
+      "grad_norm": 0.13540316590544368,
+      "learning_rate": 0.0001370609344484793,
+      "loss": 1.8113,
+      "step": 3560
+    },
+    {
+      "epoch": 0.4418417301852885,
+      "grad_norm": 0.14107784380640415,
+      "learning_rate": 0.00013685991989704105,
+      "loss": 1.8199,
+      "step": 3565
+    },
+    {
+      "epoch": 0.44246142405651606,
+      "grad_norm": 0.12911502937346492,
+      "learning_rate": 0.00013665873284083685,
+      "loss": 1.8387,
+      "step": 3570
+    },
+    {
+      "epoch": 0.4430811179277437,
+      "grad_norm": 0.13109208746838247,
+      "learning_rate": 0.000136457374221424,
+      "loss": 1.8465,
+      "step": 3575
+    },
+    {
+      "epoch": 0.4437008117989713,
+      "grad_norm": 0.1409108873500928,
+      "learning_rate": 0.00013625584498116262,
+      "loss": 1.8617,
+      "step": 3580
+    },
+    {
+      "epoch": 0.44432050567019893,
+      "grad_norm": 0.1296434499076529,
+      "learning_rate": 0.00013605414606321148,
+      "loss": 1.8629,
+      "step": 3585
+    },
+    {
+      "epoch": 0.44494019954142655,
+      "grad_norm": 0.1438673184528456,
+      "learning_rate": 0.0001358522784115234,
+      "loss": 1.8715,
+      "step": 3590
+    },
+    {
+      "epoch": 0.4455598934126542,
+      "grad_norm": 0.13609879637032504,
+      "learning_rate": 0.00013565024297084084,
+      "loss": 1.8895,
+      "step": 3595
+    },
+    {
+      "epoch": 0.44617958728388174,
+      "grad_norm": 0.13036349295369937,
+      "learning_rate": 0.0001354480406866915,
+      "loss": 1.8031,
+      "step": 3600
+    },
+    {
+      "epoch": 0.44679928115510936,
+      "grad_norm": 0.1350939728981667,
+      "learning_rate": 0.00013524567250538396,
+      "loss": 1.8594,
+      "step": 3605
+    },
+    {
+      "epoch": 0.447418975026337,
+      "grad_norm": 0.12714971692456176,
+      "learning_rate": 0.00013504313937400317,
+      "loss": 1.8668,
+      "step": 3610
+    },
+    {
+      "epoch": 0.4480386688975646,
+      "grad_norm": 0.13951214846411,
+      "learning_rate": 0.00013484044224040606,
+      "loss": 1.8418,
+      "step": 3615
+    },
+    {
+      "epoch": 0.44865836276879223,
+      "grad_norm": 0.1386490336407402,
+      "learning_rate": 0.00013463758205321715,
+      "loss": 1.8695,
+      "step": 3620
+    },
+    {
+      "epoch": 0.44927805664001985,
+      "grad_norm": 0.13782618102933727,
+      "learning_rate": 0.0001344345597618239,
+      "loss": 1.8273,
+      "step": 3625
+    },
+    {
+      "epoch": 0.4498977505112474,
+      "grad_norm": 0.13579485790542123,
+      "learning_rate": 0.00013423137631637258,
+      "loss": 1.8586,
+      "step": 3630
+    },
+    {
+      "epoch": 0.45051744438247504,
+      "grad_norm": 0.1300030317668718,
+      "learning_rate": 0.00013402803266776353,
+      "loss": 1.8652,
+      "step": 3635
+    },
+    {
+      "epoch": 0.45113713825370266,
+      "grad_norm": 0.13598316227960938,
+      "learning_rate": 0.00013382452976764693,
+      "loss": 1.8309,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4517568321249303,
+      "grad_norm": 0.12953667957338744,
+      "learning_rate": 0.00013362086856841826,
+      "loss": 1.8668,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4523765259961579,
+      "grad_norm": 0.13451954909685226,
+      "learning_rate": 0.0001334170500232138,
+      "loss": 1.8543,
+      "step": 3650
+    },
+    {
+      "epoch": 0.45299621986738553,
+      "grad_norm": 0.13195858085157425,
+      "learning_rate": 0.00013321307508590624,
+      "loss": 1.8504,
+      "step": 3655
+    },
+    {
+      "epoch": 0.45361591373861315,
+      "grad_norm": 0.1305239663789138,
+      "learning_rate": 0.00013300894471110014,
+      "loss": 1.8621,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4542356076098407,
+      "grad_norm": 0.13635413078375763,
+      "learning_rate": 0.00013280465985412757,
+      "loss": 1.827,
+      "step": 3665
+    },
+    {
+      "epoch": 0.45485530148106834,
+      "grad_norm": 0.1258025644634598,
+      "learning_rate": 0.00013260022147104354,
+      "loss": 1.85,
+      "step": 3670
+    },
+    {
+      "epoch": 0.45547499535229596,
+      "grad_norm": 0.1415524251745301,
+      "learning_rate": 0.00013239563051862158,
+      "loss": 1.8246,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4560946892235236,
+      "grad_norm": 0.13622385001598838,
+      "learning_rate": 0.00013219088795434923,
+      "loss": 1.8383,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4567143830947512,
+      "grad_norm": 0.1396238647704036,
+      "learning_rate": 0.00013198599473642354,
+      "loss": 1.8922,
+      "step": 3685
+    },
+    {
+      "epoch": 0.45733407696597883,
+      "grad_norm": 0.12825728444600357,
+      "learning_rate": 0.00013178095182374676,
+      "loss": 1.7937,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4579537708372064,
+      "grad_norm": 0.14084681880849642,
+      "learning_rate": 0.00013157576017592157,
+      "loss": 1.8258,
+      "step": 3695
+    },
+    {
+      "epoch": 0.458573464708434,
+      "grad_norm": 0.1385483862626899,
+      "learning_rate": 0.0001313704207532468,
+      "loss": 1.8488,
+      "step": 3700
+    },
+    {
+      "epoch": 0.45919315857966164,
+      "grad_norm": 0.1289536639011512,
+      "learning_rate": 0.00013116493451671279,
+      "loss": 1.8352,
+      "step": 3705
+    },
+    {
+      "epoch": 0.45981285245088926,
+      "grad_norm": 0.1289125265280001,
+      "learning_rate": 0.000130959302427997,
+      "loss": 1.7816,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4604325463221169,
+      "grad_norm": 0.13577383918982386,
+      "learning_rate": 0.00013075352544945966,
+      "loss": 1.866,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4610522401933445,
+      "grad_norm": 0.13279564421407647,
+      "learning_rate": 0.00013054760454413882,
+      "loss": 1.8707,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4616719340645721,
+      "grad_norm": 0.13165317327395187,
+      "learning_rate": 0.00013034154067574622,
+      "loss": 1.8813,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4622916279357997,
+      "grad_norm": 0.13474300432005532,
+      "learning_rate": 0.00013013533480866273,
+      "loss": 1.8141,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4629113218070273,
+      "grad_norm": 0.1348771890055806,
+      "learning_rate": 0.00012992898790793362,
+      "loss": 1.8488,
+      "step": 3735
+    },
+    {
+      "epoch": 0.46353101567825494,
+      "grad_norm": 0.13101703748539598,
+      "learning_rate": 0.00012972250093926436,
+      "loss": 1.85,
+      "step": 3740
+    },
+    {
+      "epoch": 0.46415070954948257,
+      "grad_norm": 0.12926638302593185,
+      "learning_rate": 0.0001295158748690159,
+      "loss": 1.8449,
+      "step": 3745
+    },
+    {
+      "epoch": 0.4647704034207102,
+      "grad_norm": 0.12685006085742664,
+      "learning_rate": 0.0001293091106642001,
+      "loss": 1.8676,
+      "step": 3750
+    },
+    {
+      "epoch": 0.4653900972919378,
+      "grad_norm": 0.13335175429016402,
+      "learning_rate": 0.00012910220929247538,
+      "loss": 1.8445,
+      "step": 3755
+    },
+    {
+      "epoch": 0.4660097911631654,
+      "grad_norm": 0.1334428844711798,
+      "learning_rate": 0.00012889517172214206,
+      "loss": 1.8258,
+      "step": 3760
+    },
+    {
+      "epoch": 0.466629485034393,
+      "grad_norm": 0.13235299019438176,
+      "learning_rate": 0.0001286879989221379,
+      "loss": 1.8328,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4672491789056206,
+      "grad_norm": 0.13692059560954664,
+      "learning_rate": 0.0001284806918620335,
+      "loss": 1.8285,
+      "step": 3770
+    },
+    {
+      "epoch": 0.46786887277684824,
+      "grad_norm": 0.1241361680899772,
+      "learning_rate": 0.00012827325151202782,
+      "loss": 1.8387,
+      "step": 3775
+    },
+    {
+      "epoch": 0.46848856664807587,
+      "grad_norm": 0.1339809286185403,
+      "learning_rate": 0.00012806567884294362,
+      "loss": 1.8816,
+      "step": 3780
+    },
+    {
+      "epoch": 0.4691082605193035,
+      "grad_norm": 0.12897521994670574,
+      "learning_rate": 0.00012785797482622294,
+      "loss": 1.8398,
+      "step": 3785
+    },
+    {
+      "epoch": 0.46972795439053106,
+      "grad_norm": 0.13985587399436297,
+      "learning_rate": 0.00012765014043392242,
+      "loss": 1.848,
+      "step": 3790
+    },
+    {
+      "epoch": 0.4703476482617587,
+      "grad_norm": 0.13828870435326854,
+      "learning_rate": 0.00012744217663870902,
+      "loss": 1.8797,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4709673421329863,
+      "grad_norm": 0.138650980685239,
+      "learning_rate": 0.00012723408441385521,
+      "loss": 1.8598,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4715870360042139,
+      "grad_norm": 0.13009664188771436,
+      "learning_rate": 0.0001270258647332345,
+      "loss": 1.8523,
+      "step": 3805
+    },
+    {
+      "epoch": 0.47220672987544154,
+      "grad_norm": 0.129438419879605,
+      "learning_rate": 0.00012681751857131693,
+      "loss": 1.8004,
+      "step": 3810
+    },
+    {
+      "epoch": 0.47282642374666917,
+      "grad_norm": 0.13070922171883706,
+      "learning_rate": 0.00012660904690316445,
+      "loss": 1.8504,
+      "step": 3815
+    },
+    {
+      "epoch": 0.47344611761789673,
+      "grad_norm": 0.12668656962083866,
+      "learning_rate": 0.00012640045070442643,
+      "loss": 1.8609,
+      "step": 3820
+    },
+    {
+      "epoch": 0.47406581148912436,
+      "grad_norm": 0.13741354466112743,
+      "learning_rate": 0.000126191730951335,
+      "loss": 1.809,
+      "step": 3825
+    },
+    {
+      "epoch": 0.474685505360352,
+      "grad_norm": 0.12823342134235233,
+      "learning_rate": 0.0001259828886207005,
+      "loss": 1.8492,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4753051992315796,
+      "grad_norm": 0.1379250744950109,
+      "learning_rate": 0.00012577392468990695,
+      "loss": 1.934,
+      "step": 3835
+    },
+    {
+      "epoch": 0.4759248931028072,
+      "grad_norm": 0.13202143342192066,
+      "learning_rate": 0.00012556484013690763,
+      "loss": 1.8352,
+      "step": 3840
+    },
+    {
+      "epoch": 0.47654458697403485,
+      "grad_norm": 0.13299754683798998,
+      "learning_rate": 0.00012535563594022,
+      "loss": 1.8297,
+      "step": 3845
+    },
+    {
+      "epoch": 0.47716428084526247,
+      "grad_norm": 0.12790876592097397,
+      "learning_rate": 0.0001251463130789217,
+      "loss": 1.8383,
+      "step": 3850
+    },
+    {
+      "epoch": 0.47778397471649003,
+      "grad_norm": 0.12997304481536676,
+      "learning_rate": 0.0001249368725326457,
+      "loss": 1.8055,
+      "step": 3855
+    },
+    {
+      "epoch": 0.47840366858771766,
+      "grad_norm": 0.12972319050241293,
+      "learning_rate": 0.00012472731528157563,
+      "loss": 1.8434,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4790233624589453,
+      "grad_norm": 0.13199195937817895,
+      "learning_rate": 0.00012451764230644145,
+      "loss": 1.8934,
+      "step": 3865
+    },
+    {
+      "epoch": 0.4796430563301729,
+      "grad_norm": 0.13461248829630099,
+      "learning_rate": 0.0001243078545885145,
+      "loss": 1.8602,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4802627502014005,
+      "grad_norm": 0.1355432249144203,
+      "learning_rate": 0.00012409795310960333,
+      "loss": 1.8156,
+      "step": 3875
+    },
+    {
+      "epoch": 0.48088244407262815,
+      "grad_norm": 0.14482071533572835,
+      "learning_rate": 0.00012388793885204875,
+      "loss": 1.8492,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4815021379438557,
+      "grad_norm": 0.13371691055623847,
+      "learning_rate": 0.00012367781279871946,
+      "loss": 1.782,
+      "step": 3885
+    },
+    {
+      "epoch": 0.48212183181508333,
+      "grad_norm": 0.1447992804565308,
+      "learning_rate": 0.00012346757593300733,
+      "loss": 1.8227,
+      "step": 3890
+    },
+    {
+      "epoch": 0.48274152568631096,
+      "grad_norm": 0.12767703586264148,
+      "learning_rate": 0.00012325722923882285,
+      "loss": 1.841,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4833612195575386,
+      "grad_norm": 0.1314306698392859,
+      "learning_rate": 0.00012304677370059047,
+      "loss": 1.825,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4839809134287662,
+      "grad_norm": 0.1345967559038808,
+      "learning_rate": 0.00012283621030324403,
+      "loss": 1.7437,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4846006072999938,
+      "grad_norm": 0.13106442792164216,
+      "learning_rate": 0.00012262554003222221,
+      "loss": 1.8285,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4852203011712214,
+      "grad_norm": 0.1398213559576168,
+      "learning_rate": 0.00012241476387346386,
+      "loss": 1.8332,
+      "step": 3915
+    },
+    {
+      "epoch": 0.485839995042449,
+      "grad_norm": 0.13124036303853287,
+      "learning_rate": 0.00012220388281340328,
+      "loss": 1.8324,
+      "step": 3920
+    },
+    {
+      "epoch": 0.48645968891367664,
+      "grad_norm": 0.1421186518175135,
+      "learning_rate": 0.00012199289783896582,
+      "loss": 1.8418,
+      "step": 3925
+    },
+    {
+      "epoch": 0.48707938278490426,
+      "grad_norm": 0.1364339784527228,
+      "learning_rate": 0.0001217818099375631,
+      "loss": 1.775,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4876990766561319,
+      "grad_norm": 0.1335459935708431,
+      "learning_rate": 0.00012157062009708847,
+      "loss": 1.8012,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4883187705273595,
+      "grad_norm": 0.1359760422124762,
+      "learning_rate": 0.00012135932930591232,
+      "loss": 1.8246,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4889384643985871,
+      "grad_norm": 0.12394543676184719,
+      "learning_rate": 0.00012114793855287749,
+      "loss": 1.8207,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4895581582698147,
+      "grad_norm": 0.1304584073821454,
+      "learning_rate": 0.00012093644882729473,
+      "loss": 1.8484,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4901778521410423,
+      "grad_norm": 0.1277387513791641,
+      "learning_rate": 0.0001207248611189378,
+      "loss": 1.8074,
+      "step": 3955
+    },
+    {
+      "epoch": 0.49079754601226994,
+      "grad_norm": 0.13299457188730449,
+      "learning_rate": 0.00012051317641803921,
+      "loss": 1.8664,
+      "step": 3960
+    },
+    {
+      "epoch": 0.49141723988349756,
+      "grad_norm": 0.1342304756979945,
+      "learning_rate": 0.00012030139571528534,
+      "loss": 1.8043,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4920369337547252,
+      "grad_norm": 0.13564000088494185,
+      "learning_rate": 0.00012008952000181175,
+      "loss": 1.8113,
+      "step": 3970
+    },
+    {
+      "epoch": 0.4926566276259528,
+      "grad_norm": 0.1351196466939399,
+      "learning_rate": 0.0001198775502691988,
+      "loss": 1.9066,
+      "step": 3975
+    },
+    {
+      "epoch": 0.49327632149718037,
+      "grad_norm": 0.1357026978518113,
+      "learning_rate": 0.00011966548750946678,
+      "loss": 1.7566,
+      "step": 3980
+    },
+    {
+      "epoch": 0.493896015368408,
+      "grad_norm": 0.13117708579809673,
+      "learning_rate": 0.00011945333271507138,
+      "loss": 1.8559,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4945157092396356,
+      "grad_norm": 0.13842921233959002,
+      "learning_rate": 0.00011924108687889899,
+      "loss": 1.8441,
+      "step": 3990
+    },
+    {
+      "epoch": 0.49513540311086324,
+      "grad_norm": 0.1328649899399259,
+      "learning_rate": 0.00011902875099426207,
+      "loss": 1.8445,
+      "step": 3995
+    },
+    {
+      "epoch": 0.49575509698209086,
+      "grad_norm": 0.13155897661920435,
+      "learning_rate": 0.00011881632605489457,
+      "loss": 1.8977,
+      "step": 4000
+    },
+    {
+      "epoch": 0.4963747908533185,
+      "grad_norm": 0.13187321718339992,
+      "learning_rate": 0.0001186038130549471,
+      "loss": 1.7977,
+      "step": 4005
+    },
+    {
+      "epoch": 0.49699448472454605,
+      "grad_norm": 0.13286908074253748,
+      "learning_rate": 0.00011839121298898253,
+      "loss": 1.9191,
+      "step": 4010
+    },
+    {
+      "epoch": 0.49761417859577367,
+      "grad_norm": 0.13199065193371257,
+      "learning_rate": 0.00011817852685197109,
+      "loss": 1.7773,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4982338724670013,
+      "grad_norm": 0.1373404116593631,
+      "learning_rate": 0.00011796575563928591,
+      "loss": 1.8336,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4988535663382289,
+      "grad_norm": 0.13885642586504665,
+      "learning_rate": 0.00011775290034669822,
+      "loss": 1.8387,
+      "step": 4025
+    },
+    {
+      "epoch": 0.49947326020945654,
+      "grad_norm": 0.14373947655079228,
+      "learning_rate": 0.00011753996197037272,
+      "loss": 1.8379,
+      "step": 4030
+    },
+    {
+      "epoch": 0.5000929540806841,
+      "grad_norm": 0.12835515006238413,
+      "learning_rate": 0.00011732694150686301,
+      "loss": 1.8641,
+      "step": 4035
+    },
+    {
+      "epoch": 0.5007126479519117,
+      "grad_norm": 0.13281937799249963,
+      "learning_rate": 0.00011711383995310681,
+      "loss": 1.8371,
+      "step": 4040
+    },
+    {
+      "epoch": 0.5013323418231393,
+      "grad_norm": 0.134641067216936,
+      "learning_rate": 0.00011690065830642143,
+      "loss": 1.766,
+      "step": 4045
+    },
+    {
+      "epoch": 0.501952035694367,
+      "grad_norm": 0.1406950506844368,
+      "learning_rate": 0.00011668739756449885,
+      "loss": 1.7727,
+      "step": 4050
+    },
+    {
+      "epoch": 0.5025717295655946,
+      "grad_norm": 0.1377858503969937,
+      "learning_rate": 0.00011647405872540138,
+      "loss": 1.8195,
+      "step": 4055
+    },
+    {
+      "epoch": 0.5031914234368222,
+      "grad_norm": 0.13313565675420963,
+      "learning_rate": 0.00011626064278755673,
+      "loss": 1.7883,
+      "step": 4060
+    },
+    {
+      "epoch": 0.5038111173080498,
+      "grad_norm": 0.13459471367809545,
+      "learning_rate": 0.00011604715074975347,
+      "loss": 1.8438,
+      "step": 4065
+    },
+    {
+      "epoch": 0.5044308111792775,
+      "grad_norm": 0.1311158940726712,
+      "learning_rate": 0.00011583358361113632,
+      "loss": 1.9012,
+      "step": 4070
+    },
+    {
+      "epoch": 0.5050505050505051,
+      "grad_norm": 0.13876753398913574,
+      "learning_rate": 0.00011561994237120148,
+      "loss": 1.8672,
+      "step": 4075
+    },
+    {
+      "epoch": 0.5056701989217327,
+      "grad_norm": 0.13183560468669342,
+      "learning_rate": 0.00011540622802979187,
+      "loss": 1.8582,
+      "step": 4080
+    },
+    {
+      "epoch": 0.5062898927929603,
+      "grad_norm": 0.13580340297486979,
+      "learning_rate": 0.00011519244158709263,
+      "loss": 1.8945,
+      "step": 4085
+    },
+    {
+      "epoch": 0.5069095866641878,
+      "grad_norm": 0.13991883408059513,
+      "learning_rate": 0.00011497858404362631,
+      "loss": 1.8926,
+      "step": 4090
+    },
+    {
+      "epoch": 0.5075292805354155,
+      "grad_norm": 0.14012773463184303,
+      "learning_rate": 0.00011476465640024814,
+      "loss": 1.8121,
+      "step": 4095
+    },
+    {
+      "epoch": 0.5081489744066431,
+      "grad_norm": 0.1301910255761767,
+      "learning_rate": 0.00011455065965814148,
+      "loss": 1.8758,
+      "step": 4100
+    },
+    {
+      "epoch": 0.5087686682778707,
+      "grad_norm": 0.13887521410538325,
+      "learning_rate": 0.00011433659481881307,
+      "loss": 1.841,
+      "step": 4105
+    },
+    {
+      "epoch": 0.5093883621490983,
+      "grad_norm": 0.12807030791037602,
+      "learning_rate": 0.00011412246288408835,
+      "loss": 1.8855,
+      "step": 4110
+    },
+    {
+      "epoch": 0.510008056020326,
+      "grad_norm": 0.12976882936743359,
+      "learning_rate": 0.00011390826485610675,
+      "loss": 1.8348,
+      "step": 4115
+    },
+    {
+      "epoch": 0.5106277498915536,
+      "grad_norm": 0.12871810923015659,
+      "learning_rate": 0.000113694001737317,
+      "loss": 1.8852,
+      "step": 4120
