diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14797 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 24.0,
+  "eval_steps": 500,
+  "global_step": 21096,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01138627953316254,
+      "grad_norm": 2.734767198562622,
+      "learning_rate": 0.0001999512432959532,
+      "loss": 30.333529663085937,
+      "step": 10
+    },
+    {
+      "epoch": 0.02277255906632508,
+      "grad_norm": 2.7044386863708496,
+      "learning_rate": 0.00019989706918034563,
+      "loss": 25.792864990234374,
+      "step": 20
+    },
+    {
+      "epoch": 0.034158838599487616,
+      "grad_norm": 2.4714598655700684,
+      "learning_rate": 0.00019984289506473808,
+      "loss": 24.416429138183595,
+      "step": 30
+    },
+    {
+      "epoch": 0.04554511813265016,
+      "grad_norm": 2.6829721927642822,
+      "learning_rate": 0.00019978872094913053,
+      "loss": 23.82067413330078,
+      "step": 40
+    },
+    {
+      "epoch": 0.0569313976658127,
+      "grad_norm": 3.0441884994506836,
+      "learning_rate": 0.00019973454683352296,
+      "loss": 23.323060607910158,
+      "step": 50
+    },
+    {
+      "epoch": 0.06831767719897523,
+      "grad_norm": 3.367356538772583,
+      "learning_rate": 0.0001996803727179154,
+      "loss": 22.732952880859376,
+      "step": 60
+    },
+    {
+      "epoch": 0.07970395673213777,
+      "grad_norm": 3.5883660316467285,
+      "learning_rate": 0.00019962619860230783,
+      "loss": 22.522509765625,
+      "step": 70
+    },
+    {
+      "epoch": 0.09109023626530031,
+      "grad_norm": 3.8543548583984375,
+      "learning_rate": 0.00019957202448670026,
+      "loss": 22.270034790039062,
+      "step": 80
+    },
+    {
+      "epoch": 0.10247651579846286,
+      "grad_norm": 4.374216079711914,
+      "learning_rate": 0.0001995178503710927,
+      "loss": 22.013529968261718,
+      "step": 90
+    },
+    {
+      "epoch": 0.1138627953316254,
+      "grad_norm": 4.15579891204834,
+      "learning_rate": 0.00019946367625548513,
+      "loss": 22.13349304199219,
+      "step": 100
+    },
+    {
+      "epoch": 0.12524907486478792,
+      "grad_norm": 4.63075065612793,
+      "learning_rate": 0.00019940950213987756,
+      "loss": 21.13990173339844,
+      "step": 110
+    },
+    {
+      "epoch": 0.13663535439795046,
+      "grad_norm": 5.391815662384033,
+      "learning_rate": 0.00019935532802427,
+      "loss": 21.209239196777343,
+      "step": 120
+    },
+    {
+      "epoch": 0.148021633931113,
+      "grad_norm": 5.1299896240234375,
+      "learning_rate": 0.00019930115390866243,
+      "loss": 21.18792266845703,
+      "step": 130
+    },
+    {
+      "epoch": 0.15940791346427555,
+      "grad_norm": 5.028836727142334,
+      "learning_rate": 0.00019924697979305489,
+      "loss": 20.73717346191406,
+      "step": 140
+    },
+    {
+      "epoch": 0.1707941929974381,
+      "grad_norm": 5.911370754241943,
+      "learning_rate": 0.00019919280567744734,
+      "loss": 20.920809936523437,
+      "step": 150
+    },
+    {
+      "epoch": 0.18218047253060063,
+      "grad_norm": 6.1007914543151855,
+      "learning_rate": 0.00019913863156183976,
+      "loss": 20.673001098632813,
+      "step": 160
+    },
+    {
+      "epoch": 0.19356675206376317,
+      "grad_norm": 7.011501312255859,
+      "learning_rate": 0.0001990844574462322,
+      "loss": 20.6161376953125,
+      "step": 170
+    },
+    {
+      "epoch": 0.2049530315969257,
+      "grad_norm": 6.00868558883667,
+      "learning_rate": 0.00019903028333062464,
+      "loss": 20.028971862792968,
+      "step": 180
+    },
+    {
+      "epoch": 0.21633931113008825,
+      "grad_norm": 6.089074611663818,
+      "learning_rate": 0.0001989761092150171,
+      "loss": 20.252516174316405,
+      "step": 190
+    },
+    {
+      "epoch": 0.2277255906632508,
+      "grad_norm": 6.774249076843262,
+      "learning_rate": 0.00019892193509940951,
+      "loss": 20.02190399169922,
+      "step": 200
+    },
+    {
+      "epoch": 0.23911187019641333,
+      "grad_norm": 5.999351501464844,
+      "learning_rate": 0.00019886776098380197,
+      "loss": 19.793946838378908,
+      "step": 210
+    },
+    {
+      "epoch": 0.25049814972957585,
+      "grad_norm": 6.9030256271362305,
+      "learning_rate": 0.0001988135868681944,