+    },
+    {
+      "epoch": 0.5112474437627812,
+      "grad_norm": 0.12717099476381863,
+      "learning_rate": 0.00011347967453047248,
+      "loss": 1.8328,
+      "step": 4125
+    },
+    {
+      "epoch": 0.5118671376340088,
+      "grad_norm": 0.13121903677104316,
+      "learning_rate": 0.00011326528423862653,
+      "loss": 1.8375,
+      "step": 4130
+    },
+    {
+      "epoch": 0.5124868315052364,
+      "grad_norm": 0.12986993222627213,
+      "learning_rate": 0.00011305083186512765,
+      "loss": 1.8043,
+      "step": 4135
+    },
+    {
+      "epoch": 0.5131065253764641,
+      "grad_norm": 0.13292488082615117,
+      "learning_rate": 0.00011283631841361499,
+      "loss": 1.8258,
+      "step": 4140
+    },
+    {
+      "epoch": 0.5137262192476917,
+      "grad_norm": 0.1290602203925276,
+      "learning_rate": 0.00011262174488801349,
+      "loss": 1.868,
+      "step": 4145
+    },
+    {
+      "epoch": 0.5143459131189193,
+      "grad_norm": 0.1340007696562376,
+      "learning_rate": 0.00011240711229252915,
+      "loss": 1.8273,
+      "step": 4150
+    },
+    {
+      "epoch": 0.5149656069901468,
+      "grad_norm": 0.13145813073711832,
+      "learning_rate": 0.00011219242163164457,
+      "loss": 1.9297,
+      "step": 4155
+    },
+    {
+      "epoch": 0.5155853008613744,
+      "grad_norm": 0.1313762201180359,
+      "learning_rate": 0.000111977673910114,
+      "loss": 1.9039,
+      "step": 4160
+    },
+    {
+      "epoch": 0.5162049947326021,
+      "grad_norm": 0.1323692673503486,
+      "learning_rate": 0.00011176287013295879,
+      "loss": 1.8992,
+      "step": 4165
+    },
+    {
+      "epoch": 0.5168246886038297,
+      "grad_norm": 0.14101520024243003,
+      "learning_rate": 0.0001115480113054626,
+      "loss": 1.8641,
+      "step": 4170
+    },
+    {
+      "epoch": 0.5174443824750573,
+      "grad_norm": 0.1351388006249665,
+      "learning_rate": 0.00011133309843316669,
+      "loss": 1.8871,
+      "step": 4175
+    },
+    {
+      "epoch": 0.5180640763462849,
+      "grad_norm": 0.1401631083665635,
+      "learning_rate": 0.0001111181325218653,
+      "loss": 1.7629,
+      "step": 4180
+    },
+    {
+      "epoch": 0.5186837702175126,
+      "grad_norm": 0.1385892991917342,
+      "learning_rate": 0.00011090311457760094,
+      "loss": 1.8035,
+      "step": 4185
+    },
+    {
+      "epoch": 0.5193034640887402,
+      "grad_norm": 0.13371107142455393,
+      "learning_rate": 0.0001106880456066595,
+      "loss": 1.7957,
+      "step": 4190
+    },
+    {
+      "epoch": 0.5199231579599678,
+      "grad_norm": 0.13066975486189664,
+      "learning_rate": 0.00011047292661556581,
+      "loss": 1.8027,
+      "step": 4195
+    },
+    {
+      "epoch": 0.5205428518311954,
+      "grad_norm": 0.1358953784147378,
+      "learning_rate": 0.00011025775861107874,
+      "loss": 1.827,
+      "step": 4200
+    },
+    {
+      "epoch": 0.521162545702423,
+      "grad_norm": 0.1312496989386351,
+      "learning_rate": 0.00011004254260018648,
+      "loss": 1.8441,
+      "step": 4205
+    },
+    {
+      "epoch": 0.5217822395736507,
+      "grad_norm": 0.13783425853932574,
+      "learning_rate": 0.00010982727959010201,
+      "loss": 1.8816,
+      "step": 4210
+    },
+    {
+      "epoch": 0.5224019334448782,
+      "grad_norm": 0.127457472607005,
+      "learning_rate": 0.00010961197058825817,
+      "loss": 1.8676,
+      "step": 4215
+    },
+    {
+      "epoch": 0.5230216273161058,
+      "grad_norm": 0.13820731677300183,
+      "learning_rate": 0.00010939661660230309,
+      "loss": 1.8344,
+      "step": 4220
+    },
+    {
+      "epoch": 0.5236413211873334,
+      "grad_norm": 0.13449834908571123,
+      "learning_rate": 0.00010918121864009543,
+      "loss": 1.8309,
+      "step": 4225
+    },
+    {
+      "epoch": 0.524261015058561,
+      "grad_norm": 0.13724886784231616,
+      "learning_rate": 0.00010896577770969964,
+      "loss": 1.8043,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5248807089297887,
+      "grad_norm": 0.12837832712288905,
+      "learning_rate": 0.00010875029481938126,
+      "loss": 1.902,
+      "step": 4235
+    },
+    {
+      "epoch": 0.5255004028010163,
+      "grad_norm": 0.13135449306647268,
+      "learning_rate": 0.00010853477097760222,
+      "loss": 1.8266,
+      "step": 4240
+    },
+    {
+      "epoch": 0.5261200966722439,
+      "grad_norm": 0.136587596414044,
+      "learning_rate": 0.0001083192071930161,
+      "loss": 1.8004,
+      "step": 4245
+    },
+    {
+      "epoch": 0.5267397905434715,
+      "grad_norm": 0.12616824686941636,
+      "learning_rate": 0.00010810360447446335,
+      "loss": 1.8512,
+      "step": 4250
+    },
+    {
+      "epoch": 0.5273594844146992,
+      "grad_norm": 0.13515955918539224,
+      "learning_rate": 0.00010788796383096676,
+      "loss": 1.8156,
+      "step": 4255
+    },
+    {
+      "epoch": 0.5279791782859268,
+      "grad_norm": 0.13077554655712312,
+      "learning_rate": 0.00010767228627172645,
+      "loss": 1.8566,
+      "step": 4260
+    },
+    {
+      "epoch": 0.5285988721571544,
+      "grad_norm": 0.1311515087885154,
+      "learning_rate": 0.00010745657280611552,
+      "loss": 1.8633,
+      "step": 4265
+    },
+    {
+      "epoch": 0.529218566028382,
+      "grad_norm": 0.13183696062014078,
+      "learning_rate": 0.00010724082444367485,
+      "loss": 1.8453,
+      "step": 4270
+    },
+    {
+      "epoch": 0.5298382598996096,
+      "grad_norm": 0.1326471397034417,
+      "learning_rate": 0.00010702504219410884,
+      "loss": 1.8797,
+      "step": 4275
+    },
+    {
+      "epoch": 0.5304579537708372,
+      "grad_norm": 0.13074802402419886,
+      "learning_rate": 0.00010680922706728041,
+      "loss": 1.7977,
+      "step": 4280
+    },
+    {
+      "epoch": 0.5310776476420648,
+      "grad_norm": 0.13471386870044813,
+      "learning_rate": 0.00010659338007320632,
+      "loss": 1.9219,
+      "step": 4285
+    },
+    {
+      "epoch": 0.5316973415132924,
+      "grad_norm": 0.13101063536055021,
+      "learning_rate": 0.00010637750222205253,
+      "loss": 1.8273,
+      "step": 4290
+    },
+    {
+      "epoch": 0.53231703538452,
+      "grad_norm": 0.1330552224706942,
+      "learning_rate": 0.00010616159452412939,
+      "loss": 1.8289,
+      "step": 4295
+    },
+    {
+      "epoch": 0.5329367292557476,
+      "grad_norm": 0.13232572723510502,
+      "learning_rate": 0.00010594565798988689,
+      "loss": 1.8492,
+      "step": 4300
+    },
+    {
+      "epoch": 0.5335564231269753,
+      "grad_norm": 0.14341248190041536,
+      "learning_rate": 0.00010572969362990998,
+      "loss": 1.8523,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5341761169982029,
+      "grad_norm": 0.1326750594937519,
+      "learning_rate": 0.00010551370245491394,
+      "loss": 1.843,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5347958108694305,
+      "grad_norm": 0.13873579584003962,
+      "learning_rate": 0.00010529768547573942,
+      "loss": 1.8121,
+      "step": 4315
+    },
+    {
+      "epoch": 0.5354155047406581,
+      "grad_norm": 0.13163567087274103,
+      "learning_rate": 0.00010508164370334787,
+      "loss": 1.8465,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5360351986118858,
+      "grad_norm": 0.13968960114860696,
+      "learning_rate": 0.00010486557814881686,
+      "loss": 1.8352,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5366548924831134,
+      "grad_norm": 0.13788540405986352,
+      "learning_rate": 0.00010464948982333504,
+      "loss": 1.8113,
+      "step": 4330
+    },
+    {
+      "epoch": 0.537274586354341,
+      "grad_norm": 0.1369901284801625,
+      "learning_rate": 0.00010443337973819791,
+      "loss": 1.8141,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5378942802255686,
+      "grad_norm": 0.14142406746780717,
+      "learning_rate": 0.00010421724890480258,
+      "loss": 1.8234,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5385139740967961,
+      "grad_norm": 0.14077233215811605,
+      "learning_rate": 0.00010400109833464338,
+      "loss": 1.859,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5391336679680238,
+      "grad_norm": 0.14302584088863268,
+      "learning_rate": 0.00010378492903930699,
+      "loss": 1.8238,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5397533618392514,
+      "grad_norm": 0.14375440559453245,
+      "learning_rate": 0.00010356874203046766,
+      "loss": 1.8383,
+      "step": 4355
+    },
+    {
+      "epoch": 0.540373055710479,
+      "grad_norm": 0.12698327084195696,
+      "learning_rate": 0.00010335253831988267,
+      "loss": 1.8203,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5409927495817066,
+      "grad_norm": 0.12860058438618013,
+      "learning_rate": 0.00010313631891938736,
+      "loss": 1.8637,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5416124434529342,
+      "grad_norm": 0.13436040029788598,
+      "learning_rate": 0.00010292008484089047,
+      "loss": 1.8523,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5422321373241619,
+      "grad_norm": 0.1345634167056628,
+      "learning_rate": 0.0001027038370963695,
+      "loss": 1.8539,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5428518311953895,
+      "grad_norm": 0.13589085190830785,
+      "learning_rate": 0.00010248757669786594,
+      "loss": 1.841,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5434715250666171,
+      "grad_norm": 0.1394154520082444,
+      "learning_rate": 0.00010227130465748045,
+      "loss": 1.8855,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5440912189378447,
+      "grad_norm": 0.13059784478842867,
+      "learning_rate": 0.00010205502198736816,
+      "loss": 1.7937,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5447109128090724,
+      "grad_norm": 0.12746980330896387,
+      "learning_rate": 0.00010183872969973396,
+      "loss": 1.8199,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5453306066803,
+      "grad_norm": 0.1535462716817909,
+      "learning_rate": 0.00010162242880682776,
+      "loss": 1.8605,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5459503005515276,
+      "grad_norm": 0.1342716210660633,
+      "learning_rate": 0.00010140612032093972,
+      "loss": 1.8234,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5465699944227551,
+      "grad_norm": 0.1318603678326961,
+      "learning_rate": 0.00010118980525439559,
+      "loss": 1.8414,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5471896882939827,
+      "grad_norm": 0.1344047381087586,
+      "learning_rate": 0.00010097348461955186,
+      "loss": 1.7937,
+      "step": 4415
+    },
+    {
+      "epoch": 0.5478093821652104,
+      "grad_norm": 0.13155594328043496,
+      "learning_rate": 0.00010075715942879114,
+      "loss": 1.8137,
+      "step": 4420
+    },
+    {
+      "epoch": 0.548429076036438,
+      "grad_norm": 0.13086140046383155,
+      "learning_rate": 0.00010054083069451728,
+      "loss": 1.841,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5490487699076656,
+      "grad_norm": 0.13099764443257447,
+      "learning_rate": 0.00010032449942915072,
+      "loss": 1.8141,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5496684637788932,
+      "grad_norm": 0.14093728038783374,
+      "learning_rate": 0.00010010816664512389,
+      "loss": 1.8008,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5502881576501208,
+      "grad_norm": 0.12948947274284553,
+      "learning_rate": 9.989183335487615e-05,
+      "loss": 1.8656,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5509078515213485,
+      "grad_norm": 0.1346182653084658,
+      "learning_rate": 9.96755005708493e-05,
+      "loss": 1.8719,
+      "step": 4445
+    },
+    {
+      "epoch": 0.5515275453925761,
+      "grad_norm": 0.13528704687502488,
+      "learning_rate": 9.945916930548276e-05,
+      "loss": 1.9227,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5521472392638037,
+      "grad_norm": 0.13495978964006375,
+      "learning_rate": 9.924284057120888e-05,
+      "loss": 1.8027,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5527669331350313,
+      "grad_norm": 0.13224572304109156,
+      "learning_rate": 9.902651538044814e-05,
+      "loss": 1.8789,
+      "step": 4460
+    },
+    {
+      "epoch": 0.553386627006259,
+      "grad_norm": 0.13366684428179468,
+      "learning_rate": 9.88101947456044e-05,
+      "loss": 1.8629,
+      "step": 4465
+    },
+    {
+      "epoch": 0.5540063208774865,
+      "grad_norm": 0.13117562103130084,
+      "learning_rate": 9.859387967906033e-05,
+      "loss": 1.8375,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5546260147487141,
+      "grad_norm": 0.13152848331502962,
+      "learning_rate": 9.837757119317228e-05,
+      "loss": 1.8738,
+      "step": 4475
+    },
+    {
+      "epoch": 0.5552457086199417,
+      "grad_norm": 0.1290209815631564,
+      "learning_rate": 9.816127030026607e-05,
+      "loss": 1.8594,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5558654024911693,
+      "grad_norm": 0.1415132208460771,
+      "learning_rate": 9.794497801263185e-05,
+      "loss": 1.8383,
+      "step": 4485
+    },
+    {
+      "epoch": 0.556485096362397,
+      "grad_norm": 0.13215668855742108,
+      "learning_rate": 9.772869534251956e-05,
+      "loss": 1.8262,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5571047902336246,
+      "grad_norm": 0.13620380571406365,
+      "learning_rate": 9.751242330213407e-05,
+      "loss": 1.8852,
+      "step": 4495
+    },
+    {
+      "epoch": 0.5577244841048522,
+      "grad_norm": 0.1382439074725802,
+      "learning_rate": 9.729616290363051e-05,
+      "loss": 1.8473,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5583441779760798,
+      "grad_norm": 0.13763527468041403,
+      "learning_rate": 9.707991515910957e-05,
+      "loss": 1.8625,
+      "step": 4505
+    },
+    {
+      "epoch": 0.5589638718473074,
+      "grad_norm": 0.1387628426822122,
+      "learning_rate": 9.686368108061268e-05,
+      "loss": 1.9066,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5595835657185351,
+      "grad_norm": 0.1284143925815832,
+      "learning_rate": 9.664746168011734e-05,
+      "loss": 1.8227,
+      "step": 4515
+    },
+    {
+      "epoch": 0.5602032595897627,
+      "grad_norm": 0.13073807322245737,
+      "learning_rate": 9.643125796953233e-05,
+      "loss": 1.8734,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5608229534609903,
+      "grad_norm": 0.13788095056614413,
+      "learning_rate": 9.621507096069304e-05,
+      "loss": 1.8582,
+      "step": 4525
+    },
+    {
+      "epoch": 0.5614426473322179,
+      "grad_norm": 0.13116436340123142,
+      "learning_rate": 9.599890166535665e-05,
+      "loss": 1.8668,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5620623412034454,
+      "grad_norm": 0.1417381732399684,
+      "learning_rate": 9.578275109519744e-05,
+      "loss": 1.891,
+      "step": 4535
+    },
+    {
+      "epoch": 0.5626820350746731,
+      "grad_norm": 0.1378428491993231,
+      "learning_rate": 9.556662026180212e-05,
+      "loss": 1.8562,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5633017289459007,
+      "grad_norm": 0.13590235062753475,
+      "learning_rate": 9.535051017666497e-05,
+      "loss": 1.877,
+      "step": 4545
+    },
+    {
+      "epoch": 0.5639214228171283,
+      "grad_norm": 0.13282424118769054,
+      "learning_rate": 9.513442185118319e-05,
+      "loss": 1.8711,
+      "step": 4550
+    },
+    {
+      "epoch": 0.5645411166883559,
+      "grad_norm": 0.1329650614632228,
+      "learning_rate": 9.491835629665214e-05,
+      "loss": 1.8375,
+      "step": 4555
+    },
+    {
+      "epoch": 0.5651608105595836,
+      "grad_norm": 0.13566530561601037,
+      "learning_rate": 9.470231452426059e-05,
+      "loss": 1.7836,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5657805044308112,
+      "grad_norm": 0.13493729363122003,
+      "learning_rate": 9.448629754508607e-05,
+      "loss": 1.8703,
+      "step": 4565
+    },
+    {
+      "epoch": 0.5664001983020388,
+      "grad_norm": 0.1397619680131406,
+      "learning_rate": 9.427030637009003e-05,
+      "loss": 1.8855,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5670198921732664,
+      "grad_norm": 0.13164387557122,
+      "learning_rate": 9.405434201011313e-05,
+      "loss": 1.8969,
+      "step": 4575
+    },
+    {
+      "epoch": 0.567639586044494,
+      "grad_norm": 0.12900052483979663,
+      "learning_rate": 9.383840547587064e-05,
+      "loss": 1.8848,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5682592799157217,
+      "grad_norm": 0.1345646794503773,
+      "learning_rate": 9.362249777794749e-05,
+      "loss": 1.8422,
+      "step": 4585
+    },
+    {
+      "epoch": 0.5688789737869493,
+      "grad_norm": 0.1334915488354972,
+      "learning_rate": 9.340661992679369e-05,
+      "loss": 1.8246,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5694986676581769,
+      "grad_norm": 0.13798209662252225,
+      "learning_rate": 9.319077293271961e-05,
+      "loss": 1.8746,
+      "step": 4595
+    },
+    {
+      "epoch": 0.5701183615294044,
+      "grad_norm": 0.13590886596850502,
+      "learning_rate": 9.297495780589118e-05,
+      "loss": 1.8484,
+      "step": 4600
+    },
+    {
+      "epoch": 0.570738055400632,
+      "grad_norm": 0.13290713678985736,
+      "learning_rate": 9.27591755563252e-05,
+      "loss": 1.766,
+      "step": 4605
+    },
+    {
+      "epoch": 0.5713577492718597,
+      "grad_norm": 0.12641011719135234,
+      "learning_rate": 9.254342719388454e-05,
+      "loss": 1.7781,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5719774431430873,
+      "grad_norm": 0.14121891454527805,
+      "learning_rate": 9.232771372827356e-05,
+      "loss": 1.877,
+      "step": 4615
+    },
+    {
+      "epoch": 0.5725971370143149,
+      "grad_norm": 0.13878039672119347,
+      "learning_rate": 9.211203616903328e-05,
+      "loss": 1.8391,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5732168308855425,
+      "grad_norm": 0.1266251624297058,
+      "learning_rate": 9.189639552553667e-05,
+      "loss": 1.8941,
+      "step": 4625
+    },
+    {
+      "epoch": 0.5738365247567702,
+      "grad_norm": 0.14463093120155013,
+      "learning_rate": 9.168079280698391e-05,
+      "loss": 1.8234,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5744562186279978,
+      "grad_norm": 0.1322360423319378,
+      "learning_rate": 9.146522902239781e-05,
+      "loss": 1.85,
+      "step": 4635
+    },
+    {
+      "epoch": 0.5750759124992254,
+      "grad_norm": 0.1298579011473491,
+      "learning_rate": 9.124970518061877e-05,
+      "loss": 1.8664,
+      "step": 4640
+    },
+    {
+      "epoch": 0.575695606370453,
+      "grad_norm": 0.13432468619740123,
+      "learning_rate": 9.103422229030038e-05,
+      "loss": 1.8441,
+      "step": 4645
+    },
+    {
+      "epoch": 0.5763153002416807,
+      "grad_norm": 0.13199808903312424,
+      "learning_rate": 9.081878135990458e-05,
+      "loss": 1.7465,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5769349941129083,
+      "grad_norm": 0.1348707902460552,
+      "learning_rate": 9.06033833976969e-05,
+      "loss": 1.8445,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5775546879841358,
+      "grad_norm": 0.1323248488216555,
+      "learning_rate": 9.038802941174187e-05,
+      "loss": 1.7797,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5781743818553634,
+      "grad_norm": 0.13490844471066665,
+      "learning_rate": 9.017272040989804e-05,
+      "loss": 1.8605,
+      "step": 4665
+    },
+    {
+      "epoch": 0.578794075726591,
+      "grad_norm": 0.129326446856862,
+      "learning_rate": 8.995745739981355e-05,
+      "loss": 1.8629,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5794137695978187,
+      "grad_norm": 0.1405323300926802,
+      "learning_rate": 8.974224138892127e-05,
+      "loss": 1.852,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5800334634690463,
+      "grad_norm": 0.13103817592164685,
+      "learning_rate": 8.952707338443418e-05,
+      "loss": 1.8113,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5806531573402739,
+      "grad_norm": 0.1370095616856299,
+      "learning_rate": 8.931195439334048e-05,
+      "loss": 1.8484,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5812728512115015,
+      "grad_norm": 0.13307208775504528,
+      "learning_rate": 8.90968854223991e-05,
+      "loss": 1.8207,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5818925450827291,
+      "grad_norm": 0.13560010132734596,
+      "learning_rate": 8.888186747813473e-05,
+      "loss": 1.8574,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5825122389539568,
+      "grad_norm": 0.13395887302734633,
+      "learning_rate": 8.866690156683332e-05,
+      "loss": 1.827,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5831319328251844,
+      "grad_norm": 0.13429729678044644,
+      "learning_rate": 8.845198869453742e-05,
+      "loss": 1.8605,
+      "step": 4705
+    },
+    {
+      "epoch": 0.583751626696412,
+      "grad_norm": 0.1363221017644863,
+      "learning_rate": 8.823712986704121e-05,
+      "loss": 1.7816,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5843713205676396,
+      "grad_norm": 0.1335714214307853,
+      "learning_rate": 8.802232608988604e-05,
+      "loss": 1.8316,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5849910144388673,
+      "grad_norm": 0.15216425952602824,
+      "learning_rate": 8.780757836835544e-05,
+      "loss": 1.8223,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5856107083100948,
+      "grad_norm": 0.13275202428606528,
+      "learning_rate": 8.759288770747087e-05,
+      "loss": 1.8527,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5862304021813224,
+      "grad_norm": 0.1549023027007513,
+      "learning_rate": 8.737825511198654e-05,
+      "loss": 1.8434,
+      "step": 4730
+    },
+    {
+      "epoch": 0.58685009605255,
+      "grad_norm": 0.14026036781473247,
+      "learning_rate": 8.7163681586385e-05,
+      "loss": 1.8254,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5874697899237776,
+      "grad_norm": 0.13233339755282925,
+      "learning_rate": 8.694916813487233e-05,
+      "loss": 1.834,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5880894837950053,
+      "grad_norm": 0.13863065649650821,
+      "learning_rate": 8.67347157613735e-05,
+      "loss": 1.7969,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5887091776662329,
+      "grad_norm": 0.13446458501695888,
+      "learning_rate": 8.652032546952754e-05,
+      "loss": 1.8039,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5893288715374605,
+      "grad_norm": 0.1349155466617635,
+      "learning_rate": 8.630599826268303e-05,
+      "loss": 1.9152,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5899485654086881,
+      "grad_norm": 0.1364181833568848,
+      "learning_rate": 8.609173514389328e-05,
+      "loss": 1.8367,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5905682592799157,
+      "grad_norm": 0.13658301611381946,
+      "learning_rate": 8.587753711591166e-05,
+      "loss": 1.7961,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5911879531511434,
+      "grad_norm": 0.13948936070832027,
+      "learning_rate": 8.566340518118695e-05,
+      "loss": 1.8777,
+      "step": 4770
+    },
+    {
+      "epoch": 0.591807647022371,
+      "grad_norm": 0.13193053858601933,
+      "learning_rate": 8.544934034185854e-05,
+      "loss": 1.8289,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5924273408935986,
+      "grad_norm": 0.1291550431062787,
+      "learning_rate": 8.523534359975189e-05,
+      "loss": 1.8105,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5930470347648262,
+      "grad_norm": 0.1337244591947349,
+      "learning_rate": 8.502141595637371e-05,
+      "loss": 1.8254,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5936667286360537,
+      "grad_norm": 0.13532142100296568,
+      "learning_rate": 8.480755841290736e-05,
+      "loss": 1.8703,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5942864225072814,
+      "grad_norm": 0.1377470717648544,
+      "learning_rate": 8.459377197020813e-05,
+      "loss": 1.8426,
+      "step": 4795
+    },
+    {
+      "epoch": 0.594906116378509,
+      "grad_norm": 0.13709338754981154,
+      "learning_rate": 8.438005762879856e-05,
+      "loss": 1.8359,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5955258102497366,
+      "grad_norm": 0.13560842681432764,
+      "learning_rate": 8.416641638886369e-05,
+      "loss": 1.8703,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5961455041209642,
+      "grad_norm": 0.1419275087886935,
+      "learning_rate": 8.395284925024654e-05,
+      "loss": 1.8141,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5967651979921919,
+      "grad_norm": 0.14348376755535794,
+      "learning_rate": 8.373935721244329e-05,
+      "loss": 1.8414,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5973848918634195,
+      "grad_norm": 0.13863281114527268,
+      "learning_rate": 8.352594127459865e-05,
+      "loss": 1.7605,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5980045857346471,
+      "grad_norm": 0.1374738138416099,
+      "learning_rate": 8.331260243550119e-05,
+      "loss": 1.8258,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5986242796058747,
+      "grad_norm": 0.13038481110656677,
+      "learning_rate": 8.309934169357862e-05,
+      "loss": 1.8102,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5992439734771023,
+      "grad_norm": 0.13222459650512203,
+      "learning_rate": 8.28861600468932e-05,
+      "loss": 1.7926,
+      "step": 4835
+    },
+    {
+      "epoch": 0.59986366734833,
+      "grad_norm": 0.1341387150049404,
+      "learning_rate": 8.267305849313702e-05,
+      "loss": 1.8582,
+      "step": 4840
+    },
+    {
+      "epoch": 0.6004833612195576,
+      "grad_norm": 0.14125522067847465,
+      "learning_rate": 8.246003802962732e-05,
+      "loss": 1.7863,
+      "step": 4845
+    },
+    {
+      "epoch": 0.6011030550907851,
+      "grad_norm": 0.13606861615713078,
+      "learning_rate": 8.224709965330182e-05,
+      "loss": 1.916,
+      "step": 4850
+    },
+    {
+      "epoch": 0.6017227489620127,
+      "grad_norm": 0.14156145445397822,
+      "learning_rate": 8.203424436071413e-05,
+      "loss": 1.8277,
+      "step": 4855
+    },
+    {
+      "epoch": 0.6023424428332403,
+      "grad_norm": 0.13589787935379263,
+      "learning_rate": 8.182147314802892e-05,
+      "loss": 1.8293,
+      "step": 4860
+    },
+    {
+      "epoch": 0.602962136704468,
+      "grad_norm": 0.13209541273739087,
+      "learning_rate": 8.160878701101751e-05,
+      "loss": 1.8113,
+      "step": 4865
+    },
+    {
+      "epoch": 0.6035818305756956,
+      "grad_norm": 0.1422468671145226,
+      "learning_rate": 8.139618694505292e-05,
+      "loss": 1.7863,
+      "step": 4870
+    },
+    {
+      "epoch": 0.6042015244469232,
+      "grad_norm": 0.13502272532252504,
+      "learning_rate": 8.118367394510544e-05,
+      "loss": 1.8195,
+      "step": 4875
+    },
+    {
+      "epoch": 0.6048212183181508,
+      "grad_norm": 0.1288915778981244,
+      "learning_rate": 8.097124900573795e-05,
+      "loss": 1.7996,
+      "step": 4880
+    },
+    {
+      "epoch": 0.6054409121893785,
+      "grad_norm": 0.13113784023771752,
+      "learning_rate": 8.075891312110104e-05,
+      "loss": 1.8668,
+      "step": 4885
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 0.13594399497080922,
+      "learning_rate": 8.054666728492864e-05,
+      "loss": 1.818,
+      "step": 4890
+    },
+    {
+      "epoch": 0.6066802999318337,
+      "grad_norm": 0.13803949866445053,
+      "learning_rate": 8.033451249053324e-05,
+      "loss": 1.8051,
+      "step": 4895
+    },
+    {
+      "epoch": 0.6072999938030613,
+      "grad_norm": 0.1319319227151321,
+      "learning_rate": 8.01224497308012e-05,
+      "loss": 1.9059,
+      "step": 4900
+    },
+    {
+      "epoch": 0.6079196876742889,
+      "grad_norm": 0.13577815653327518,
+      "learning_rate": 7.991047999818825e-05,
+      "loss": 1.8473,
+      "step": 4905
+    },
+    {
+      "epoch": 0.6085393815455166,
+      "grad_norm": 0.13549175693333843,
+      "learning_rate": 7.969860428471472e-05,
+      "loss": 1.8254,
+      "step": 4910
+    },
+    {
+      "epoch": 0.6091590754167441,
+      "grad_norm": 0.13452651891535064,
+      "learning_rate": 7.948682358196081e-05,
+      "loss": 1.8109,
+      "step": 4915
+    },
+    {
+      "epoch": 0.6097787692879717,
+      "grad_norm": 0.13296839704090957,
+      "learning_rate": 7.927513888106222e-05,
+      "loss": 1.8684,
+      "step": 4920
+    },
+    {
+      "epoch": 0.6103984631591993,
+      "grad_norm": 0.13298206942158278,
+      "learning_rate": 7.90635511727053e-05,
+      "loss": 1.8617,
+      "step": 4925
+    },
+    {
+      "epoch": 0.611018157030427,
+      "grad_norm": 0.13471862453317543,
+      "learning_rate": 7.88520614471225e-05,
+      "loss": 1.8207,
+      "step": 4930
+    },
+    {
+      "epoch": 0.6116378509016546,
+      "grad_norm": 0.1356902660201222,
+      "learning_rate": 7.864067069408773e-05,
+      "loss": 1.841,
+      "step": 4935
+    },
+    {
+      "epoch": 0.6122575447728822,
+      "grad_norm": 0.13190339436344312,
+      "learning_rate": 7.842937990291157e-05,
+      "loss": 1.7984,
+      "step": 4940
+    },
+    {
+      "epoch": 0.6128772386441098,
+      "grad_norm": 0.13568034294738784,
+      "learning_rate": 7.821819006243691e-05,
+      "loss": 1.8543,
+      "step": 4945
+    },
+    {
+      "epoch": 0.6134969325153374,
+      "grad_norm": 0.1417235964582138,
+      "learning_rate": 7.80071021610342e-05,
+      "loss": 1.9066,
+      "step": 4950
+    },
+    {
+      "epoch": 0.6141166263865651,
+      "grad_norm": 0.13549923024658295,
+      "learning_rate": 7.779611718659674e-05,
+      "loss": 1.8219,
+      "step": 4955
+    },
+    {
+      "epoch": 0.6147363202577927,
+      "grad_norm": 0.13335577403941054,
+      "learning_rate": 7.758523612653617e-05,
+      "loss": 1.7711,
+      "step": 4960
+    },
+    {
+      "epoch": 0.6153560141290203,
+      "grad_norm": 0.13770283681260875,
+      "learning_rate": 7.737445996777784e-05,
+      "loss": 1.8008,
+      "step": 4965
+    },
+    {
+      "epoch": 0.6159757080002479,
+      "grad_norm": 0.1350132818669121,
+      "learning_rate": 7.716378969675599e-05,
+      "loss": 1.8332,
+      "step": 4970
+    },
+    {
+      "epoch": 0.6165954018714755,
+      "grad_norm": 0.1379620012305667,
+      "learning_rate": 7.695322629940957e-05,
+      "loss": 1.8152,
+      "step": 4975
+    },
+    {
+      "epoch": 0.6172150957427031,
+      "grad_norm": 0.14223438579861916,
+      "learning_rate": 7.674277076117718e-05,
+      "loss": 1.8676,
+      "step": 4980
+    },
+    {
+      "epoch": 0.6178347896139307,
+      "grad_norm": 0.13631387246079188,
+      "learning_rate": 7.653242406699267e-05,
+      "loss": 1.8195,
+      "step": 4985
+    },
+    {
+      "epoch": 0.6184544834851583,
+      "grad_norm": 0.13139403130867852,
+      "learning_rate": 7.632218720128056e-05,
+      "loss": 1.8789,
+      "step": 4990
+    },
+    {
+      "epoch": 0.6190741773563859,
+      "grad_norm": 0.1333637361950243,
+      "learning_rate": 7.611206114795126e-05,
+      "loss": 1.8102,
+      "step": 4995
+    },
+    {
+      "epoch": 0.6196938712276135,
+      "grad_norm": 0.13724109963380457,
+      "learning_rate": 7.590204689039671e-05,
+      "loss": 1.8473,
+      "step": 5000
+    },
+    {
+      "epoch": 0.6203135650988412,
+      "grad_norm": 0.13124493232609577,
+      "learning_rate": 7.569214541148551e-05,
+      "loss": 1.8051,
+      "step": 5005
+    },
+    {
+      "epoch": 0.6209332589700688,
+      "grad_norm": 0.13887359555527148,
+      "learning_rate": 7.548235769355858e-05,
+      "loss": 1.8527,
+      "step": 5010
+    },
+    {
+      "epoch": 0.6215529528412964,
+      "grad_norm": 0.14358380718074817,
+      "learning_rate": 7.527268471842436e-05,
+      "loss": 1.8605,
+      "step": 5015
+    },
+    {
+      "epoch": 0.622172646712524,
+      "grad_norm": 0.137106352425573,
+      "learning_rate": 7.506312746735432e-05,
+      "loss": 1.8133,
+      "step": 5020
+    },
+    {
+      "epoch": 0.6227923405837517,
+      "grad_norm": 0.1397418008866006,
+      "learning_rate": 7.48536869210783e-05,
+      "loss": 1.8531,
+      "step": 5025
+    },
+    {
+      "epoch": 0.6234120344549793,
+      "grad_norm": 0.1297107756247531,
+      "learning_rate": 7.464436405978002e-05,
+      "loss": 1.8203,
+      "step": 5030
+    },
+    {
+      "epoch": 0.6240317283262069,
+      "grad_norm": 0.13710425450182642,
+      "learning_rate": 7.44351598630924e-05,
+      "loss": 1.8586,
+      "step": 5035
+    },
+    {
+      "epoch": 0.6246514221974344,
+      "grad_norm": 0.1317466014455055,
+      "learning_rate": 7.422607531009302e-05,
+      "loss": 1.8246,
+      "step": 5040
+    },
+    {
+      "epoch": 0.625271116068662,
+      "grad_norm": 0.13346546234173723,
+      "learning_rate": 7.401711137929955e-05,
+      "loss": 1.8824,
+      "step": 5045
+    },
+    {
+      "epoch": 0.6258908099398897,
+      "grad_norm": 0.12791424524198686,
+      "learning_rate": 7.380826904866504e-05,
+      "loss": 1.8742,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6265105038111173,
+      "grad_norm": 0.13961087823065946,
+      "learning_rate": 7.35995492955736e-05,
+      "loss": 1.8375,
+      "step": 5055
+    },
+    {
+      "epoch": 0.6271301976823449,
+      "grad_norm": 0.1294602335761151,
+      "learning_rate": 7.339095309683557e-05,
+      "loss": 1.852,
+      "step": 5060
+    },
+    {
+      "epoch": 0.6277498915535725,
+      "grad_norm": 0.1335668188096444,
+      "learning_rate": 7.31824814286831e-05,
+      "loss": 1.8281,
+      "step": 5065
+    },
+    {
+      "epoch": 0.6283695854248001,
+      "grad_norm": 0.1457801204886514,
+      "learning_rate": 7.297413526676553e-05,
+      "loss": 1.7699,
+      "step": 5070
+    },
+    {
+      "epoch": 0.6289892792960278,
+      "grad_norm": 0.1333202164334989,
+      "learning_rate": 7.276591558614482e-05,
+      "loss": 1.8578,
+      "step": 5075
+    },
+    {
+      "epoch": 0.6296089731672554,
+      "grad_norm": 0.12964753000237725,
+      "learning_rate": 7.2557823361291e-05,
+      "loss": 1.8234,
+      "step": 5080
+    },
+    {
+      "epoch": 0.630228667038483,
+      "grad_norm": 0.1335935100958171,
+      "learning_rate": 7.23498595660776e-05,
+      "loss": 1.8457,
+      "step": 5085
+    },
+    {
+      "epoch": 0.6308483609097106,
+      "grad_norm": 0.13566562420556766,
+      "learning_rate": 7.21420251737771e-05,
+      "loss": 1.8629,
+      "step": 5090
+    },
+    {
+      "epoch": 0.6314680547809383,
+      "grad_norm": 0.1350181379111902,
+      "learning_rate": 7.19343211570564e-05,
+      "loss": 1.8711,
+      "step": 5095
+    },
+    {
+      "epoch": 0.6320877486521659,
+      "grad_norm": 0.13697942997561122,
+      "learning_rate": 7.172674848797219e-05,
+      "loss": 1.7922,
+      "step": 5100
+    },
+    {
+      "epoch": 0.6327074425233934,
+      "grad_norm": 0.14031317331319004,
+      "learning_rate": 7.151930813796655e-05,
+      "loss": 1.8117,
+      "step": 5105
+    },
+    {
+      "epoch": 0.633327136394621,
+      "grad_norm": 0.13755025256595127,
+      "learning_rate": 7.131200107786214e-05,
+      "loss": 1.8062,
+      "step": 5110
+    },
+    {
+      "epoch": 0.6339468302658486,
+      "grad_norm": 0.13894392598493008,
+      "learning_rate": 7.110482827785796e-05,
+      "loss": 1.7805,
+      "step": 5115
+    },
+    {
+      "epoch": 0.6345665241370763,
+      "grad_norm": 0.14482768933598048,
+      "learning_rate": 7.089779070752463e-05,
+      "loss": 1.798,
+      "step": 5120
+    },
+    {
+      "epoch": 0.6351862180083039,
+      "grad_norm": 0.14158917920715833,
+      "learning_rate": 7.069088933579988e-05,
+      "loss": 1.8164,
+      "step": 5125
+    },
+    {
+      "epoch": 0.6358059118795315,
+      "grad_norm": 0.1353622024235782,
+      "learning_rate": 7.048412513098412e-05,
+      "loss": 1.8758,
+      "step": 5130
+    },
+    {
+      "epoch": 0.6364256057507591,
+      "grad_norm": 0.13637740239228766,
+      "learning_rate": 7.027749906073564e-05,
+      "loss": 1.7945,
+      "step": 5135
+    },
+    {
+      "epoch": 0.6370452996219867,
+      "grad_norm": 0.1388858528887692,
+      "learning_rate": 7.007101209206639e-05,
+      "loss": 1.8609,
+      "step": 5140
+    },
+    {
+      "epoch": 0.6376649934932144,
+      "grad_norm": 0.13169802010265266,
+      "learning_rate": 6.98646651913373e-05,
+      "loss": 1.8457,
+      "step": 5145
+    },
+    {
+      "epoch": 0.638284687364442,
+      "grad_norm": 0.1432891542870216,
+      "learning_rate": 6.965845932425377e-05,
+      "loss": 1.8676,
+      "step": 5150
+    },
+    {
+      "epoch": 0.6389043812356696,
+      "grad_norm": 0.13521941306370572,
+      "learning_rate": 6.945239545586117e-05,
+      "loss": 1.8477,
+      "step": 5155
+    },
+    {
+      "epoch": 0.6395240751068972,
+      "grad_norm": 0.13771072724035743,
+      "learning_rate": 6.924647455054036e-05,
+      "loss": 1.7895,
+      "step": 5160
+    },
+    {
+      "epoch": 0.6401437689781249,
+      "grad_norm": 0.135151419363287,
+      "learning_rate": 6.9040697572003e-05,
+      "loss": 1.8367,
+      "step": 5165
+    },
+    {
+      "epoch": 0.6407634628493524,
+      "grad_norm": 0.1384020407161956,
+      "learning_rate": 6.883506548328724e-05,
+      "loss": 1.8074,
+      "step": 5170
+    },
+    {
+      "epoch": 0.64138315672058,
+      "grad_norm": 0.14469575370154378,
+      "learning_rate": 6.862957924675323e-05,
+      "loss": 1.8559,
+      "step": 5175
+    },
+    {
+      "epoch": 0.6420028505918076,
+      "grad_norm": 0.1359705775673227,
+      "learning_rate": 6.842423982407842e-05,
+      "loss": 1.8418,
+      "step": 5180
+    },
+    {
+      "epoch": 0.6426225444630352,
+      "grad_norm": 0.1381154848467315,
+      "learning_rate": 6.821904817625326e-05,
+      "loss": 1.8617,
+      "step": 5185
+    },
+    {
+      "epoch": 0.6432422383342629,
+      "grad_norm": 0.13127462304573068,
+      "learning_rate": 6.801400526357647e-05,
+      "loss": 1.8523,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6438619322054905,
+      "grad_norm": 0.14055038700162212,
+      "learning_rate": 6.78091120456508e-05,
+      "loss": 1.7988,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6444816260767181,
+      "grad_norm": 0.13763358948089932,
+      "learning_rate": 6.760436948137843e-05,
+      "loss": 1.8621,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6451013199479457,
+      "grad_norm": 0.13962503302729862,
+      "learning_rate": 6.739977852895647e-05,
+      "loss": 1.8379,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6457210138191734,
+      "grad_norm": 0.14185188312406627,
+      "learning_rate": 6.719534014587243e-05,
+      "loss": 1.8309,
+      "step": 5210
+    },
+    {
+      "epoch": 0.646340707690401,
+      "grad_norm": 0.13486063326929387,
+      "learning_rate": 6.699105528889988e-05,
+      "loss": 1.873,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6469604015616286,
+      "grad_norm": 0.1362204464921119,
+      "learning_rate": 6.678692491409378e-05,
+      "loss": 1.8094,
+      "step": 5220
+    },
+    {
+      "epoch": 0.6475800954328562,
+      "grad_norm": 0.1336262838647592,
+      "learning_rate": 6.658294997678621e-05,
+      "loss": 1.8859,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6481997893040837,
+      "grad_norm": 0.1426175317216606,
+      "learning_rate": 6.637913143158175e-05,
+      "loss": 1.8219,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6488194831753114,
+      "grad_norm": 0.13139603334052216,
+      "learning_rate": 6.617547023235309e-05,
+      "loss": 1.8805,
+      "step": 5235
+    },
+    {
+      "epoch": 0.649439177046539,
+      "grad_norm": 0.13499721675068688,
+      "learning_rate": 6.597196733223651e-05,
+      "loss": 1.8562,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6500588709177666,
+      "grad_norm": 0.14124375108338427,
+      "learning_rate": 6.576862368362747e-05,
+      "loss": 1.8074,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6506785647889942,
+      "grad_norm": 0.13385751122133496,
+      "learning_rate": 6.556544023817613e-05,
+      "loss": 1.9098,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6512982586602218,
+      "grad_norm": 0.13027632164466119,
+      "learning_rate": 6.536241794678288e-05,
+      "loss": 1.8102,
+      "step": 5255