+      "loss": 19.749664306640625,
+      "step": 220
+    },
+    {
+      "epoch": 0.2618844292627384,
+      "grad_norm": 6.463044166564941,
+      "learning_rate": 0.00019875941275258681,
+      "loss": 19.66212615966797,
+      "step": 230
+    },
+    {
+      "epoch": 0.27327070879590093,
+      "grad_norm": 6.855940818786621,
+      "learning_rate": 0.00019870523863697924,
+      "loss": 19.732081604003906,
+      "step": 240
+    },
+    {
+      "epoch": 0.2846569883290635,
+      "grad_norm": 5.966732025146484,
+      "learning_rate": 0.0001986510645213717,
+      "loss": 19.346263122558593,
+      "step": 250
+    },
+    {
+      "epoch": 0.296043267862226,
+      "grad_norm": 7.0768632888793945,
+      "learning_rate": 0.00019859689040576412,
+      "loss": 19.514622497558594,
+      "step": 260
+    },
+    {
+      "epoch": 0.3074295473953886,
+      "grad_norm": 6.5192694664001465,
+      "learning_rate": 0.00019854271629015657,
+      "loss": 19.53545379638672,
+      "step": 270
+    },
+    {
+      "epoch": 0.3188158269285511,
+      "grad_norm": 6.5199103355407715,
+      "learning_rate": 0.00019848854217454902,
+      "loss": 18.95964813232422,
+      "step": 280
+    },
+    {
+      "epoch": 0.33020210646171366,
+      "grad_norm": 6.990855693817139,
+      "learning_rate": 0.00019843436805894144,
+      "loss": 18.899111938476562,
+      "step": 290
+    },
+    {
+      "epoch": 0.3415883859948762,
+      "grad_norm": 7.77662992477417,
+      "learning_rate": 0.0001983801939433339,
+      "loss": 18.688560485839844,
+      "step": 300
+    },
+    {
+      "epoch": 0.3529746655280387,
+      "grad_norm": 7.311860084533691,
+      "learning_rate": 0.00019832601982772632,
+      "loss": 18.872056579589845,
+      "step": 310
+    },
+    {
+      "epoch": 0.36436094506120126,
+      "grad_norm": 8.168448448181152,
+      "learning_rate": 0.00019827184571211877,
+      "loss": 18.846502685546874,
+      "step": 320
+    },
+    {
+      "epoch": 0.37574722459436377,
+      "grad_norm": 7.2531046867370605,
+      "learning_rate": 0.0001982176715965112,
+      "loss": 18.281404113769533,
+      "step": 330
+    },
+    {
+      "epoch": 0.38713350412752634,
+      "grad_norm": 7.322739601135254,
+      "learning_rate": 0.00019816349748090365,
+      "loss": 18.37748565673828,
+      "step": 340
+    },
+    {
+      "epoch": 0.39851978366068885,
+      "grad_norm": 7.208354949951172,
+      "learning_rate": 0.00019810932336529607,
+      "loss": 18.53016815185547,
+      "step": 350
+    },
+    {
+      "epoch": 0.4099060631938514,
+      "grad_norm": 8.217365264892578,
+      "learning_rate": 0.00019805514924968852,
+      "loss": 18.07104949951172,
+      "step": 360
+    },
+    {
+      "epoch": 0.42129234272701394,
+      "grad_norm": 6.938902854919434,
+      "learning_rate": 0.00019800097513408095,
+      "loss": 18.317633056640624,
+      "step": 370
+    },
+    {
+      "epoch": 0.4326786222601765,
+      "grad_norm": 7.835426330566406,
+      "learning_rate": 0.00019794680101847337,
+      "loss": 18.164633178710936,
+      "step": 380
+    },
+    {
+      "epoch": 0.444064901793339,
+      "grad_norm": 6.851658821105957,
+      "learning_rate": 0.00019789262690286582,
+      "loss": 18.049453735351562,
+      "step": 390
+    },
+    {
+      "epoch": 0.4554511813265016,
+      "grad_norm": 6.942612648010254,
+      "learning_rate": 0.00019783845278725825,
+      "loss": 18.057662963867188,
+      "step": 400
+    },
+    {
+      "epoch": 0.4668374608596641,
+      "grad_norm": 7.211177349090576,
+      "learning_rate": 0.0001977842786716507,
+      "loss": 17.956227111816407,
+      "step": 410
+    },
+    {
+      "epoch": 0.47822374039282667,
+      "grad_norm": 7.398573398590088,
+      "learning_rate": 0.00019773010455604312,
+      "loss": 17.89140167236328,
+      "step": 420
+    },
+    {
+      "epoch": 0.4896100199259892,
+      "grad_norm": 7.789492607116699,
+      "learning_rate": 0.00019767593044043558,
+      "loss": 17.603182983398437,
+      "step": 430
+    },
+    {
+      "epoch": 0.5009962994591517,
+      "grad_norm": 6.86952018737793,
+      "learning_rate": 0.000197621756324828,
+      "loss": 17.51262969970703,
+      "step": 440
+    },
+    {
+      "epoch": 0.5123825789923142,
+      "grad_norm": 6.798729419708252,