+    },
+    {
+      "epoch": 0.6519179525314495,
+      "grad_norm": 0.12988529012895247,
+      "learning_rate": 6.515955775959394e-05,
+      "loss": 1.8711,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6525376464026771,
+      "grad_norm": 0.13882339488018036,
+      "learning_rate": 6.495686062599684e-05,
+      "loss": 1.7949,
+      "step": 5265
+    },
+    {
+      "epoch": 0.6531573402739047,
+      "grad_norm": 0.133952007292679,
+      "learning_rate": 6.475432749461607e-05,
+      "loss": 1.8203,
+      "step": 5270
+    },
+    {
+      "epoch": 0.6537770341451323,
+      "grad_norm": 0.1348251697867989,
+      "learning_rate": 6.455195931330855e-05,
+      "loss": 1.8379,
+      "step": 5275
+    },
+    {
+      "epoch": 0.65439672801636,
+      "grad_norm": 0.13015983597998307,
+      "learning_rate": 6.43497570291592e-05,
+      "loss": 1.8641,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6550164218875876,
+      "grad_norm": 0.14151362596069492,
+      "learning_rate": 6.414772158847661e-05,
+      "loss": 1.8773,
+      "step": 5285
+    },
+    {
+      "epoch": 0.6556361157588152,
+      "grad_norm": 0.1392005904812876,
+      "learning_rate": 6.394585393678851e-05,
+      "loss": 1.8027,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6562558096300427,
+      "grad_norm": 0.13728007241822773,
+      "learning_rate": 6.374415501883741e-05,
+      "loss": 1.8016,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6568755035012703,
+      "grad_norm": 0.13755360326233745,
+      "learning_rate": 6.354262577857606e-05,
+      "loss": 1.8641,
+      "step": 5300
+    },
+    {
+      "epoch": 0.657495197372498,
+      "grad_norm": 0.1407832990027108,
+      "learning_rate": 6.334126715916318e-05,
+      "loss": 1.8004,
+      "step": 5305
+    },
+    {
+      "epoch": 0.6581148912437256,
+      "grad_norm": 0.13848872290550496,
+      "learning_rate": 6.314008010295897e-05,
+      "loss": 1.7961,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6587345851149532,
+      "grad_norm": 0.13814509728706104,
+      "learning_rate": 6.293906555152072e-05,
+      "loss": 1.8469,
+      "step": 5315
+    },
+    {
+      "epoch": 0.6593542789861808,
+      "grad_norm": 0.1310131249847111,
+      "learning_rate": 6.273822444559839e-05,
+      "loss": 1.8582,
+      "step": 5320
+    },
+    {
+      "epoch": 0.6599739728574084,
+      "grad_norm": 0.13818644699764637,
+      "learning_rate": 6.253755772513024e-05,
+      "loss": 1.7691,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6605936667286361,
+      "grad_norm": 0.13439005537004714,
+      "learning_rate": 6.233706632923832e-05,
+      "loss": 1.8617,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6612133605998637,
+      "grad_norm": 0.14117004629994614,
+      "learning_rate": 6.213675119622425e-05,
+      "loss": 1.8156,
+      "step": 5335
+    },
+    {
+      "epoch": 0.6618330544710913,
+      "grad_norm": 0.13441504473083238,
+      "learning_rate": 6.19366132635647e-05,
+      "loss": 1.8191,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6624527483423189,
+      "grad_norm": 0.1376510356237994,
+      "learning_rate": 6.173665346790704e-05,
+      "loss": 1.9027,
+      "step": 5345
+    },
+    {
+      "epoch": 0.6630724422135466,
+      "grad_norm": 0.13260573599872516,
+      "learning_rate": 6.153687274506501e-05,
+      "loss": 1.8062,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6636921360847742,
+      "grad_norm": 0.14452496979165114,
+      "learning_rate": 6.133727203001415e-05,
+      "loss": 1.8031,
+      "step": 5355
+    },
+    {
+      "epoch": 0.6643118299560017,
+      "grad_norm": 0.13310538735250482,
+      "learning_rate": 6.113785225688772e-05,
+      "loss": 1.8523,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6649315238272293,
+      "grad_norm": 0.14020085644370098,
+      "learning_rate": 6.093861435897208e-05,
+      "loss": 1.8535,
+      "step": 5365
+    },
+    {
+      "epoch": 0.6655512176984569,
+      "grad_norm": 0.13281311353444464,
+      "learning_rate": 6.073955926870243e-05,
+      "loss": 1.8172,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6661709115696846,
+      "grad_norm": 0.14163086589006268,
+      "learning_rate": 6.0540687917658445e-05,
+      "loss": 1.7875,
+      "step": 5375
+    },
+    {
+      "epoch": 0.6667906054409122,
+      "grad_norm": 0.1311369179551235,
+      "learning_rate": 6.034200123655993e-05,
+      "loss": 1.8496,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6674102993121398,
+      "grad_norm": 0.13835525169206805,
+      "learning_rate": 6.01435001552623e-05,
+      "loss": 1.8555,
+      "step": 5385
+    },
+    {
+      "epoch": 0.6680299931833674,
+      "grad_norm": 0.13402711274212736,
+      "learning_rate": 5.9945185602752496e-05,
+      "loss": 1.8148,
+      "step": 5390
+    },
+    {
+      "epoch": 0.668649687054595,
+      "grad_norm": 0.1406576203494184,
+      "learning_rate": 5.974705850714444e-05,
+      "loss": 1.8578,
+      "step": 5395
+    },
+    {
+      "epoch": 0.6692693809258227,
+      "grad_norm": 0.1334842815628208,
+      "learning_rate": 5.95491197956748e-05,
+      "loss": 1.8332,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6698890747970503,
+      "grad_norm": 0.13248412735805298,
+      "learning_rate": 5.9351370394698604e-05,
+      "loss": 1.8344,
+      "step": 5405
+    },
+    {
+      "epoch": 0.6705087686682779,
+      "grad_norm": 0.13323431139084796,
+      "learning_rate": 5.9153811229684794e-05,
+      "loss": 1.8582,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6711284625395055,
+      "grad_norm": 0.13598450846192175,
+      "learning_rate": 5.895644322521212e-05,
+      "loss": 1.8555,
+      "step": 5415
+    },
+    {
+      "epoch": 0.671748156410733,
+      "grad_norm": 0.13924831817973832,
+      "learning_rate": 5.875926730496471e-05,
+      "loss": 1.8695,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6723678502819607,
+      "grad_norm": 0.13200834015914215,
+      "learning_rate": 5.856228439172764e-05,
+      "loss": 1.9035,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6729875441531883,
+      "grad_norm": 0.13431588152455146,
+      "learning_rate": 5.836549540738281e-05,
+      "loss": 1.827,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6736072380244159,
+      "grad_norm": 0.13074786666140048,
+      "learning_rate": 5.816890127290446e-05,
+      "loss": 1.8262,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6742269318956435,
+      "grad_norm": 0.14226796953838972,
+      "learning_rate": 5.7972502908354954e-05,
+      "loss": 1.8348,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6748466257668712,
+      "grad_norm": 0.13089302513048215,
+      "learning_rate": 5.777630123288046e-05,
+      "loss": 1.8863,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6754663196380988,
+      "grad_norm": 0.13586048592762726,
+      "learning_rate": 5.758029716470664e-05,
+      "loss": 1.8871,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6760860135093264,
+      "grad_norm": 0.1445088669511517,
+      "learning_rate": 5.738449162113435e-05,
+      "loss": 1.8043,
+      "step": 5455
+    },
+    {
+      "epoch": 0.676705707380554,
+      "grad_norm": 0.1396596555264478,
+      "learning_rate": 5.7188885518535365e-05,
+      "loss": 1.8555,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6773254012517816,
+      "grad_norm": 0.1356062175747125,
+      "learning_rate": 5.699347977234799e-05,
+      "loss": 1.8465,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6779450951230093,
+      "grad_norm": 0.14319873842375655,
+      "learning_rate": 5.679827529707295e-05,
+      "loss": 1.7801,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6785647889942369,
+      "grad_norm": 0.13252156474119606,
+      "learning_rate": 5.660327300626902e-05,
+      "loss": 1.873,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6791844828654645,
+      "grad_norm": 0.14296503617575715,
+      "learning_rate": 5.640847381254869e-05,
+      "loss": 1.8297,
+      "step": 5480
+    },
+    {
+      "epoch": 0.679804176736692,
+      "grad_norm": 0.13214280074229467,
+      "learning_rate": 5.621387862757397e-05,
+      "loss": 1.8082,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6804238706079196,
+      "grad_norm": 0.13982102437227564,
+      "learning_rate": 5.6019488362052255e-05,
+      "loss": 1.8121,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6810435644791473,
+      "grad_norm": 0.1354634885782795,
+      "learning_rate": 5.582530392573164e-05,
+      "loss": 1.823,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6816632583503749,
+      "grad_norm": 0.13692962283590215,
+      "learning_rate": 5.563132622739713e-05,
+      "loss": 1.8406,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6822829522216025,
+      "grad_norm": 0.1344469592023015,
+      "learning_rate": 5.5437556174866156e-05,
+      "loss": 1.7934,
+      "step": 5505
+    },
+    {
+      "epoch": 0.6829026460928301,
+      "grad_norm": 0.1360058691875614,
+      "learning_rate": 5.5243994674984345e-05,
+      "loss": 1.8754,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6835223399640578,
+      "grad_norm": 0.14351473803114279,
+      "learning_rate": 5.505064263362136e-05,
+      "loss": 1.866,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6841420338352854,
+      "grad_norm": 0.1322311208239839,
+      "learning_rate": 5.485750095566644e-05,
+      "loss": 1.9109,
+      "step": 5520
+    },
+    {
+      "epoch": 0.684761727706513,
+      "grad_norm": 0.1384464211302108,
+      "learning_rate": 5.46645705450245e-05,
+      "loss": 1.8535,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6853814215777406,
+      "grad_norm": 0.13694935698109853,
+      "learning_rate": 5.447185230461156e-05,
+      "loss": 1.784,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6860011154489682,
+      "grad_norm": 0.13546497838866572,
+      "learning_rate": 5.427934713635088e-05,
+      "loss": 1.8625,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6866208093201959,
+      "grad_norm": 0.1400633356132821,
+      "learning_rate": 5.4087055941168384e-05,
+      "loss": 1.8258,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6872405031914235,
+      "grad_norm": 0.14126924542721914,
+      "learning_rate": 5.389497961898866e-05,
+      "loss": 1.8609,
+      "step": 5545
+    },
+    {
+      "epoch": 0.687860197062651,
+      "grad_norm": 0.1320352339912381,
+      "learning_rate": 5.370311906873062e-05,
+      "loss": 1.8625,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6884798909338786,
+      "grad_norm": 0.13076785413411318,
+      "learning_rate": 5.351147518830345e-05,
+      "loss": 1.8238,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6890995848051062,
+      "grad_norm": 0.13316624015952805,
+      "learning_rate": 5.3320048874602266e-05,
+      "loss": 1.8465,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6897192786763339,
+      "grad_norm": 0.13440609604282788,
+      "learning_rate": 5.3128841023504e-05,
+      "loss": 1.8742,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6903389725475615,
+      "grad_norm": 0.13545277719506474,
+      "learning_rate": 5.293785252986321e-05,
+      "loss": 1.8367,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6909586664187891,
+      "grad_norm": 0.1417270903396745,
+      "learning_rate": 5.274708428750765e-05,
+      "loss": 1.7941,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6915783602900167,
+      "grad_norm": 0.12988305752922388,
+      "learning_rate": 5.255653718923463e-05,
+      "loss": 1.8363,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6921980541612444,
+      "grad_norm": 0.13799787499371047,
+      "learning_rate": 5.236621212680628e-05,
+      "loss": 1.8273,
+      "step": 5585
+    },
+    {
+      "epoch": 0.692817748032472,
+      "grad_norm": 0.13054648174080216,
+      "learning_rate": 5.217610999094563e-05,
+      "loss": 1.8426,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6934374419036996,
+      "grad_norm": 0.1510525206781296,
+      "learning_rate": 5.1986231671332454e-05,
+      "loss": 1.8438,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6940571357749272,
+      "grad_norm": 0.1333950578838988,
+      "learning_rate": 5.179657805659908e-05,
+      "loss": 1.8633,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6946768296461548,
+      "grad_norm": 0.1349388448110292,
+      "learning_rate": 5.160715003432608e-05,
+      "loss": 1.8672,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6952965235173824,
+      "grad_norm": 0.13659731635234268,
+      "learning_rate": 5.1417948491038416e-05,
+      "loss": 1.8078,
+      "step": 5610
+    },
+    {
+      "epoch": 0.69591621738861,
+      "grad_norm": 0.149573593747872,
+      "learning_rate": 5.122897431220104e-05,
+      "loss": 1.8465,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6965359112598376,
+      "grad_norm": 0.14078615205837303,
+      "learning_rate": 5.104022838221487e-05,
+      "loss": 1.7676,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6971556051310652,
+      "grad_norm": 0.1380562039018842,
+      "learning_rate": 5.085171158441261e-05,
+      "loss": 1.8547,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6977752990022928,
+      "grad_norm": 0.15447513080222938,
+      "learning_rate": 5.0663424801054595e-05,
+      "loss": 1.8164,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6983949928735205,
+      "grad_norm": 0.13906912145428393,
+      "learning_rate": 5.047536891332473e-05,
+      "loss": 1.8,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6990146867447481,
+      "grad_norm": 0.13378728455920846,
+      "learning_rate": 5.0287544801326293e-05,
+      "loss": 1.8328,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6996343806159757,
+      "grad_norm": 0.1380817489955438,
+      "learning_rate": 5.0099953344077885e-05,
+      "loss": 1.843,
+      "step": 5645
+    },
+    {
+      "epoch": 0.7002540744872033,
+      "grad_norm": 0.13842527792443207,
+      "learning_rate": 4.991259541950924e-05,
+      "loss": 1.8582,
+      "step": 5650
+    },
+    {
+      "epoch": 0.700873768358431,
+      "grad_norm": 0.13436379086426098,
+      "learning_rate": 4.972547190445723e-05,
+      "loss": 1.8711,
+      "step": 5655
+    },
+    {
+      "epoch": 0.7014934622296586,
+      "grad_norm": 0.145219301193658,
+      "learning_rate": 4.953858367466155e-05,
+      "loss": 1.8957,
+      "step": 5660
+    },
+    {
+      "epoch": 0.7021131561008862,
+      "grad_norm": 0.14332316898614764,
+      "learning_rate": 4.9351931604760907e-05,
+      "loss": 1.8605,
+      "step": 5665
+    },
+    {
+      "epoch": 0.7027328499721138,
+      "grad_norm": 0.1313147508930661,
+      "learning_rate": 4.9165516568288674e-05,
+      "loss": 1.8637,
+      "step": 5670
+    },
+    {
+      "epoch": 0.7033525438433413,
+      "grad_norm": 0.13599383374559293,
+      "learning_rate": 4.897933943766897e-05,
+      "loss": 1.8156,
+      "step": 5675
+    },
+    {
+      "epoch": 0.703972237714569,
+      "grad_norm": 0.1310129091650981,
+      "learning_rate": 4.879340108421248e-05,
+      "loss": 1.8793,
+      "step": 5680
+    },
+    {
+      "epoch": 0.7045919315857966,
+      "grad_norm": 0.1358272525240438,
+      "learning_rate": 4.8607702378112415e-05,
+      "loss": 1.8398,
+      "step": 5685
+    },
+    {
+      "epoch": 0.7052116254570242,
+      "grad_norm": 0.13857163612920126,
+      "learning_rate": 4.842224418844045e-05,
+      "loss": 1.8082,
+      "step": 5690
+    },
+    {
+      "epoch": 0.7058313193282518,
+      "grad_norm": 0.13217816459332826,
+      "learning_rate": 4.823702738314262e-05,
+      "loss": 1.8242,
+      "step": 5695
+    },
+    {
+      "epoch": 0.7064510131994794,
+      "grad_norm": 0.13929364642164002,
+      "learning_rate": 4.8052052829035275e-05,
+      "loss": 1.882,
+      "step": 5700
+    },
+    {
+      "epoch": 0.7070707070707071,
+      "grad_norm": 0.13409665553182737,
+      "learning_rate": 4.7867321391801065e-05,
+      "loss": 1.8777,
+      "step": 5705
+    },
+    {
+      "epoch": 0.7076904009419347,
+      "grad_norm": 0.13751526203517048,
+      "learning_rate": 4.768283393598484e-05,
+      "loss": 1.8227,
+      "step": 5710
+    },
+    {
+      "epoch": 0.7083100948131623,
+      "grad_norm": 0.14070595152235213,
+      "learning_rate": 4.749859132498953e-05,
+      "loss": 1.8531,
+      "step": 5715
+    },
+    {
+      "epoch": 0.7089297886843899,
+      "grad_norm": 0.14049704552673886,
+      "learning_rate": 4.73145944210723e-05,
+      "loss": 1.8289,
+      "step": 5720
+    },
+    {
+      "epoch": 0.7095494825556176,
+      "grad_norm": 0.13542766328955186,
+      "learning_rate": 4.713084408534035e-05,
+      "loss": 1.7961,
+      "step": 5725
+    },
+    {
+      "epoch": 0.7101691764268452,
+      "grad_norm": 0.13784595495438537,
+      "learning_rate": 4.6947341177746926e-05,
+      "loss": 1.7789,
+      "step": 5730
+    },
+    {
+      "epoch": 0.7107888702980728,
+      "grad_norm": 0.1346102286615244,
+      "learning_rate": 4.6764086557087406e-05,
+      "loss": 1.8078,
+      "step": 5735
+    },
+    {
+      "epoch": 0.7114085641693003,
+      "grad_norm": 0.1399061673906458,
+      "learning_rate": 4.65810810809951e-05,
+      "loss": 1.8445,
+      "step": 5740
+    },
+    {
+      "epoch": 0.7120282580405279,
+      "grad_norm": 0.1392295817810214,
+      "learning_rate": 4.6398325605937265e-05,
+      "loss": 1.8391,
+      "step": 5745
+    },
+    {
+      "epoch": 0.7126479519117556,
+      "grad_norm": 0.13422887919186408,
+      "learning_rate": 4.621582098721124e-05,
+      "loss": 1.8699,
+      "step": 5750
+    },
+    {
+      "epoch": 0.7132676457829832,
+      "grad_norm": 0.13029477241952278,
+      "learning_rate": 4.6033568078940345e-05,
+      "loss": 1.8422,
+      "step": 5755
+    },
+    {
+      "epoch": 0.7138873396542108,
+      "grad_norm": 0.13926763308772894,
+      "learning_rate": 4.585156773406986e-05,
+      "loss": 1.8414,
+      "step": 5760
+    },
+    {
+      "epoch": 0.7145070335254384,
+      "grad_norm": 0.13603681047097105,
+      "learning_rate": 4.5669820804363116e-05,
+      "loss": 1.8285,
+      "step": 5765
+    },
+    {
+      "epoch": 0.715126727396666,
+      "grad_norm": 0.13825962890714413,
+      "learning_rate": 4.5488328140397364e-05,
+      "loss": 1.8645,
+      "step": 5770
+    },
+    {
+      "epoch": 0.7157464212678937,
+      "grad_norm": 0.13476991285028014,
+      "learning_rate": 4.530709059155994e-05,
+      "loss": 1.8305,
+      "step": 5775
+    },
+    {
+      "epoch": 0.7163661151391213,
+      "grad_norm": 0.13893798411440325,
+      "learning_rate": 4.512610900604434e-05,
+      "loss": 1.8445,
+      "step": 5780
+    },
+    {
+      "epoch": 0.7169858090103489,
+      "grad_norm": 0.13255106459645197,
+      "learning_rate": 4.4945384230846e-05,
+      "loss": 1.8152,
+      "step": 5785
+    },
+    {
+      "epoch": 0.7176055028815765,
+      "grad_norm": 0.13775635236702505,
+      "learning_rate": 4.476491711175854e-05,
+      "loss": 1.8016,
+      "step": 5790
+    },
+    {
+      "epoch": 0.7182251967528042,
+      "grad_norm": 0.134500964920823,
+      "learning_rate": 4.45847084933698e-05,
+      "loss": 1.8723,
+      "step": 5795
+    },
+    {
+      "epoch": 0.7188448906240317,
+      "grad_norm": 0.13597330490660237,
+      "learning_rate": 4.440475921905768e-05,
+      "loss": 1.7895,
+      "step": 5800
+    },
+    {
+      "epoch": 0.7194645844952593,
+      "grad_norm": 0.1315917432903297,
+      "learning_rate": 4.422507013098651e-05,
+      "loss": 1.8133,
+      "step": 5805
+    },
+    {
+      "epoch": 0.7200842783664869,
+      "grad_norm": 0.13875462075386993,
+      "learning_rate": 4.404564207010288e-05,
+      "loss": 1.857,
+      "step": 5810
+    },
+    {
+      "epoch": 0.7207039722377145,
+      "grad_norm": 0.13431544947110896,
+      "learning_rate": 4.3866475876131764e-05,
+      "loss": 1.7969,
+      "step": 5815
+    },
+    {
+      "epoch": 0.7213236661089422,
+      "grad_norm": 0.13155675895860156,
+      "learning_rate": 4.3687572387572605e-05,
+      "loss": 1.8605,
+      "step": 5820
+    },
+    {
+      "epoch": 0.7219433599801698,
+      "grad_norm": 0.13995415211194065,
+      "learning_rate": 4.350893244169541e-05,
+      "loss": 1.8555,
+      "step": 5825
+    },
+    {
+      "epoch": 0.7225630538513974,
+      "grad_norm": 0.13449577476692642,
+      "learning_rate": 4.333055687453673e-05,
+      "loss": 1.8195,
+      "step": 5830
+    },
+    {
+      "epoch": 0.723182747722625,
+      "grad_norm": 0.14401065172175462,
+      "learning_rate": 4.315244652089592e-05,
+      "loss": 1.7914,
+      "step": 5835
+    },
+    {
+      "epoch": 0.7238024415938527,
+      "grad_norm": 0.1389024704353045,
+      "learning_rate": 4.297460221433104e-05,
+      "loss": 1.8316,
+      "step": 5840
+    },
+    {
+      "epoch": 0.7244221354650803,
+      "grad_norm": 0.13282353788709725,
+      "learning_rate": 4.2797024787155114e-05,
+      "loss": 1.8426,
+      "step": 5845
+    },
+    {
+      "epoch": 0.7250418293363079,
+      "grad_norm": 0.13791586639383402,
+      "learning_rate": 4.2619715070432174e-05,
+      "loss": 1.8484,
+      "step": 5850
+    },
+    {
+      "epoch": 0.7256615232075355,
+      "grad_norm": 0.13459363626623866,
+      "learning_rate": 4.244267389397326e-05,
+      "loss": 1.825,
+      "step": 5855
+    },
+    {
+      "epoch": 0.7262812170787631,
+      "grad_norm": 0.13114527523173383,
+      "learning_rate": 4.226590208633275e-05,
+      "loss": 1.827,
+      "step": 5860
+    },
+    {
+      "epoch": 0.7269009109499907,
+      "grad_norm": 0.13950898565386535,
+      "learning_rate": 4.208940047480434e-05,
+      "loss": 1.8,
+      "step": 5865
+    },
+    {
+      "epoch": 0.7275206048212183,
+      "grad_norm": 0.13754357299749095,
+      "learning_rate": 4.191316988541721e-05,
+      "loss": 1.8449,
+      "step": 5870
+    },
+    {
+      "epoch": 0.7281402986924459,
+      "grad_norm": 0.1412420242297336,
+      "learning_rate": 4.173721114293214e-05,
+      "loss": 1.8195,
+      "step": 5875
+    },
+    {
+      "epoch": 0.7287599925636735,
+      "grad_norm": 0.13460299907847717,
+      "learning_rate": 4.156152507083767e-05,
+      "loss": 1.8094,
+      "step": 5880
+    },
+    {
+      "epoch": 0.7293796864349011,
+      "grad_norm": 0.1335314650244059,
+      "learning_rate": 4.1386112491346255e-05,
+      "loss": 1.8129,
+      "step": 5885
+    },
+    {
+      "epoch": 0.7299993803061288,
+      "grad_norm": 0.13934436501401282,
+      "learning_rate": 4.121097422539036e-05,
+      "loss": 1.8836,
+      "step": 5890
+    },
+    {
+      "epoch": 0.7306190741773564,
+      "grad_norm": 0.13065863803776073,
+      "learning_rate": 4.1036111092618725e-05,
+      "loss": 1.8102,
+      "step": 5895
+    },
+    {
+      "epoch": 0.731238768048584,
+      "grad_norm": 0.139176383450031,
+      "learning_rate": 4.0861523911392406e-05,
+      "loss": 1.784,
+      "step": 5900
+    },
+    {
+      "epoch": 0.7318584619198116,
+      "grad_norm": 0.1382728042720779,
+      "learning_rate": 4.068721349878107e-05,
+      "loss": 1.8781,
+      "step": 5905
+    },
+    {
+      "epoch": 0.7324781557910393,
+      "grad_norm": 0.13071903823760816,
+      "learning_rate": 4.051318067055898e-05,
+      "loss": 1.8363,
+      "step": 5910
+    },
+    {
+      "epoch": 0.7330978496622669,
+      "grad_norm": 0.13148997495963968,
+      "learning_rate": 4.033942624120143e-05,
+      "loss": 1.8223,
+      "step": 5915
+    },
+    {
+      "epoch": 0.7337175435334945,
+      "grad_norm": 0.13306227335689444,
+      "learning_rate": 4.0165951023880746e-05,
+      "loss": 1.8449,
+      "step": 5920
+    },
+    {
+      "epoch": 0.7343372374047221,
+      "grad_norm": 0.13278952164174923,
+      "learning_rate": 3.999275583046256e-05,
+      "loss": 1.7824,
+      "step": 5925
+    },
+    {
+      "epoch": 0.7349569312759496,
+      "grad_norm": 0.14316705856810816,
+      "learning_rate": 3.981984147150196e-05,
+      "loss": 1.8535,
+      "step": 5930
+    },
+    {
+      "epoch": 0.7355766251471773,
+      "grad_norm": 0.13612015046365858,
+      "learning_rate": 3.964720875623976e-05,
+      "loss": 1.8055,
+      "step": 5935
+    },
+    {
+      "epoch": 0.7361963190184049,
+      "grad_norm": 0.13203944006386598,
+      "learning_rate": 3.9474858492598653e-05,
+      "loss": 1.8031,
+      "step": 5940
+    },
+    {
+      "epoch": 0.7368160128896325,
+      "grad_norm": 0.1327410609332686,
+      "learning_rate": 3.930279148717948e-05,
+      "loss": 1.7957,
+      "step": 5945
+    },
+    {
+      "epoch": 0.7374357067608601,
+      "grad_norm": 0.1412184409098123,
+      "learning_rate": 3.913100854525742e-05,
+      "loss": 1.8562,
+      "step": 5950
+    },
+    {
+      "epoch": 0.7380554006320877,
+      "grad_norm": 0.14211921706904118,
+      "learning_rate": 3.895951047077821e-05,
+      "loss": 1.8051,
+      "step": 5955
+    },
+    {
+      "epoch": 0.7386750945033154,
+      "grad_norm": 0.14363924491542174,
+      "learning_rate": 3.8788298066354464e-05,
+      "loss": 1.8109,
+      "step": 5960
+    },
+    {
+      "epoch": 0.739294788374543,
+      "grad_norm": 0.13367304925942577,
+      "learning_rate": 3.8617372133261766e-05,
+      "loss": 1.8359,
+      "step": 5965
+    },
+    {
+      "epoch": 0.7399144822457706,
+      "grad_norm": 0.1283371909012973,
+      "learning_rate": 3.844673347143507e-05,
+      "loss": 1.8594,
+      "step": 5970
+    },
+    {
+      "epoch": 0.7405341761169982,
+      "grad_norm": 0.140355325772872,
+      "learning_rate": 3.827638287946489e-05,
+      "loss": 1.8551,
+      "step": 5975
+    },
+    {
+      "epoch": 0.7411538699882259,
+      "grad_norm": 0.13603925926463906,
+      "learning_rate": 3.8106321154593605e-05,
+      "loss": 1.8039,
+      "step": 5980
+    },
+    {
+      "epoch": 0.7417735638594535,
+      "grad_norm": 0.13333303827812676,
+      "learning_rate": 3.793654909271169e-05,
+      "loss": 1.7832,
+      "step": 5985
+    },
+    {
+      "epoch": 0.742393257730681,
+      "grad_norm": 0.14092753476620795,
+      "learning_rate": 3.776706748835388e-05,
+      "loss": 1.8727,
+      "step": 5990
+    },
+    {
+      "epoch": 0.7430129516019086,
+      "grad_norm": 0.13526664213664327,
+      "learning_rate": 3.759787713469569e-05,
+      "loss": 1.8406,
+      "step": 5995
+    },
+    {
+      "epoch": 0.7436326454731362,
+      "grad_norm": 0.13890336096217432,
+      "learning_rate": 3.7428978823549545e-05,
+      "loss": 1.8184,
+      "step": 6000
+    },
+    {
+      "epoch": 0.7442523393443639,
+      "grad_norm": 0.136193007857896,
+      "learning_rate": 3.726037334536109e-05,
+      "loss": 1.8562,
+      "step": 6005
+    },
+    {
+      "epoch": 0.7448720332155915,
+      "grad_norm": 0.14442667788304958,
+      "learning_rate": 3.709206148920553e-05,
+      "loss": 1.9062,
+      "step": 6010
+    },
+    {
+      "epoch": 0.7454917270868191,
+      "grad_norm": 0.13679257439666426,
+      "learning_rate": 3.692404404278395e-05,
+      "loss": 1.8246,
+      "step": 6015
+    },
+    {
+      "epoch": 0.7461114209580467,
+      "grad_norm": 0.13120658103553706,
+      "learning_rate": 3.675632179241946e-05,
+      "loss": 1.8305,
+      "step": 6020
+    },
+    {
+      "epoch": 0.7467311148292743,
+      "grad_norm": 0.14446220470432317,
+      "learning_rate": 3.658889552305376e-05,
+      "loss": 1.7777,
+      "step": 6025
+    },
+    {
+      "epoch": 0.747350808700502,
+      "grad_norm": 0.13770389116571016,
+      "learning_rate": 3.642176601824341e-05,
+      "loss": 1.8184,
+      "step": 6030
+    },
+    {
+      "epoch": 0.7479705025717296,
+      "grad_norm": 0.13267683993530827,
+      "learning_rate": 3.6254934060156e-05,
+      "loss": 1.8637,
+      "step": 6035
+    },
+    {
+      "epoch": 0.7485901964429572,
+      "grad_norm": 0.13804463607426604,
+      "learning_rate": 3.608840042956666e-05,
+      "loss": 1.8715,
+      "step": 6040
+    },
+    {
+      "epoch": 0.7492098903141848,
+      "grad_norm": 0.13803991697296203,
+      "learning_rate": 3.592216590585427e-05,
+      "loss": 1.8703,
+      "step": 6045
+    },
+    {
+      "epoch": 0.7498295841854125,
+      "grad_norm": 0.13771290613002898,
+      "learning_rate": 3.5756231266997965e-05,
+      "loss": 1.8266,
+      "step": 6050
+    },
+    {
+      "epoch": 0.75044927805664,
+      "grad_norm": 0.1377426301197608,
+      "learning_rate": 3.559059728957338e-05,
+      "loss": 1.8387,
+      "step": 6055
+    },
+    {
+      "epoch": 0.7510689719278676,
+      "grad_norm": 0.13142143493494793,
+      "learning_rate": 3.5425264748749074e-05,
+      "loss": 1.8754,
+      "step": 6060
+    },
+    {
+      "epoch": 0.7516886657990952,
+      "grad_norm": 0.13870584512058637,
+      "learning_rate": 3.5260234418282865e-05,
+      "loss": 1.8102,
+      "step": 6065
+    },
+    {
+      "epoch": 0.7523083596703228,
+      "grad_norm": 0.13263561172267807,
+      "learning_rate": 3.509550707051823e-05,
+      "loss": 1.8395,
+      "step": 6070
+    },
+    {
+      "epoch": 0.7529280535415505,
+      "grad_norm": 0.13347345228759588,
+      "learning_rate": 3.493108347638067e-05,
+      "loss": 1.8488,
+      "step": 6075
+    },
+    {
+      "epoch": 0.7535477474127781,
+      "grad_norm": 0.1348178977080782,
+      "learning_rate": 3.476696440537413e-05,
+      "loss": 1.7828,
+      "step": 6080
+    },
+    {
+      "epoch": 0.7541674412840057,
+      "grad_norm": 0.14134614205432544,
+      "learning_rate": 3.460315062557737e-05,
+      "loss": 1.793,
+      "step": 6085
+    },
+    {
+      "epoch": 0.7547871351552333,
+      "grad_norm": 0.1302945011937037,
+      "learning_rate": 3.443964290364041e-05,
+      "loss": 1.802,
+      "step": 6090
+    },
+    {
+      "epoch": 0.755406829026461,
+      "grad_norm": 0.13968371731188192,
+      "learning_rate": 3.4276442004780916e-05,
+      "loss": 1.9215,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7560265228976886,
+      "grad_norm": 0.1455653462790958,
+      "learning_rate": 3.411354869278056e-05,
+      "loss": 1.8215,
+      "step": 6100
+    },
+    {
+      "epoch": 0.7566462167689162,
+      "grad_norm": 0.14760360535248854,
+      "learning_rate": 3.3950963729981565e-05,
+      "loss": 1.8395,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7572659106401438,
+      "grad_norm": 0.13441648798291483,
+      "learning_rate": 3.378868787728308e-05,
+      "loss": 1.891,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7578856045113714,
+      "grad_norm": 0.13945591613635544,
+      "learning_rate": 3.362672189413756e-05,
+      "loss": 1.832,
+      "step": 6115
+    },
+    {
+      "epoch": 0.758505298382599,
+      "grad_norm": 0.1356119340760843,
+      "learning_rate": 3.346506653854734e-05,
+      "loss": 1.8215,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7591249922538266,
+      "grad_norm": 0.14180993594873473,
+      "learning_rate": 3.3303722567060956e-05,
+      "loss": 1.7512,
+      "step": 6125
+    },
+    {
+      "epoch": 0.7597446861250542,
+      "grad_norm": 0.12910186660875014,
+      "learning_rate": 3.31426907347697e-05,
+      "loss": 1.8566,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7603643799962818,
+      "grad_norm": 0.14017170111652089,
+      "learning_rate": 3.2981971795304026e-05,
+      "loss": 1.8836,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7609840738675094,
+      "grad_norm": 0.13260660637237812,
+      "learning_rate": 3.282156650083006e-05,
+      "loss": 1.8633,
+      "step": 6140
+    },
+    {
+      "epoch": 0.7616037677387371,
+      "grad_norm": 0.13413829542520983,
+      "learning_rate": 3.266147560204608e-05,
+      "loss": 1.8672,
+      "step": 6145
+    },
+    {
+      "epoch": 0.7622234616099647,
+      "grad_norm": 0.14500649061949047,
+      "learning_rate": 3.250169984817897e-05,
+      "loss": 1.8227,
+      "step": 6150
+    },
+    {
+      "epoch": 0.7628431554811923,
+      "grad_norm": 0.13143903720477662,
+      "learning_rate": 3.23422399869807e-05,
+      "loss": 1.8137,
+      "step": 6155
+    },
+    {
+      "epoch": 0.7634628493524199,
+      "grad_norm": 0.13853349840431195,
+      "learning_rate": 3.2183096764724915e-05,
+      "loss": 1.8305,
+      "step": 6160
+    },
+    {
+      "epoch": 0.7640825432236475,
+      "grad_norm": 0.13860343379835846,
+      "learning_rate": 3.2024270926203384e-05,
+      "loss": 1.7961,
+      "step": 6165
+    },
+    {
+      "epoch": 0.7647022370948752,
+      "grad_norm": 0.1384361618622998,
+      "learning_rate": 3.1865763214722474e-05,
+      "loss": 1.8492,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7653219309661028,
+      "grad_norm": 0.13199549343512088,
+      "learning_rate": 3.1707574372099754e-05,
+      "loss": 1.8434,
+      "step": 6175
+    },
+    {
+      "epoch": 0.7659416248373303,
+      "grad_norm": 0.13451064848409564,
+      "learning_rate": 3.154970513866047e-05,
+      "loss": 1.8215,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7665613187085579,
+      "grad_norm": 0.1378993997426509,
+      "learning_rate": 3.1392156253234086e-05,
+      "loss": 1.8219,
+      "step": 6185
+    },
+    {
+      "epoch": 0.7671810125797855,
+      "grad_norm": 0.13271589858843205,
+      "learning_rate": 3.123492845315086e-05,
+      "loss": 1.8707,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7678007064510132,
+      "grad_norm": 0.13419734586499776,
+      "learning_rate": 3.1078022474238334e-05,
+      "loss": 1.8223,
+      "step": 6195
+    },
+    {
+      "epoch": 0.7684204003222408,
+      "grad_norm": 0.13926944052109394,
+      "learning_rate": 3.092143905081794e-05,
+      "loss": 1.8313,
+      "step": 6200
+    },
+    {
+      "epoch": 0.7690400941934684,
+      "grad_norm": 0.13678235355553914,
+      "learning_rate": 3.07651789157016e-05,
+      "loss": 1.8035,
+      "step": 6205
+    },
+    {
+      "epoch": 0.769659788064696,
+      "grad_norm": 0.13151996809359814,
+      "learning_rate": 3.060924280018811e-05,
+      "loss": 1.7555,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7702794819359237,
+      "grad_norm": 0.14276236865321887,
+      "learning_rate": 3.0453631434059958e-05,
+      "loss": 1.7926,
+      "step": 6215
+    },
+    {
+      "epoch": 0.7708991758071513,
+      "grad_norm": 0.13592822741431418,
+      "learning_rate": 3.0298345545579787e-05,
+      "loss": 1.8324,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7715188696783789,
+      "grad_norm": 0.15065267997122453,
+      "learning_rate": 3.0143385861486974e-05,
+      "loss": 1.8789,
+      "step": 6225
+    },
+    {
+      "epoch": 0.7721385635496065,
+      "grad_norm": 0.13878430796743169,
+      "learning_rate": 2.9988753106994306e-05,
+      "loss": 1.848,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7727582574208341,
+      "grad_norm": 0.13441719402081825,
+      "learning_rate": 2.983444800578452e-05,
+      "loss": 1.8508,
+      "step": 6235
+    },
+    {
+      "epoch": 0.7733779512920618,
+      "grad_norm": 0.1409089714119921,
+      "learning_rate": 2.9680471280006848e-05,
+      "loss": 1.8441,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7739976451632893,
+      "grad_norm": 0.13250438881223234,
+      "learning_rate": 2.9526823650273837e-05,
+      "loss": 1.8266,
+      "step": 6245
+    },
+    {
+      "epoch": 0.7746173390345169,
+      "grad_norm": 0.14901974178220379,
+      "learning_rate": 2.93735058356578e-05,
+      "loss": 1.7891,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7752370329057445,
+      "grad_norm": 0.13981158606565733,
+      "learning_rate": 2.9220518553687526e-05,
+      "loss": 1.8555,
+      "step": 6255
+    },
+    {
+      "epoch": 0.7758567267769721,
+      "grad_norm": 0.13508067673905183,
+      "learning_rate": 2.9067862520344956e-05,
+      "loss": 1.8375,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7764764206481998,
+      "grad_norm": 0.1329903186719515,
+      "learning_rate": 2.891553845006165e-05,
+      "loss": 1.825,
+      "step": 6265
+    },
+    {
+      "epoch": 0.7770961145194274,
+      "grad_norm": 0.1379727364911777,
+      "learning_rate": 2.87635470557157e-05,
+      "loss": 1.8465,
+      "step": 6270
+    },
+    {
+      "epoch": 0.777715808390655,
+      "grad_norm": 0.1410408068306655,
+      "learning_rate": 2.861188904862827e-05,
+      "loss": 1.8449,
+      "step": 6275
+    },
+    {
+      "epoch": 0.7783355022618826,
+      "grad_norm": 0.1314735833953845,
+      "learning_rate": 2.8460565138560212e-05,
+      "loss": 1.8219,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7789551961331103,
+      "grad_norm": 0.13288247697952463,
+      "learning_rate": 2.830957603370883e-05,
+      "loss": 1.8398,
+      "step": 6285
+    },
+    {
+      "epoch": 0.7795748900043379,
+      "grad_norm": 0.13510806378576828,
+      "learning_rate": 2.815892244070455e-05,
+      "loss": 1.8801,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7801945838755655,
+      "grad_norm": 0.13593787427552473,
+      "learning_rate": 2.8008605064607528e-05,
+      "loss": 1.8637,
+      "step": 6295
+    },
+    {
+      "epoch": 0.7808142777467931,
+      "grad_norm": 0.13570594578366504,
+      "learning_rate": 2.7858624608904515e-05,
+      "loss": 1.9043,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7814339716180208,
+      "grad_norm": 0.13671078512318602,
+      "learning_rate": 2.7708981775505416e-05,
+      "loss": 1.8309,
+      "step": 6305
+    },
+    {
+      "epoch": 0.7820536654892483,
+      "grad_norm": 0.1411639120932341,
+      "learning_rate": 2.755967726474007e-05,
+      "loss": 1.8258,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7826733593604759,
+      "grad_norm": 0.13774898564586507,
+      "learning_rate": 2.741071177535499e-05,
+      "loss": 1.8227,
+      "step": 6315
+    },
+    {
+      "epoch": 0.7832930532317035,
+      "grad_norm": 0.13076963037262845,
+      "learning_rate": 2.7262086004510023e-05,
+      "loss": 1.8277,
+      "step": 6320
+    },
+    {
+      "epoch": 0.7839127471029311,
+      "grad_norm": 0.1343651996214831,
+      "learning_rate": 2.7113800647775156e-05,
+      "loss": 1.7766,
+      "step": 6325
+    },
+    {
+      "epoch": 0.7845324409741588,
+      "grad_norm": 0.13468642096762548,
+      "learning_rate": 2.6965856399127232e-05,
+      "loss": 1.8441,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7851521348453864,
+      "grad_norm": 0.14720986177976159,
+      "learning_rate": 2.6818253950946704e-05,
+      "loss": 1.8488,
+      "step": 6335
+    },
+    {
+      "epoch": 0.785771828716614,
+      "grad_norm": 0.1560079026652874,
+      "learning_rate": 2.6670993994014394e-05,
+      "loss": 1.875,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7863915225878416,
+      "grad_norm": 0.13378317082115668,
+      "learning_rate": 2.6524077217508292e-05,
+      "loss": 1.8277,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7870112164590692,
+      "grad_norm": 0.14218358804125789,
+      "learning_rate": 2.63775043090002e-05,
+      "loss": 1.8301,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7876309103302969,
+      "grad_norm": 0.13527135821037153,
+      "learning_rate": 2.623127595445274e-05,
+      "loss": 1.7805,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7882506042015245,
+      "grad_norm": 0.14734614025952317,
+      "learning_rate": 2.6085392838215938e-05,
+      "loss": 1.875,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7888702980727521,
+      "grad_norm": 0.14019751079756237,
+      "learning_rate": 2.5939855643024136e-05,
+      "loss": 1.8344,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7894899919439796,
+      "grad_norm": 0.1363368429297401,
+      "learning_rate": 2.5794665049992762e-05,
+      "loss": 1.7402,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7901096858152072,
+      "grad_norm": 0.13954747652635188,
+      "learning_rate": 2.564982173861512e-05,
+      "loss": 1.8367,
+      "step": 6375
+    },
+    {
+      "epoch": 0.7907293796864349,
+      "grad_norm": 0.13494365879866646,
+      "learning_rate": 2.5505326386759254e-05,
+      "loss": 1.8523,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7913490735576625,
+      "grad_norm": 0.13314683645391906,
+      "learning_rate": 2.536117967066476e-05,
+      "loss": 1.7867,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7919687674288901,
+      "grad_norm": 0.13327408799707985,
+      "learning_rate": 2.521738226493957e-05,
+      "loss": 1.8102,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7925884613001177,
+      "grad_norm": 0.13992946116216445,
+      "learning_rate": 2.50739348425569e-05,
+      "loss": 1.7902,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7932081551713454,
+      "grad_norm": 0.14100811061457322,
+      "learning_rate": 2.4930838074852026e-05,
+      "loss": 1.8656,
+      "step": 6400
+    },
+    {
+      "epoch": 0.793827849042573,
+      "grad_norm": 0.1474245956678477,
+      "learning_rate": 2.47880926315191e-05,
+      "loss": 1.8477,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7944475429138006,
+      "grad_norm": 0.13533489898855472,
+      "learning_rate": 2.4645699180608127e-05,
+      "loss": 1.8414,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7950672367850282,
+      "grad_norm": 0.136517801601565,
+      "learning_rate": 2.45036583885218e-05,
+      "loss": 1.8234,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7956869306562558,
+      "grad_norm": 0.13002283300770023,
+      "learning_rate": 2.4361970920012313e-05,
+      "loss": 1.9266,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7963066245274835,
+      "grad_norm": 0.1379400020050846,
+      "learning_rate": 2.4220637438178317e-05,
+      "loss": 1.8352,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7969263183987111,
+      "grad_norm": 0.1336448212657819,
+      "learning_rate": 2.4079658604461896e-05,
+      "loss": 1.8168,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7975460122699386,
+      "grad_norm": 0.13573713481362645,
+      "learning_rate": 2.393903507864521e-05,
+      "loss": 1.8309,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7981657061411662,
+      "grad_norm": 0.1386764771891764,
+      "learning_rate": 2.3798767518847687e-05,
+      "loss": 1.859,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7987854000123938,
+      "grad_norm": 0.13800165635093364,
+      "learning_rate": 2.3658856581522804e-05,
+      "loss": 1.8344,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7994050938836215,
+      "grad_norm": 0.14205563126041354,
+      "learning_rate": 2.3519302921455033e-05,
+      "loss": 1.8223,
+      "step": 6450
+    },
+    {
+      "epoch": 0.8000247877548491,
+      "grad_norm": 0.1367572578358323,
+      "learning_rate": 2.338010719175684e-05,
+      "loss": 1.7699,
+      "step": 6455
+    },
+    {
+      "epoch": 0.8006444816260767,
+      "grad_norm": 0.13208109576129815,
+      "learning_rate": 2.324127004386546e-05,
+      "loss": 1.8512,
+      "step": 6460
+    },
+    {
+      "epoch": 0.8012641754973043,
+      "grad_norm": 0.13593091577531968,
+      "learning_rate": 2.310279212754006e-05,
+      "loss": 1.8379,
+      "step": 6465
+    },
+    {
+      "epoch": 0.801883869368532,
+      "grad_norm": 0.1342950750798231,
+      "learning_rate": 2.296467409085853e-05,
+      "loss": 1.8473,
+      "step": 6470
+    },
+    {
+      "epoch": 0.8025035632397596,
+      "grad_norm": 0.14153433233252935,
+      "learning_rate": 2.2826916580214632e-05,
+      "loss": 1.775,
+      "step": 6475
+    },
+    {
+      "epoch": 0.8031232571109872,
+      "grad_norm": 0.1384192064540353,
+      "learning_rate": 2.2689520240314755e-05,
+      "loss": 1.8586,
+      "step": 6480
+    },
+    {
+      "epoch": 0.8037429509822148,
+      "grad_norm": 0.13797923662673559,
+      "learning_rate": 2.2552485714175064e-05,
+      "loss": 1.8453,
+      "step": 6485
+    },
+    {
+      "epoch": 0.8043626448534424,
+      "grad_norm": 0.1318062253579002,
+      "learning_rate": 2.2415813643118356e-05,
+      "loss": 1.8266,
+      "step": 6490
+    },
+    {
+      "epoch": 0.8049823387246701,
+      "grad_norm": 0.1380940923690682,
+      "learning_rate": 2.227950466677121e-05,
+      "loss": 1.8016,
+      "step": 6495
+    },
+    {
+      "epoch": 0.8056020325958976,
+      "grad_norm": 0.13292787617402785,
+      "learning_rate": 2.21435594230609e-05,
+      "loss": 1.8113,
+      "step": 6500
+    },
+    {
+      "epoch": 0.8062217264671252,
+      "grad_norm": 0.14153364728285703,
+      "learning_rate": 2.2007978548212425e-05,
+      "loss": 1.7828,
+      "step": 6505
+    },
+    {
+      "epoch": 0.8068414203383528,
+      "grad_norm": 0.13985017879700073,
+      "learning_rate": 2.1872762676745563e-05,
+      "loss": 1.8672,
+      "step": 6510
+    },
+    {
+      "epoch": 0.8074611142095804,
+      "grad_norm": 0.1316374604121444,
+      "learning_rate": 2.1737912441471787e-05,
+      "loss": 1.7961,
+      "step": 6515
+    },
+    {
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.13947504872687227,
+      "learning_rate": 2.160342847349144e-05,
+      "loss": 1.8305,
+      "step": 6520
+    },
+    {
+      "epoch": 0.8087005019520357,
+      "grad_norm": 0.1415289846578962,
+      "learning_rate": 2.1469311402190794e-05,
+      "loss": 1.8211,
+      "step": 6525
+    },
+    {
+      "epoch": 0.8093201958232633,
+      "grad_norm": 0.13209481549795832,
+      "learning_rate": 2.133556185523895e-05,
+      "loss": 1.8215,
+      "step": 6530
+    },
+    {
+      "epoch": 0.8099398896944909,
+      "grad_norm": 0.13213273877464646,
+      "learning_rate": 2.120218045858503e-05,
+      "loss": 1.8617,
+      "step": 6535
+    },
+    {
+      "epoch": 0.8105595835657186,
+      "grad_norm": 0.14048998695177167,
+      "learning_rate": 2.1069167836455228e-05,
+      "loss": 1.8254,
+      "step": 6540
+    },
+    {
+      "epoch": 0.8111792774369462,
+      "grad_norm": 0.13984057341213088,
+      "learning_rate": 2.0936524611349795e-05,
+      "loss": 1.7922,
+      "step": 6545
+    },
+    {
+      "epoch": 0.8117989713081738,
+      "grad_norm": 0.14234566575879473,
+      "learning_rate": 2.080425140404029e-05,
+      "loss": 1.85,
+      "step": 6550
+    },
+    {
+      "epoch": 0.8124186651794014,
+      "grad_norm": 0.13619722325941194,
+      "learning_rate": 2.0672348833566512e-05,
+      "loss": 1.8723,
+      "step": 6555
+    },
+    {
+      "epoch": 0.813038359050629,
+      "grad_norm": 0.13354456391525196,
+      "learning_rate": 2.0540817517233735e-05,
+      "loss": 1.8309,
+      "step": 6560
+    },
+    {
+      "epoch": 0.8136580529218566,
+      "grad_norm": 0.1306309793562597,
+      "learning_rate": 2.0409658070609738e-05,
+      "loss": 1.8254,
+      "step": 6565
+    },
+    {
+      "epoch": 0.8142777467930842,
+      "grad_norm": 0.13739179800016613,
+      "learning_rate": 2.0278871107521936e-05,
+      "loss": 1.8023,
+      "step": 6570
+    },
+    {
+      "epoch": 0.8148974406643118,
+      "grad_norm": 0.13666140093469806,
+      "learning_rate": 2.014845724005453e-05,
+      "loss": 1.8449,
+      "step": 6575
+    },
+    {
+      "epoch": 0.8155171345355394,
+      "grad_norm": 0.13040603198376724,
+      "learning_rate": 2.0018417078545614e-05,
+      "loss": 1.8324,
+      "step": 6580
+    },
+    {
+      "epoch": 0.816136828406767,
+      "grad_norm": 0.14022816506048247,
+      "learning_rate": 1.988875123158437e-05,
+      "loss": 1.8051,
+      "step": 6585
+    },
+    {
+      "epoch": 0.8167565222779947,
+      "grad_norm": 0.13382954051159468,
+      "learning_rate": 1.975946030600814e-05,
+      "loss": 1.7973,
+      "step": 6590
+    },
+    {
+      "epoch": 0.8173762161492223,
+      "grad_norm": 0.14241355395415345,
+      "learning_rate": 1.9630544906899672e-05,
+      "loss": 1.791,
+      "step": 6595
+    },
+    {
+      "epoch": 0.8179959100204499,
+      "grad_norm": 0.13466759194537686,
+      "learning_rate": 1.9502005637584198e-05,
+      "loss": 1.8574,
+      "step": 6600
+    },
+    {
+      "epoch": 0.8186156038916775,
+      "grad_norm": 0.13508090700592257,
+      "learning_rate": 1.93738430996267e-05,
+      "loss": 1.8477,
+      "step": 6605
+    },
+    {
+      "epoch": 0.8192352977629052,
+      "grad_norm": 0.14482298397934196,
+      "learning_rate": 1.9246057892829038e-05,
+      "loss": 1.8273,
+      "step": 6610
+    },
+    {
+      "epoch": 0.8198549916341328,
+      "grad_norm": 0.1407416637413334,
+      "learning_rate": 1.9118650615227162e-05,
+      "loss": 1.8258,
+      "step": 6615
+    },
+    {
+      "epoch": 0.8204746855053604,
+      "grad_norm": 0.13884770681711742,
+      "learning_rate": 1.8991621863088315e-05,
+      "loss": 1.827,
+      "step": 6620
+    },
+    {
+      "epoch": 0.8210943793765879,
+      "grad_norm": 0.13406222131993933,
+      "learning_rate": 1.886497223090823e-05,
+      "loss": 1.8367,
+      "step": 6625
+    },
+    {
+      "epoch": 0.8217140732478155,
+      "grad_norm": 0.13345740447218,
+      "learning_rate": 1.8738702311408352e-05,
+      "loss": 1.85,
+      "step": 6630
+    },
+    {
+      "epoch": 0.8223337671190432,
+      "grad_norm": 0.1361039949242507,
+      "learning_rate": 1.8612812695533077e-05,
+      "loss": 1.9008,
+      "step": 6635
+    },
+    {
+      "epoch": 0.8229534609902708,
+      "grad_norm": 0.14532071771167876,
+      "learning_rate": 1.8487303972446966e-05,
+      "loss": 1.8203,
+      "step": 6640
+    },
+    {
+      "epoch": 0.8235731548614984,
+      "grad_norm": 0.13415668268856182,
+      "learning_rate": 1.836217672953201e-05,
+      "loss": 1.8738,
+      "step": 6645
+    },
+    {
+      "epoch": 0.824192848732726,
+      "grad_norm": 0.14031869927136761,
+      "learning_rate": 1.8237431552384887e-05,
+      "loss": 1.8176,
+      "step": 6650
+    },
+    {
+      "epoch": 0.8248125426039536,
+      "grad_norm": 0.13674539846811842,
+      "learning_rate": 1.811306902481412e-05,
+      "loss": 1.7992,
+      "step": 6655
+    },
+    {
+      "epoch": 0.8254322364751813,
+      "grad_norm": 0.14014061563694274,
+      "learning_rate": 1.798908972883754e-05,
+      "loss": 1.8609,
+      "step": 6660
+    },
+    {
+      "epoch": 0.8260519303464089,
+      "grad_norm": 0.14206135339382653,
+      "learning_rate": 1.786549424467936e-05,
+      "loss": 1.9484,
+      "step": 6665
+    },
+    {
+      "epoch": 0.8266716242176365,
+      "grad_norm": 0.14489028988827174,
+      "learning_rate": 1.7742283150767614e-05,
+      "loss": 1.9082,
+      "step": 6670
+    },
+    {
+      "epoch": 0.8272913180888641,
+      "grad_norm": 0.13006211826169928,
+      "learning_rate": 1.7619457023731355e-05,
+      "loss": 1.8477,
+      "step": 6675
+    },
+    {
+      "epoch": 0.8279110119600918,
+      "grad_norm": 0.13916462938836544,
+      "learning_rate": 1.7497016438397984e-05,
+      "loss": 1.8336,
+      "step": 6680
+    },
+    {
+      "epoch": 0.8285307058313194,
+      "grad_norm": 0.13832093116873143,
+      "learning_rate": 1.737496196779059e-05,
+      "loss": 1.798,
+      "step": 6685
+    },
+    {
+      "epoch": 0.8291503997025469,
+      "grad_norm": 0.14176420760152497,
+      "learning_rate": 1.7253294183125223e-05,
+      "loss": 1.8395,
+      "step": 6690
+    },
+    {
+      "epoch": 0.8297700935737745,
+      "grad_norm": 0.14071427615818557,
+      "learning_rate": 1.7132013653808222e-05,
+      "loss": 1.8344,
+      "step": 6695
+    },
+    {
+      "epoch": 0.8303897874450021,
+      "grad_norm": 0.13551316956515488,
+      "learning_rate": 1.70111209474336e-05,
+      "loss": 1.8719,
+      "step": 6700
+    },
+    {
+      "epoch": 0.8310094813162298,
+      "grad_norm": 0.1369378832967966,
+      "learning_rate": 1.6890616629780364e-05,
+      "loss": 1.8125,
+      "step": 6705
+    },
+    {
+      "epoch": 0.8316291751874574,
+      "grad_norm": 0.13670780086941425,
+      "learning_rate": 1.6770501264809778e-05,
+      "loss": 1.8508,
+      "step": 6710
+    },
+    {
+      "epoch": 0.832248869058685,
+      "grad_norm": 0.14168797345046086,
+      "learning_rate": 1.665077541466289e-05,
+      "loss": 1.8492,
+      "step": 6715
+    },
+    {
+      "epoch": 0.8328685629299126,
+      "grad_norm": 0.12899525112771104,
+      "learning_rate": 1.6531439639657776e-05,
+      "loss": 1.8504,
+      "step": 6720
+    },
+    {
+      "epoch": 0.8334882568011402,
+      "grad_norm": 0.13276523815484698,
+      "learning_rate": 1.641249449828699e-05,
+      "loss": 1.9055,
+      "step": 6725
+    },
+    {
+      "epoch": 0.8341079506723679,
+      "grad_norm": 0.13386303228577315,
+      "learning_rate": 1.6293940547214905e-05,
+      "loss": 1.7434,
+      "step": 6730
+    },
+    {
+      "epoch": 0.8347276445435955,
+      "grad_norm": 0.1383082816059096,
+      "learning_rate": 1.617577834127506e-05,
+      "loss": 1.7766,
+      "step": 6735
+    },
+    {
+      "epoch": 0.8353473384148231,
+      "grad_norm": 0.13758595605104323,
+      "learning_rate": 1.6058008433467698e-05,
+      "loss": 1.7832,
+      "step": 6740
+    },
+    {
+      "epoch": 0.8359670322860507,
+      "grad_norm": 0.13142188279747996,
+      "learning_rate": 1.594063137495707e-05,
+      "loss": 1.8305,
+      "step": 6745
+    },
+    {
+      "epoch": 0.8365867261572784,
+      "grad_norm": 0.13834331995085114,
+      "learning_rate": 1.582364771506891e-05,
+      "loss": 1.8633,
+      "step": 6750
+    },
+    {
+      "epoch": 0.8372064200285059,
+      "grad_norm": 0.1426359228453992,
+      "learning_rate": 1.570705800128781e-05,
+      "loss": 1.7992,
+      "step": 6755
+    },
+    {
+      "epoch": 0.8378261138997335,
+      "grad_norm": 0.14010058644661322,
+      "learning_rate": 1.5590862779254746e-05,
+      "loss": 1.8043,
+      "step": 6760
+    },
+    {
+      "epoch": 0.8384458077709611,
+      "grad_norm": 0.1363378846450556,
+      "learning_rate": 1.5475062592764346e-05,
+      "loss": 1.8078,
+      "step": 6765
+    },
+    {
+      "epoch": 0.8390655016421887,
+      "grad_norm": 0.13928139362824718,
+      "learning_rate": 1.5359657983762632e-05,
+      "loss": 1.8055,
+      "step": 6770
+    },
+    {
+      "epoch": 0.8396851955134164,
+      "grad_norm": 0.1372046018249784,
+      "learning_rate": 1.524464949234422e-05,
+      "loss": 1.8211,
+      "step": 6775
+    },
+    {
+      "epoch": 0.840304889384644,
+      "grad_norm": 0.13039972587829407,
+      "learning_rate": 1.5130037656749918e-05,
+      "loss": 1.7777,
+      "step": 6780
+    },
+    {
+      "epoch": 0.8409245832558716,
+      "grad_norm": 0.13925521199854124,
+      "learning_rate": 1.5015823013364183e-05,
+      "loss": 1.8609,
+      "step": 6785
+    },
+    {
+      "epoch": 0.8415442771270992,
+      "grad_norm": 0.1318479986936156,
+      "learning_rate": 1.4902006096712572e-05,
+      "loss": 1.7832,
+      "step": 6790
+    },
+    {
+      "epoch": 0.8421639709983268,
+      "grad_norm": 0.13360512591823362,
+      "learning_rate": 1.4788587439459323e-05,
+      "loss": 1.8039,
+      "step": 6795
+    },
+    {
+      "epoch": 0.8427836648695545,
+      "grad_norm": 0.13712021933762084,
+      "learning_rate": 1.4675567572404803e-05,
+      "loss": 1.8434,
+      "step": 6800
+    },
+    {
+      "epoch": 0.8434033587407821,
+      "grad_norm": 0.13608874848867214,
+      "learning_rate": 1.4562947024483031e-05,
+      "loss": 1.7625,
+      "step": 6805
+    },
+    {
+      "epoch": 0.8440230526120097,
+      "grad_norm": 0.13426634730012743,
+      "learning_rate": 1.4450726322759223e-05,
+      "loss": 1.8086,
+      "step": 6810
+    },
+    {
+      "epoch": 0.8446427464832372,
+      "grad_norm": 0.13382844194945753,
+      "learning_rate": 1.4338905992427287e-05,
+      "loss": 1.8695,
+      "step": 6815
+    },
+    {
+      "epoch": 0.8452624403544649,
+      "grad_norm": 0.14966906376149425,
+      "learning_rate": 1.4227486556807412e-05,
+      "loss": 1.8121,
+      "step": 6820
+    },
+    {
+      "epoch": 0.8458821342256925,
+      "grad_norm": 0.13735951665790422,
+      "learning_rate": 1.4116468537343585e-05,
+      "loss": 1.8242,
+      "step": 6825
+    },
+    {
+      "epoch": 0.8465018280969201,
+      "grad_norm": 0.14377264999175535,
+      "learning_rate": 1.4005852453601164e-05,
+      "loss": 1.8648,
+      "step": 6830
+    },
+    {
+      "epoch": 0.8471215219681477,
+      "grad_norm": 0.1350064432331249,
+      "learning_rate": 1.3895638823264446e-05,
+      "loss": 1.8465,
+      "step": 6835
+    },
+    {
+      "epoch": 0.8477412158393753,
+      "grad_norm": 0.13075932246700298,
+      "learning_rate": 1.3785828162134252e-05,
+      "loss": 1.8348,
+      "step": 6840
+    },
+    {
+      "epoch": 0.848360909710603,
+      "grad_norm": 0.13247777446205836,
+      "learning_rate": 1.367642098412546e-05,
+      "loss": 1.8629,
+      "step": 6845
+    },
+    {
+      "epoch": 0.8489806035818306,
+      "grad_norm": 0.13338219328252035,
+      "learning_rate": 1.3567417801264692e-05,
+      "loss": 1.8297,
+      "step": 6850
+    },
+    {
+      "epoch": 0.8496002974530582,
+      "grad_norm": 0.13787557857168983,
+      "learning_rate": 1.345881912368785e-05,
+      "loss": 1.8621,
+      "step": 6855
+    },
+    {
+      "epoch": 0.8502199913242858,
+      "grad_norm": 0.13952008714498,
+      "learning_rate": 1.3350625459637744e-05,
+      "loss": 1.8117,
+      "step": 6860
+    },
+    {
+      "epoch": 0.8508396851955135,
+      "grad_norm": 0.13937544594263107,
+      "learning_rate": 1.3242837315461732e-05,
+      "loss": 1.8418,
+      "step": 6865
+    },
+    {
+      "epoch": 0.8514593790667411,
+      "grad_norm": 0.13851152521957338,
+      "learning_rate": 1.3135455195609325e-05,
+      "loss": 1.8137,
+      "step": 6870
+    },
+    {
+      "epoch": 0.8520790729379687,
+      "grad_norm": 0.13245292293717914,
+      "learning_rate": 1.3028479602629839e-05,
+      "loss": 1.8445,
+      "step": 6875
+    },
+    {
+      "epoch": 0.8526987668091962,
+      "grad_norm": 0.13176137950037486,
+      "learning_rate": 1.292191103717002e-05,
+      "loss": 1.8457,
+      "step": 6880
+    },
+    {
+      "epoch": 0.8533184606804238,
+      "grad_norm": 0.14274560973973033,
+      "learning_rate": 1.281574999797176e-05,
+      "loss": 1.8793,
+      "step": 6885
+    },
+    {
+      "epoch": 0.8539381545516515,
+      "grad_norm": 0.13669793644265746,
+      "learning_rate": 1.2709996981869699e-05,
+      "loss": 1.7793,
+      "step": 6890
+    },
+    {
+      "epoch": 0.8545578484228791,
+      "grad_norm": 0.13401311583823233,
+      "learning_rate": 1.2604652483788948e-05,
+      "loss": 1.841,
+      "step": 6895
+    },
+    {
+      "epoch": 0.8551775422941067,
+      "grad_norm": 0.14562512309081663,
+      "learning_rate": 1.2499716996742694e-05,
+      "loss": 1.8477,
+      "step": 6900
+    },
+    {
+      "epoch": 0.8557972361653343,
+      "grad_norm": 0.13981374112922862,
+      "learning_rate": 1.2395191011829999e-05,
+      "loss": 1.8508,
+      "step": 6905
+    },
+    {
+      "epoch": 0.8564169300365619,
+      "grad_norm": 0.13795922585737252,
+      "learning_rate": 1.2291075018233445e-05,
+      "loss": 1.8637,
+      "step": 6910
+    },
+    {
+      "epoch": 0.8570366239077896,
+      "grad_norm": 0.13648363243463277,
+      "learning_rate": 1.218736950321685e-05,
+      "loss": 1.7707,
+      "step": 6915
+    },
+    {
+      "epoch": 0.8576563177790172,
+      "grad_norm": 0.13536889404424596,
+      "learning_rate": 1.2084074952122959e-05,
+      "loss": 1.8629,
+      "step": 6920
+    },
+    {
+      "epoch": 0.8582760116502448,
+      "grad_norm": 0.14023215465264866,
+      "learning_rate": 1.1981191848371287e-05,
+      "loss": 1.8355,
+      "step": 6925
+    },
+    {
+      "epoch": 0.8588957055214724,
+      "grad_norm": 0.13669990706649254,
+      "learning_rate": 1.1878720673455645e-05,
+      "loss": 1.8586,
+      "step": 6930
+    },
+    {
+      "epoch": 0.8595153993927,
+      "grad_norm": 0.13993547428958963,
+      "learning_rate": 1.1776661906942099e-05,
+      "loss": 1.8172,
+      "step": 6935
+    },
+    {
+      "epoch": 0.8601350932639277,
+      "grad_norm": 0.15053693553773634,
+      "learning_rate": 1.1675016026466633e-05,
+      "loss": 1.7953,
+      "step": 6940
+    },
+    {
+      "epoch": 0.8607547871351552,
+      "grad_norm": 0.1400512786653631,
+      "learning_rate": 1.1573783507732893e-05,
+      "loss": 1.8559,
+      "step": 6945
+    },
+    {
+      "epoch": 0.8613744810063828,
+      "grad_norm": 0.14309923516749187,
+      "learning_rate": 1.1472964824510035e-05,
+      "loss": 1.8348,
+      "step": 6950
+    },
+    {
+      "epoch": 0.8619941748776104,
+      "grad_norm": 0.13484282365125583,
+      "learning_rate": 1.1372560448630376e-05,
+      "loss": 1.809,
+      "step": 6955
+    },
+    {
+      "epoch": 0.862613868748838,
+      "grad_norm": 0.13986355786084606,
+      "learning_rate": 1.1272570849987351e-05,
+      "loss": 1.8152,
+      "step": 6960
+    },
+    {
+      "epoch": 0.8632335626200657,
+      "grad_norm": 0.1378893093004656,
+      "learning_rate": 1.1172996496533194e-05,
+      "loss": 1.7738,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8638532564912933,
+      "grad_norm": 0.1345430991025174,
+      "learning_rate": 1.1073837854276826e-05,
+      "loss": 1.8082,
+      "step": 6970
+    },
+    {
+      "epoch": 0.8644729503625209,
+      "grad_norm": 0.1441895547115774,
+      "learning_rate": 1.0975095387281587e-05,
+      "loss": 1.8234,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8650926442337485,
+      "grad_norm": 0.1374998969278752,
+      "learning_rate": 1.087676955766318e-05,
+      "loss": 1.841,
+      "step": 6980
+    },
+    {
+      "epoch": 0.8657123381049762,
+      "grad_norm": 0.13606561355593746,
+      "learning_rate": 1.0778860825587323e-05,
+      "loss": 1.8352,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8663320319762038,
+      "grad_norm": 0.13749250934423712,
+      "learning_rate": 1.0681369649267836e-05,
+      "loss": 1.8496,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8669517258474314,
+      "grad_norm": 0.1368331262299894,
+      "learning_rate": 1.0584296484964318e-05,
+      "loss": 1.8711,
+      "step": 6995
+    },
+    {
+      "epoch": 0.867571419718659,
+      "grad_norm": 0.13863183942887752,
+      "learning_rate": 1.0487641786980063e-05,
+      "loss": 1.8391,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8681911135898865,
+      "grad_norm": 0.13434259874384205,
+      "learning_rate": 1.0391406007659964e-05,
+      "loss": 1.7699,
+      "step": 7005
+    },
+    {
+      "epoch": 0.8688108074611142,
+      "grad_norm": 0.14095478527168656,
+      "learning_rate": 1.0295589597388355e-05,
+      "loss": 1.8336,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8694305013323418,
+      "grad_norm": 0.1348345066351467,
+      "learning_rate": 1.0200193004586922e-05,
+      "loss": 1.802,
+      "step": 7015
+    },
+    {
+      "epoch": 0.8700501952035694,
+      "grad_norm": 0.13470060368449044,
+      "learning_rate": 1.0105216675712592e-05,
+      "loss": 1.8281,
+      "step": 7020
+    },
+    {
+      "epoch": 0.870669889074797,
+      "grad_norm": 0.1498496809323783,
+      "learning_rate": 1.0010661055255488e-05,
+      "loss": 1.8391,
+      "step": 7025
+    },
+    {
+      "epoch": 0.8712895829460247,
+      "grad_norm": 0.146463161739714,
+      "learning_rate": 9.916526585736763e-06,
+      "loss": 1.8535,
+      "step": 7030
+    },
+    {
+      "epoch": 0.8719092768172523,
+      "grad_norm": 0.14092293708406953,
+      "learning_rate": 9.822813707706625e-06,
+      "loss": 1.8406,
+      "step": 7035
+    },
+    {
+      "epoch": 0.8725289706884799,
+      "grad_norm": 0.14019511041970145,
+      "learning_rate": 9.729522859742191e-06,
+      "loss": 1.8938,
+      "step": 7040
+    },
+    {
+      "epoch": 0.8731486645597075,
+      "grad_norm": 0.1411623147758512,
+      "learning_rate": 9.636654478445494e-06,
+      "loss": 1.8387,
+      "step": 7045
+    },
+    {
+      "epoch": 0.8737683584309351,
+      "grad_norm": 0.1401437249903466,
+      "learning_rate": 9.544208998441428e-06,
+      "loss": 1.8867,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8743880523021628,
+      "grad_norm": 0.13345487580370216,
+      "learning_rate": 9.452186852375678e-06,
+      "loss": 1.8891,
+      "step": 7055
+    },
+    {
+      "epoch": 0.8750077461733904,
+      "grad_norm": 0.13538270401639796,
+      "learning_rate": 9.360588470912756e-06,
+      "loss": 1.8609,
+      "step": 7060
+    },
+    {
+      "epoch": 0.875627440044618,
+      "grad_norm": 0.13939914091565453,
+      "learning_rate": 9.269414282733924e-06,
+      "loss": 1.8781,
+      "step": 7065
+    },
+    {
+      "epoch": 0.8762471339158455,
+      "grad_norm": 0.138068735546144,
+      "learning_rate": 9.178664714535235e-06,
+      "loss": 1.8359,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8768668277870731,
+      "grad_norm": 0.13765544825239814,
+      "learning_rate": 9.088340191025501e-06,
+      "loss": 1.8586,
+      "step": 7075
+    },
+    {
+      "epoch": 0.8774865216583008,
+      "grad_norm": 0.13486521127469114,
+      "learning_rate": 8.998441134924318e-06,
+      "loss": 1.8758,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8781062155295284,
+      "grad_norm": 0.13651144599742507,
+      "learning_rate": 8.908967966960124e-06,
+      "loss": 1.8922,
+      "step": 7085
+    },
+    {
+      "epoch": 0.878725909400756,
+      "grad_norm": 0.1447116043420326,
+      "learning_rate": 8.819921105868167e-06,
+      "loss": 1.8168,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8793456032719836,
+      "grad_norm": 0.12956578472131167,
+      "learning_rate": 8.731300968388556e-06,
+      "loss": 1.8375,
+      "step": 7095
+    },
+    {
+      "epoch": 0.8799652971432113,
+      "grad_norm": 0.14217108418937452,
+      "learning_rate": 8.643107969264375e-06,
+      "loss": 1.8793,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8805849910144389,
+      "grad_norm": 0.1475313460315241,
+      "learning_rate": 8.555342521239662e-06,
+      "loss": 1.8191,
+      "step": 7105
+    },
+    {
+      "epoch": 0.8812046848856665,
+      "grad_norm": 0.14056958492979643,
+      "learning_rate": 8.468005035057536e-06,
+      "loss": 1.8664,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8818243787568941,
+      "grad_norm": 0.14479567172868937,
+      "learning_rate": 8.381095919458226e-06,
+      "loss": 1.8539,
+      "step": 7115
+    },
+    {
+      "epoch": 0.8824440726281217,
+      "grad_norm": 0.14164836437320075,
+      "learning_rate": 8.294615581177223e-06,
+      "loss": 1.8445,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8830637664993494,
+      "grad_norm": 0.14339217053169717,
+      "learning_rate": 8.208564424943288e-06,
+      "loss": 1.8113,
+      "step": 7125
+    },
+    {
+      "epoch": 0.883683460370577,
+      "grad_norm": 0.13508727975928309,
+      "learning_rate": 8.122942853476633e-06,
+      "loss": 1.8961,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8843031542418045,
+      "grad_norm": 0.1431420751411207,
+      "learning_rate": 8.037751267487003e-06,
+      "loss": 1.852,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8849228481130321,
+      "grad_norm": 0.13776837332926897,
+      "learning_rate": 7.952990065671817e-06,
+      "loss": 1.8367,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8855425419842597,
+      "grad_norm": 0.1357035406570165,
+      "learning_rate": 7.868659644714294e-06,
+      "loss": 1.8164,
+      "step": 7145
+    },
+    {
+      "epoch": 0.8861622358554874,
+      "grad_norm": 0.13774817584994603,
+      "learning_rate": 7.784760399281554e-06,
+      "loss": 1.8715,
+      "step": 7150
+    },
+    {
+      "epoch": 0.886781929726715,
+      "grad_norm": 0.13709267085065505,
+      "learning_rate": 7.701292722022846e-06,
+      "loss": 1.8035,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8874016235979426,
+      "grad_norm": 0.1381104044552148,
+      "learning_rate": 7.618257003567675e-06,
+      "loss": 1.8172,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8880213174691702,
+      "grad_norm": 0.13578670341989943,
+      "learning_rate": 7.5356536325239755e-06,
+      "loss": 1.7742,
+      "step": 7165
+    },
+    {
+      "epoch": 0.8886410113403979,
+      "grad_norm": 0.1375594605300216,
+      "learning_rate": 7.453482995476291e-06,
+      "loss": 1.8172,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8892607052116255,
+      "grad_norm": 0.13878029380784612,
+      "learning_rate": 7.371745476983982e-06,
+      "loss": 1.7582,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8898803990828531,
+      "grad_norm": 0.13908494179294276,
+      "learning_rate": 7.2904414595793556e-06,
+      "loss": 1.8398,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8905000929540807,
+      "grad_norm": 0.13339193424993012,
+      "learning_rate": 7.209571323765973e-06,
+      "loss": 1.7902,
+      "step": 7185
+    },
+    {
+      "epoch": 0.8911197868253083,
+      "grad_norm": 0.1331999537324601,
+      "learning_rate": 7.129135448016821e-06,
+      "loss": 1.8301,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8917394806965359,
+      "grad_norm": 0.13428548155659575,
+      "learning_rate": 7.049134208772545e-06,
+      "loss": 1.852,
+      "step": 7195
+    },
+    {
+      "epoch": 0.8923591745677635,
+      "grad_norm": 0.13624592336240599,
+      "learning_rate": 6.969567980439706e-06,
+      "loss": 1.8602,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8929788684389911,
+      "grad_norm": 0.13683342194643683,
+      "learning_rate": 6.890437135388939e-06,
+      "loss": 1.8613,
+      "step": 7205
+    },
+    {
+      "epoch": 0.8935985623102187,
+      "grad_norm": 0.13394086830888138,
+      "learning_rate": 6.8117420439533615e-06,
+      "loss": 1.8535,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8942182561814463,
+      "grad_norm": 0.1308504194307238,
+      "learning_rate": 6.733483074426716e-06,
+      "loss": 1.8527,
+      "step": 7215
+    },
+    {
+      "epoch": 0.894837950052674,
+      "grad_norm": 0.142886229081844,
+      "learning_rate": 6.655660593061719e-06,
+      "loss": 1.8387,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8954576439239016,
+      "grad_norm": 0.1394768164104558,
+      "learning_rate": 6.578274964068298e-06,
+      "loss": 1.891,
+      "step": 7225
+    },
+    {
+      "epoch": 0.8960773377951292,
+      "grad_norm": 0.13125339904213718,
+      "learning_rate": 6.50132654961193e-06,
+      "loss": 1.8387,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8966970316663568,
+      "grad_norm": 0.14041832103312948,
+      "learning_rate": 6.424815709811871e-06,
+      "loss": 1.8852,
+      "step": 7235
+    },
+    {
+      "epoch": 0.8973167255375845,
+      "grad_norm": 0.13965593340949,
+      "learning_rate": 6.3487428027395715e-06,
+      "loss": 1.7867,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8979364194088121,
+      "grad_norm": 0.1326849396556491,
+      "learning_rate": 6.273108184416943e-06,
+      "loss": 1.8414,
+      "step": 7245
+    },
+    {
+      "epoch": 0.8985561132800397,
+      "grad_norm": 0.14154992344541478,
+      "learning_rate": 6.197912208814694e-06,
+      "loss": 1.8711,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8991758071512673,
+      "grad_norm": 0.14900298957887134,
+      "learning_rate": 6.123155227850708e-06,
+      "loss": 1.8496,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8997955010224948,
+      "grad_norm": 0.13378506525283548,
+      "learning_rate": 6.048837591388301e-06,
+      "loss": 1.8355,
+      "step": 7260
+    },
+    {
+      "epoch": 0.9004151948937225,
+      "grad_norm": 0.1364642041434309,
+      "learning_rate": 5.974959647234746e-06,
+      "loss": 1.827,
+      "step": 7265
+    },
+    {
+      "epoch": 0.9010348887649501,
+      "grad_norm": 0.13884954804156385,
+      "learning_rate": 5.901521741139482e-06,
+      "loss": 1.8055,
+      "step": 7270
+    },
+    {
+      "epoch": 0.9016545826361777,
+      "grad_norm": 0.13583763712007216,
+      "learning_rate": 5.828524216792586e-06,
+      "loss": 1.7793,
+      "step": 7275
+    },
+    {
+      "epoch": 0.9022742765074053,
+      "grad_norm": 0.1373270304349077,
+      "learning_rate": 5.75596741582316e-06,
+      "loss": 1.8512,
+      "step": 7280
+    },
+    {
+      "epoch": 0.902893970378633,
+      "grad_norm": 0.1352208909295617,
+      "learning_rate": 5.6838516777977135e-06,
+      "loss": 1.8242,
+      "step": 7285
+    },
+    {
+      "epoch": 0.9035136642498606,
+      "grad_norm": 0.14464708429301287,
+      "learning_rate": 5.6121773402185385e-06,
+      "loss": 1.7852,
+      "step": 7290
+    },
+    {
+      "epoch": 0.9041333581210882,
+      "grad_norm": 0.13241462166706447,
+      "learning_rate": 5.540944738522203e-06,
+      "loss": 1.8809,
+      "step": 7295
+    },
+    {
+      "epoch": 0.9047530519923158,
+      "grad_norm": 0.13280584084305314,
+      "learning_rate": 5.470154206077949e-06,
+      "loss": 1.8277,
+      "step": 7300
+    },
+    {
+      "epoch": 0.9053727458635434,
+      "grad_norm": 0.14278733047702366,
+      "learning_rate": 5.3998060741861314e-06,
+      "loss": 1.9148,
+      "step": 7305
+    },
+    {
+      "epoch": 0.9059924397347711,
+      "grad_norm": 0.14264404944438486,
+      "learning_rate": 5.329900672076637e-06,
+      "loss": 1.8313,
+      "step": 7310
+    },
+    {
+      "epoch": 0.9066121336059987,
+      "grad_norm": 0.13404018286932412,
+      "learning_rate": 5.260438326907413e-06,
+      "loss": 1.8496,
+      "step": 7315
+    },
+    {
+      "epoch": 0.9072318274772263,
+      "grad_norm": 0.139299854893475,
+      "learning_rate": 5.191419363762873e-06,
+      "loss": 1.8215,
+      "step": 7320
+    },
+    {
+      "epoch": 0.9078515213484538,
+      "grad_norm": 0.12938710234881426,
+      "learning_rate": 5.122844105652402e-06,
+      "loss": 1.9102,
+      "step": 7325
+    },
+    {
+      "epoch": 0.9084712152196814,
+      "grad_norm": 0.1431339768847217,
+      "learning_rate": 5.054712873508827e-06,
+      "loss": 1.8809,
+      "step": 7330
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.14185976129658012,
+      "learning_rate": 4.987025986186966e-06,
+      "loss": 1.8551,
+      "step": 7335
+    },
+    {
+      "epoch": 0.9097106029621367,
+      "grad_norm": 0.13984745090181067,
+      "learning_rate": 4.919783760462082e-06,
+      "loss": 1.7578,
+      "step": 7340
+    },
+    {
+      "epoch": 0.9103302968333643,
+      "grad_norm": 0.1316498232846164,
+      "learning_rate": 4.85298651102839e-06,
+      "loss": 1.8746,
+      "step": 7345
+    },
+    {
+      "epoch": 0.9109499907045919,
+      "grad_norm": 0.13622065727170202,
+      "learning_rate": 4.786634550497637e-06,
+      "loss": 1.8641,
+      "step": 7350
+    },
+    {
+      "epoch": 0.9115696845758195,
+      "grad_norm": 0.13872927734480942,
+      "learning_rate": 4.720728189397628e-06,
+      "loss": 1.7785,
+      "step": 7355
+    },
+    {
+      "epoch": 0.9121893784470472,
+      "grad_norm": 0.14013017608713682,
+      "learning_rate": 4.655267736170732e-06,
+      "loss": 1.8273,
+      "step": 7360
+    },
+    {
+      "epoch": 0.9128090723182748,
+      "grad_norm": 0.140766561179319,
+      "learning_rate": 4.5902534971724806e-06,
+      "loss": 1.8023,
+      "step": 7365
+    },
+    {
+      "epoch": 0.9134287661895024,
+      "grad_norm": 0.13550910020269402,
+      "learning_rate": 4.525685776670108e-06,
+      "loss": 1.7758,
+      "step": 7370
+    },
+    {
+      "epoch": 0.91404846006073,
+      "grad_norm": 0.1456560002195272,
+      "learning_rate": 4.46156487684114e-06,
+      "loss": 1.8203,
+      "step": 7375
+    },
+    {
+      "epoch": 0.9146681539319577,
+      "grad_norm": 0.13985502488308021,
+      "learning_rate": 4.397891097771989e-06,
+      "loss": 1.8293,
+      "step": 7380
+    },
+    {
+      "epoch": 0.9152878478031852,
+      "grad_norm": 0.14148217516325698,
+      "learning_rate": 4.334664737456539e-06,
+      "loss": 1.8039,
+      "step": 7385
+    },
+    {
+      "epoch": 0.9159075416744128,
+      "grad_norm": 0.13723569523716944,
+      "learning_rate": 4.271886091794719e-06,
+      "loss": 1.8836,
+      "step": 7390
+    },
+    {
+      "epoch": 0.9165272355456404,
+      "grad_norm": 0.13503953990532955,
+      "learning_rate": 4.209555454591197e-06,
+      "loss": 1.7797,
+      "step": 7395
+    },
+    {
+      "epoch": 0.917146929416868,
+      "grad_norm": 0.13814876703450732,
+      "learning_rate": 4.147673117553896e-06,
+      "loss": 1.8531,
+      "step": 7400
+    },
+    {
+      "epoch": 0.9177666232880957,
+      "grad_norm": 0.13586514530672164,
+      "learning_rate": 4.086239370292755e-06,
+      "loss": 1.848,
+      "step": 7405
+    },
+    {
+      "epoch": 0.9183863171593233,
+      "grad_norm": 0.14182653498050868,
+      "learning_rate": 4.025254500318265e-06,
+      "loss": 1.8422,
+      "step": 7410
+    },
+    {
+      "epoch": 0.9190060110305509,
+      "grad_norm": 0.1388334635973296,
+      "learning_rate": 3.964718793040178e-06,
+      "loss": 1.8148,
+      "step": 7415
+    },
+    {
+      "epoch": 0.9196257049017785,
+      "grad_norm": 0.1423659419451689,
+      "learning_rate": 3.904632531766195e-06,
+      "loss": 1.7766,
+      "step": 7420
+    },
+    {
+      "epoch": 0.9202453987730062,
+      "grad_norm": 0.14335609098468574,
+      "learning_rate": 3.84499599770054e-06,
+      "loss": 1.8473,
+      "step": 7425
+    },
+    {
+      "epoch": 0.9208650926442338,
+      "grad_norm": 0.13248049762452474,
+      "learning_rate": 3.785809469942758e-06,
+      "loss": 1.8238,
+      "step": 7430
+    },
+    {
+      "epoch": 0.9214847865154614,
+      "grad_norm": 0.13729453036726105,
+      "learning_rate": 3.727073225486344e-06,
+      "loss": 1.8168,
+      "step": 7435
+    },
+    {
+      "epoch": 0.922104480386689,
+      "grad_norm": 0.1353112000663283,
+      "learning_rate": 3.6687875392174665e-06,
+      "loss": 1.8223,
+      "step": 7440
+    },
+    {
+      "epoch": 0.9227241742579166,
+      "grad_norm": 0.1447311577090011,
+      "learning_rate": 3.61095268391366e-06,
+      "loss": 1.8316,
+      "step": 7445
+    },
+    {
+      "epoch": 0.9233438681291442,
+      "grad_norm": 0.13228279149862204,
+      "learning_rate": 3.5535689302426236e-06,
+      "loss": 1.8383,
+      "step": 7450
+    },
+    {
+      "epoch": 0.9239635620003718,
+      "grad_norm": 0.13613961418557433,
+      "learning_rate": 3.496636546760812e-06,
+      "loss": 1.777,
+      "step": 7455
+    },
+    {
+      "epoch": 0.9245832558715994,
+      "grad_norm": 0.14235960272198284,
+      "learning_rate": 3.4401557999123146e-06,
+      "loss": 1.8613,
+      "step": 7460
+    },
+    {
+      "epoch": 0.925202949742827,
+      "grad_norm": 0.12842862529934493,
+      "learning_rate": 3.3841269540275553e-06,
+      "loss": 1.8629,
+      "step": 7465
+    },
+    {
+      "epoch": 0.9258226436140546,
+      "grad_norm": 0.13216080090217125,
+      "learning_rate": 3.3285502713220617e-06,
+      "loss": 1.9145,
+      "step": 7470
+    },
+    {
+      "epoch": 0.9264423374852823,
+      "grad_norm": 0.13750703177725587,
+      "learning_rate": 3.2734260118952307e-06,
+      "loss": 1.8406,
+      "step": 7475
+    },
+    {
+      "epoch": 0.9270620313565099,
+      "grad_norm": 0.1362535989445485,
+      "learning_rate": 3.218754433729065e-06,
+      "loss": 1.8477,
+      "step": 7480
+    },
+    {
+      "epoch": 0.9276817252277375,
+      "grad_norm": 0.13293349873817525,
+      "learning_rate": 3.1645357926870955e-06,
+      "loss": 1.8473,
+      "step": 7485
+    },
+    {
+      "epoch": 0.9283014190989651,
+      "grad_norm": 0.12370366878412518,
+      "learning_rate": 3.110770342513036e-06,
+      "loss": 1.8441,
+      "step": 7490
+    },
+    {
+      "epoch": 0.9289211129701928,
+      "grad_norm": 0.14093551922685052,
+      "learning_rate": 3.057458334829699e-06,
+      "loss": 1.8262,
+      "step": 7495
+    },
+    {
+      "epoch": 0.9295408068414204,
+      "grad_norm": 0.14089913168838317,
+      "learning_rate": 3.0046000191377934e-06,
+      "loss": 1.8363,
+      "step": 7500
+    },
+    {
+      "epoch": 0.930160500712648,
+      "grad_norm": 0.13655806820666927,
+      "learning_rate": 2.9521956428146923e-06,
+      "loss": 1.7809,
+      "step": 7505
+    },
+    {
+      "epoch": 0.9307801945838756,
+      "grad_norm": 0.12930406578508882,
+      "learning_rate": 2.9002454511133923e-06,
+      "loss": 1.8426,
+      "step": 7510
+    },
+    {
+      "epoch": 0.9313998884551031,
+      "grad_norm": 0.13223003361326363,
+      "learning_rate": 2.8487496871612453e-06,
+      "loss": 1.8457,
+      "step": 7515
+    },
+    {
+      "epoch": 0.9320195823263308,
+      "grad_norm": 0.13191348370431163,
+      "learning_rate": 2.7977085919589254e-06,
+      "loss": 1.8461,
+      "step": 7520
+    },
+    {
+      "epoch": 0.9326392761975584,
+      "grad_norm": 0.13149475248273712,
+      "learning_rate": 2.7471224043792098e-06,
+      "loss": 1.8406,
+      "step": 7525
+    },
+    {
+      "epoch": 0.933258970068786,
+      "grad_norm": 0.13701793086079692,
+      "learning_rate": 2.6969913611659457e-06,
+      "loss": 1.907,
+      "step": 7530
+    },
+    {
+      "epoch": 0.9338786639400136,
+      "grad_norm": 0.132647067429414,
+      "learning_rate": 2.6473156969328503e-06,
+      "loss": 1.7945,
+      "step": 7535
+    },
+    {
+      "epoch": 0.9344983578112412,
+      "grad_norm": 0.13813325094756712,
+      "learning_rate": 2.5980956441625236e-06,
+      "loss": 1.802,
+      "step": 7540
+    },
+    {
+      "epoch": 0.9351180516824689,
+      "grad_norm": 0.13672748834661805,
+      "learning_rate": 2.5493314332052377e-06,
+      "loss": 1.8391,
+      "step": 7545
+    },
+    {
+      "epoch": 0.9357377455536965,
+      "grad_norm": 0.14163135141821864,
+      "learning_rate": 2.501023292277971e-06,
+      "loss": 1.85,
+      "step": 7550
+    },
+    {
+      "epoch": 0.9363574394249241,
+      "grad_norm": 0.13621074708206138,
+      "learning_rate": 2.453171447463265e-06,
+      "loss": 1.8648,
+      "step": 7555
+    },
+    {
+      "epoch": 0.9369771332961517,
+      "grad_norm": 0.143929080240879,
+      "learning_rate": 2.4057761227081923e-06,
+      "loss": 1.793,
+      "step": 7560
+    },
+    {
+      "epoch": 0.9375968271673794,
+      "grad_norm": 0.13558520445018998,
+      "learning_rate": 2.358837539823311e-06,
+      "loss": 1.8395,
+      "step": 7565
+    },
+    {
+      "epoch": 0.938216521038607,
+      "grad_norm": 0.1321488531469034,
+      "learning_rate": 2.3123559184816344e-06,
+      "loss": 1.7812,
+      "step": 7570
+    },
+    {
+      "epoch": 0.9388362149098345,
+      "grad_norm": 0.13733894017904172,
+      "learning_rate": 2.2663314762175647e-06,
+      "loss": 1.8727,
+      "step": 7575
+    },
+    {
+      "epoch": 0.9394559087810621,
+      "grad_norm": 0.13649001800785526,
+      "learning_rate": 2.2207644284259256e-06,
+      "loss": 1.8367,
+      "step": 7580
+    },
+    {
+      "epoch": 0.9400756026522897,
+      "grad_norm": 0.136966039329494,
+      "learning_rate": 2.1756549883609313e-06,
+      "loss": 1.8402,
+      "step": 7585
+    },
+    {
+      "epoch": 0.9406952965235174,
+      "grad_norm": 0.1323250881965697,
+      "learning_rate": 2.131003367135154e-06,
+      "loss": 1.8465,
+      "step": 7590
+    },
+    {
+      "epoch": 0.941314990394745,
+      "grad_norm": 0.1293883428081351,
+      "learning_rate": 2.086809773718601e-06,
+      "loss": 1.8121,
+      "step": 7595
+    },
+    {
+      "epoch": 0.9419346842659726,
+      "grad_norm": 0.13859211529936802,
+      "learning_rate": 2.0430744149377177e-06,
+      "loss": 1.8469,
+      "step": 7600
+    },
+    {
+      "epoch": 0.9425543781372002,
+      "grad_norm": 0.14096409695978848,
+      "learning_rate": 1.999797495474365e-06,
+      "loss": 1.8234,
+      "step": 7605
+    },
+    {
+      "epoch": 0.9431740720084278,
+      "grad_norm": 0.13640618682039807,
+      "learning_rate": 1.9569792178649405e-06,
+      "loss": 1.8687,
+      "step": 7610
+    },
+    {
+      "epoch": 0.9437937658796555,
+      "grad_norm": 0.12992959333703022,
+      "learning_rate": 1.914619782499383e-06,
+      "loss": 1.8938,
+      "step": 7615
+    },
+    {
+      "epoch": 0.9444134597508831,
+      "grad_norm": 0.14310321729194458,
+      "learning_rate": 1.8727193876202143e-06,
+      "loss": 1.8379,
+      "step": 7620
+    },
+    {
+      "epoch": 0.9450331536221107,
+      "grad_norm": 0.13433525346251443,
+      "learning_rate": 1.8312782293216979e-06,
+      "loss": 1.8344,
+      "step": 7625
+    },
+    {
+      "epoch": 0.9456528474933383,
+      "grad_norm": 0.131679794414834,
+      "learning_rate": 1.7902965015488381e-06,
+      "loss": 1.8887,
+      "step": 7630
+    },
+    {
+      "epoch": 0.946272541364566,
+      "grad_norm": 0.14674405226127388,
+      "learning_rate": 1.749774396096482e-06,
+      "loss": 1.8754,
+      "step": 7635
+    },
+    {
+      "epoch": 0.9468922352357935,
+      "grad_norm": 0.14156250279993213,
+      "learning_rate": 1.709712102608463e-06,
+      "loss": 1.8133,
+      "step": 7640
+    },
+    {
+      "epoch": 0.9475119291070211,
+      "grad_norm": 0.1383482897158417,
+      "learning_rate": 1.6701098085767031e-06,
+      "loss": 1.8031,
+      "step": 7645
+    },
+    {
+      "epoch": 0.9481316229782487,
+      "grad_norm": 0.1314285827216017,
+      "learning_rate": 1.630967699340269e-06,
+      "loss": 1.8559,
+      "step": 7650
+    },
+    {
+      "epoch": 0.9487513168494763,
+      "grad_norm": 0.1311075376517613,
+      "learning_rate": 1.5922859580846271e-06,
+      "loss": 1.8535,
+      "step": 7655
+    },
+    {
+      "epoch": 0.949371010720704,
+      "grad_norm": 0.14104548842656045,
+      "learning_rate": 1.5540647658406682e-06,
+      "loss": 1.8109,
+      "step": 7660
+    },
+    {
+      "epoch": 0.9499907045919316,
+      "grad_norm": 0.147336471013033,
+      "learning_rate": 1.5163043014839284e-06,
+      "loss": 1.7785,
+      "step": 7665
+    },
+    {
+      "epoch": 0.9506103984631592,
+      "grad_norm": 0.13166400385713117,
+      "learning_rate": 1.479004741733736e-06,
+      "loss": 1.8582,
+      "step": 7670
+    },
+    {
+      "epoch": 0.9512300923343868,
+      "grad_norm": 0.13572116993162228,
+      "learning_rate": 1.4421662611523667e-06,
+      "loss": 1.7457,
+      "step": 7675
+    },
+    {
+      "epoch": 0.9518497862056144,
+      "grad_norm": 0.14180501417698368,
+      "learning_rate": 1.4057890321442558e-06,
+      "loss": 1.8422,
+      "step": 7680
+    },
+    {
+      "epoch": 0.9524694800768421,
+      "grad_norm": 0.1401530258048634,
+      "learning_rate": 1.3698732249551648e-06,
+      "loss": 1.7668,
+      "step": 7685
+    },
+    {
+      "epoch": 0.9530891739480697,
+      "grad_norm": 0.13397195109554486,
+      "learning_rate": 1.3344190076714059e-06,
+      "loss": 1.9152,
+      "step": 7690
+    },
+    {
+      "epoch": 0.9537088678192973,
+      "grad_norm": 0.1338896772880402,
+      "learning_rate": 1.2994265462190513e-06,
+      "loss": 1.7973,
+      "step": 7695
+    },
+    {
+      "epoch": 0.9543285616905249,
+      "grad_norm": 0.13683971013494062,
+      "learning_rate": 1.2648960043631474e-06,
+      "loss": 1.8883,
+      "step": 7700
+    },
+    {
+      "epoch": 0.9549482555617524,
+      "grad_norm": 0.1340322150174887,
+      "learning_rate": 1.230827543706925e-06,
+      "loss": 1.7891,
+      "step": 7705
+    },
+    {
+      "epoch": 0.9555679494329801,
+      "grad_norm": 0.13540828994071039,
+      "learning_rate": 1.1972213236911112e-06,
+      "loss": 1.8691,
+      "step": 7710
+    },
+    {
+      "epoch": 0.9561876433042077,
+      "grad_norm": 0.13431224805457315,
+      "learning_rate": 1.1640775015931304e-06,
+      "loss": 1.8598,
+      "step": 7715
+    },
+    {
+      "epoch": 0.9568073371754353,
+      "grad_norm": 0.14041245444776923,
+      "learning_rate": 1.1313962325263717e-06,
+      "loss": 1.8453,
+      "step": 7720
+    },
+    {
+      "epoch": 0.9574270310466629,
+      "grad_norm": 0.13706149413387225,
+      "learning_rate": 1.0991776694394883e-06,
+      "loss": 1.8711,
+      "step": 7725
+    },
+    {
+      "epoch": 0.9580467249178906,
+      "grad_norm": 0.13378885841123092,
+      "learning_rate": 1.0674219631156334e-06,
+      "loss": 1.8477,
+      "step": 7730
+    },
+    {
+      "epoch": 0.9586664187891182,
+      "grad_norm": 0.13369307726742216,
+      "learning_rate": 1.0361292621718145e-06,
+      "loss": 1.9383,
+      "step": 7735
+    },
+    {
+      "epoch": 0.9592861126603458,
+      "grad_norm": 0.13470678685337034,
+      "learning_rate": 1.005299713058161e-06,
+      "loss": 1.8449,
+      "step": 7740
+    },
+    {
+      "epoch": 0.9599058065315734,
+      "grad_norm": 0.13827959853680175,
+      "learning_rate": 9.74933460057248e-07,
+      "loss": 1.8227,
+      "step": 7745
+    },
+    {
+      "epoch": 0.960525500402801,
+      "grad_norm": 0.13861638685269106,
+      "learning_rate": 9.450306452834179e-07,
+      "loss": 1.8598,
+      "step": 7750
+    },
+    {
+      "epoch": 0.9611451942740287,
+      "grad_norm": 0.13966148760312724,
+      "learning_rate": 9.15591408682126e-07,
+      "loss": 1.807,
+      "step": 7755
+    },
+    {
+      "epoch": 0.9617648881452563,
+      "grad_norm": 0.14081009116641513,
+      "learning_rate": 8.866158880292741e-07,
+      "loss": 1.782,
+      "step": 7760
+    },
+    {
+      "epoch": 0.9623845820164838,
+      "grad_norm": 0.14057620583361458,
+      "learning_rate": 8.581042189305555e-07,
+      "loss": 1.827,
+      "step": 7765
+    },
+    {
+      "epoch": 0.9630042758877114,
+      "grad_norm": 0.13937796915185546,
+      "learning_rate": 8.300565348208556e-07,
+      "loss": 1.8313,
+      "step": 7770
+    },
+    {
+      "epoch": 0.963623969758939,
+      "grad_norm": 0.13273514279259374,
+      "learning_rate": 8.024729669635967e-07,
+      "loss": 1.868,
+      "step": 7775
+    },
+    {
+      "epoch": 0.9642436636301667,
+      "grad_norm": 0.13598653532834637,
+      "learning_rate": 7.753536444501164e-07,
+      "loss": 1.7715,
+      "step": 7780
+    },
+    {
+      "epoch": 0.9648633575013943,
+      "grad_norm": 0.13907272985299546,
+      "learning_rate": 7.486986941991125e-07,
+      "loss": 1.7699,
+      "step": 7785
+    },
+    {
+      "epoch": 0.9654830513726219,
+      "grad_norm": 0.13321673642848147,
+      "learning_rate": 7.225082409559881e-07,
+      "loss": 1.8223,
+      "step": 7790
+    },
+    {
+      "epoch": 0.9661027452438495,
+      "grad_norm": 0.12879372395742059,
+      "learning_rate": 6.967824072923068e-07,
+      "loss": 1.768,
+      "step": 7795
+    },
+    {
+      "epoch": 0.9667224391150772,
+      "grad_norm": 0.13586715394976565,
+      "learning_rate": 6.715213136052056e-07,
+      "loss": 1.823,
+      "step": 7800
+    },
+    {
+      "epoch": 0.9673421329863048,
+      "grad_norm": 0.14497724016910185,
+      "learning_rate": 6.467250781168499e-07,
+      "loss": 1.8527,
+      "step": 7805
+    },
+    {
+      "epoch": 0.9679618268575324,
+      "grad_norm": 0.13612908889114356,
+      "learning_rate": 6.223938168738341e-07,
+      "loss": 1.8047,
+      "step": 7810
+    },
+    {
+      "epoch": 0.96858152072876,
+      "grad_norm": 0.1358005796129076,
+      "learning_rate": 5.985276437467046e-07,
+      "loss": 1.8422,
+      "step": 7815
+    },
+    {
+      "epoch": 0.9692012145999876,
+      "grad_norm": 0.14280390232394205,
+      "learning_rate": 5.751266704293601e-07,
+      "loss": 1.8168,
+      "step": 7820
+    },
+    {
+      "epoch": 0.9698209084712153,
+      "grad_norm": 0.13237756581743151,
+      "learning_rate": 5.521910064385627e-07,
+      "loss": 1.8402,
+      "step": 7825
+    },
+    {
+      "epoch": 0.9704406023424428,
+      "grad_norm": 0.13587015108604314,
+      "learning_rate": 5.297207591134612e-07,
+      "loss": 1.7957,
+      "step": 7830
+    },
+    {
+      "epoch": 0.9710602962136704,
+      "grad_norm": 0.13443918341905792,
+      "learning_rate": 5.077160336149911e-07,
+      "loss": 1.8457,
+      "step": 7835
+    },
+    {
+      "epoch": 0.971679990084898,
+      "grad_norm": 0.14765379085939143,
+      "learning_rate": 4.861769329254862e-07,
+      "loss": 1.8121,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9722996839561256,
+      "grad_norm": 0.1394457904355103,
+      "learning_rate": 4.651035578481344e-07,
+      "loss": 1.8746,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9729193778273533,
+      "grad_norm": 0.13752679317753222,
+      "learning_rate": 4.4449600700652296e-07,
+      "loss": 1.8461,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9735390716985809,
+      "grad_norm": 0.14296404692218137,
+      "learning_rate": 4.243543768441827e-07,
+      "loss": 1.85,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9741587655698085,
+      "grad_norm": 0.13087948478602446,
+      "learning_rate": 4.046787616241221e-07,
+      "loss": 1.8859,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9747784594410361,
+      "grad_norm": 0.1344356130652384,
+      "learning_rate": 3.8546925342842764e-07,
+      "loss": 1.7363,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9753981533122638,
+      "grad_norm": 0.1370666125721412,
+      "learning_rate": 3.6672594215774183e-07,
+      "loss": 1.8355,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9760178471834914,
+      "grad_norm": 0.13939543925141024,
+      "learning_rate": 3.484489155309523e-07,
+      "loss": 1.8383,
+      "step": 7875
+    },
+    {
+      "epoch": 0.976637541054719,
+      "grad_norm": 0.13817647293293314,
+      "learning_rate": 3.3063825908471456e-07,
+      "loss": 1.8535,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9772572349259466,
+      "grad_norm": 0.13705290483161534,
+      "learning_rate": 3.132940561730524e-07,
+      "loss": 1.7863,
+      "step": 7885
+    },
+    {
+      "epoch": 0.9778769287971742,
+      "grad_norm": 0.12778167892179987,
+      "learning_rate": 2.9641638796701344e-07,
+      "loss": 1.8418,
+      "step": 7890
+    },
+    {
+      "epoch": 0.9784966226684018,
+      "grad_norm": 0.13146765481563064,
+      "learning_rate": 2.800053334542363e-07,
+      "loss": 1.8629,
+      "step": 7895
+    },
+    {
+      "epoch": 0.9791163165396294,
+      "grad_norm": 0.13236765959057267,
+      "learning_rate": 2.6406096943859537e-07,
+      "loss": 1.8734,
+      "step": 7900
+    },
+    {
+      "epoch": 0.979736010410857,
+      "grad_norm": 0.1353069444117587,
+      "learning_rate": 2.485833705398677e-07,
+      "loss": 1.8191,
+      "step": 7905
+    },
+    {
+      "epoch": 0.9803557042820846,
+      "grad_norm": 0.1332817560129625,
+      "learning_rate": 2.3357260919336654e-07,
+      "loss": 1.8543,
+      "step": 7910
+    },
+    {
+      "epoch": 0.9809753981533123,
+      "grad_norm": 0.1427542111025049,
+      "learning_rate": 2.1902875564958624e-07,
+      "loss": 1.8406,
+      "step": 7915
+    },
+    {
+      "epoch": 0.9815950920245399,
+      "grad_norm": 0.13737706138886846,
+      "learning_rate": 2.0495187797390236e-07,
+      "loss": 1.8676,
+      "step": 7920
+    },
+    {
+      "epoch": 0.9822147858957675,
+      "grad_norm": 0.14196006592167704,
+      "learning_rate": 1.913420420462164e-07,
+      "loss": 1.8176,
+      "step": 7925
+    },
+    {
+      "epoch": 0.9828344797669951,
+      "grad_norm": 0.1346411791497383,
+      "learning_rate": 1.7819931156071168e-07,
+      "loss": 1.8715,
+      "step": 7930
+    },
+    {
+      "epoch": 0.9834541736382227,
+      "grad_norm": 0.13353971756702795,
+      "learning_rate": 1.6552374802546454e-07,
+      "loss": 1.8387,
+      "step": 7935
+    },
+    {
+      "epoch": 0.9840738675094504,
+      "grad_norm": 0.13534999592660807,
+      "learning_rate": 1.5331541076225585e-07,
+      "loss": 1.8742,
+      "step": 7940
+    },
+    {
+      "epoch": 0.984693561380678,
+      "grad_norm": 0.1375002640543611,
+      "learning_rate": 1.4157435690619337e-07,
+      "loss": 1.7785,
+      "step": 7945
+    },
+    {
+      "epoch": 0.9853132552519056,
+      "grad_norm": 0.13478477532193128,
+      "learning_rate": 1.30300641405523e-07,
+      "loss": 1.8578,
+      "step": 7950
+    },
+    {
+      "epoch": 0.9859329491231331,
+      "grad_norm": 0.13734004121483884,
+      "learning_rate": 1.194943170213403e-07,
+      "loss": 1.8367,
+      "step": 7955
+    },
+    {
+      "epoch": 0.9865526429943607,
+      "grad_norm": 0.13266755629771385,
+      "learning_rate": 1.0915543432733488e-07,
+      "loss": 1.8426,
+      "step": 7960
+    },
+    {
+      "epoch": 0.9871723368655884,
+      "grad_norm": 0.13522525243545225,
+      "learning_rate": 9.928404170959082e-08,
+      "loss": 1.8188,
+      "step": 7965
+    },
+    {
+      "epoch": 0.987792030736816,
+      "grad_norm": 0.13802548293591557,
+      "learning_rate": 8.988018536630893e-08,
+      "loss": 1.8395,
+      "step": 7970
+    },
+    {
+      "epoch": 0.9884117246080436,
+      "grad_norm": 0.1506854239518021,
+      "learning_rate": 8.094390930762919e-08,
+      "loss": 1.7988,
+      "step": 7975
+    },
+    {
+      "epoch": 0.9890314184792712,
+      "grad_norm": 0.13219180433160424,
+      "learning_rate": 7.247525535538647e-08,
+      "loss": 1.7918,
+      "step": 7980
+    },
+    {
+      "epoch": 0.9896511123504989,
+      "grad_norm": 0.14248580603180083,
+      "learning_rate": 6.447426314297734e-08,
+      "loss": 1.832,
+      "step": 7985
+    },
+    {
+      "epoch": 0.9902708062217265,
+      "grad_norm": 0.1332276335475975,
+      "learning_rate": 5.6940970115115785e-08,
+      "loss": 1.8535,
+      "step": 7990
+    },
+    {
+      "epoch": 0.9908905000929541,
+      "grad_norm": 0.13540820808598406,
+      "learning_rate": 4.9875411527677826e-08,
+      "loss": 1.7641,
+      "step": 7995
+    },
+    {
+      "epoch": 0.9915101939641817,
+      "grad_norm": 0.13658683750615433,
+      "learning_rate": 4.327762044755712e-08,
+      "loss": 1.8637,
+      "step": 8000
+    },
+    {
+      "epoch": 0.9921298878354093,
+      "grad_norm": 0.13582846490136452,
+      "learning_rate": 3.714762775245406e-08,
+      "loss": 1.7602,
+      "step": 8005
+    },
+    {
+      "epoch": 0.992749581706637,
+      "grad_norm": 0.13741100922644967,
+      "learning_rate": 3.148546213080916e-08,
+      "loss": 1.8711,
+      "step": 8010
+    },
+    {
+      "epoch": 0.9933692755778646,
+      "grad_norm": 0.13336480139796178,
+      "learning_rate": 2.6291150081603212e-08,
+      "loss": 1.8867,
+      "step": 8015
+    },
+    {
+      "epoch": 0.9939889694490921,
+      "grad_norm": 0.14304602235765015,
+      "learning_rate": 2.156471591426845e-08,
+      "loss": 1.8926,
+      "step": 8020
+    },
+    {
+      "epoch": 0.9946086633203197,
+      "grad_norm": 0.13292064745465249,
+      "learning_rate": 1.7306181748566463e-08,
+      "loss": 1.8152,
+      "step": 8025
+    },
+    {
+      "epoch": 0.9952283571915473,
+      "grad_norm": 0.1350953647523571,
+      "learning_rate": 1.351556751445493e-08,
+      "loss": 1.8414,
+      "step": 8030
+    },
+    {
+      "epoch": 0.995848051062775,
+      "grad_norm": 0.14228963587482882,
+      "learning_rate": 1.0192890952054334e-08,
+      "loss": 1.802,
+      "step": 8035
+    },
+    {
+      "epoch": 0.9964677449340026,
+      "grad_norm": 0.14132019863120235,
+      "learning_rate": 7.338167611536939e-09,
+      "loss": 1.7988,
+      "step": 8040
+    },
+    {
+      "epoch": 0.9970874388052302,
+      "grad_norm": 0.13716626276276406,
+      "learning_rate": 4.9514108530157585e-09,
+      "loss": 1.7957,
+      "step": 8045
+    },
+    {
+      "epoch": 0.9977071326764578,
+      "grad_norm": 0.14215685681688475,
+      "learning_rate": 3.0326318465334625e-09,
+      "loss": 1.8348,
+      "step": 8050
+    },
+    {
+      "epoch": 0.9983268265476855,
+      "grad_norm": 0.13642854353006256,
+      "learning_rate": 1.5818395720068602e-09,
+      "loss": 1.7578,
+      "step": 8055
+    },
+    {
+      "epoch": 0.9989465204189131,
+      "grad_norm": 0.1342584967064569,
+      "learning_rate": 5.99040819149188e-10,
+      "loss": 1.8578,
+      "step": 8060
+    },
+    {
+      "epoch": 0.9995662142901407,
+      "grad_norm": 0.13380846893093876,
+      "learning_rate": 8.42401874701082e-11,
+      "loss": 1.8281,
+      "step": 8065
+    },
+    {
+      "epoch": 0.9999380306128772,
+      "eval_loss": 1.7868006229400635,
+      "eval_runtime": 207.8873,
+      "eval_samples_per_second": 68.686,
+      "eval_steps_per_second": 4.296,
+      "step": 8068
+    },
+    {
+      "epoch": 0.9999380306128772,
+      "step": 8068,
+      "total_flos": 4248917998829568.0,
+      "train_loss": 1.863956661347298,
+      "train_runtime": 6909.389,
+      "train_samples_per_second": 18.684,
+      "train_steps_per_second": 1.168
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 8068,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4248917998829568.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}