+      "learning_rate": 0.00019756758220922045,
+      "loss": 17.721162414550783,
+      "step": 450
+    },
+    {
+      "epoch": 0.5237688585254768,
+      "grad_norm": 7.071723461151123,
+      "learning_rate": 0.00019751340809361288,
+      "loss": 17.830340576171874,
+      "step": 460
+    },
+    {
+      "epoch": 0.5351551380586393,
+      "grad_norm": 7.081210136413574,
+      "learning_rate": 0.00019745923397800533,
+      "loss": 17.28885498046875,
+      "step": 470
+    },
+    {
+      "epoch": 0.5465414175918019,
+      "grad_norm": 7.516231536865234,
+      "learning_rate": 0.00019740505986239775,
+      "loss": 17.6906005859375,
+      "step": 480
+    },
+    {
+      "epoch": 0.5579276971249644,
+      "grad_norm": 6.515166759490967,
+      "learning_rate": 0.0001973508857467902,
+      "loss": 17.18426971435547,
+      "step": 490
+    },
+    {
+      "epoch": 0.569313976658127,
+      "grad_norm": 7.4100189208984375,
+      "learning_rate": 0.00019729671163118263,
+      "loss": 17.34864501953125,
+      "step": 500
+    },
+    {
+      "epoch": 0.5807002561912895,
+      "grad_norm": 6.820520401000977,
+      "learning_rate": 0.00019724253751557505,
+      "loss": 17.613563537597656,
+      "step": 510
+    },
+    {
+      "epoch": 0.592086535724452,
+      "grad_norm": 7.629634857177734,
+      "learning_rate": 0.0001971883633999675,
+      "loss": 17.18858947753906,
+      "step": 520
+    },
+    {
+      "epoch": 0.6034728152576145,
+      "grad_norm": 7.522180080413818,
+      "learning_rate": 0.00019713418928435993,
+      "loss": 17.541119384765626,
+      "step": 530
+    },
+    {
+      "epoch": 0.6148590947907772,
+      "grad_norm": 7.64506196975708,
+      "learning_rate": 0.00019708001516875238,
+      "loss": 17.047702026367187,
+      "step": 540
+    },
+    {
+      "epoch": 0.6262453743239397,
+      "grad_norm": 6.857038497924805,
+      "learning_rate": 0.0001970258410531448,
+      "loss": 17.394444274902344,
+      "step": 550
+    },
+    {
+      "epoch": 0.6376316538571022,
+      "grad_norm": 6.667849063873291,
+      "learning_rate": 0.00019697166693753726,
+      "loss": 17.071144104003906,
+      "step": 560
+    },
+    {
+      "epoch": 0.6490179333902647,
+      "grad_norm": 7.24165153503418,
+      "learning_rate": 0.00019691749282192968,
+      "loss": 16.311009216308594,
+      "step": 570
+    },
+    {
+      "epoch": 0.6604042129234273,
+      "grad_norm": 7.296228408813477,
+      "learning_rate": 0.00019686331870632213,
+      "loss": 16.9224609375,
+      "step": 580
+    },
+    {
+      "epoch": 0.6717904924565898,
+      "grad_norm": 6.7600016593933105,
+      "learning_rate": 0.00019680914459071456,
+      "loss": 17.025440979003907,
+      "step": 590
+    },
+    {
+      "epoch": 0.6831767719897524,
+      "grad_norm": 7.559263229370117,
+      "learning_rate": 0.000196754970475107,
+      "loss": 16.395526123046874,
+      "step": 600
+    },
+    {
+      "epoch": 0.6945630515229149,
+      "grad_norm": 7.639865398406982,
+      "learning_rate": 0.00019670079635949946,
+      "loss": 17.01804962158203,
+      "step": 610
+    },
+    {
+      "epoch": 0.7059493310560774,
+      "grad_norm": 7.71447229385376,
+      "learning_rate": 0.00019664662224389188,
+      "loss": 16.441880798339845,
+      "step": 620
+    },
+    {
+      "epoch": 0.71733561058924,
+      "grad_norm": 6.845354080200195,
+      "learning_rate": 0.00019659244812828434,
+      "loss": 16.84131622314453,
+      "step": 630
+    },
+    {
+      "epoch": 0.7287218901224025,
+      "grad_norm": 6.601531505584717,
+      "learning_rate": 0.00019653827401267676,
+      "loss": 16.1950439453125,
+      "step": 640
+    },
+    {
+      "epoch": 0.740108169655565,
+      "grad_norm": 6.828949928283691,
+      "learning_rate": 0.00019648409989706918,
+      "loss": 16.982223510742188,
+      "step": 650
+    },
+    {
+      "epoch": 0.7514944491887275,
+      "grad_norm": 7.29005765914917,
+      "learning_rate": 0.0001964299257814616,
+      "loss": 16.249069213867188,
+      "step": 660
+    },
+    {
+      "epoch": 0.7628807287218902,
+      "grad_norm": 7.450255870819092,
+      "learning_rate": 0.00019637575166585406,
+      "loss": 16.10637664794922,
+      "step": 670
+    },
+    {
+      "epoch": 0.7742670082550527,
+      "grad_norm": 7.040278434753418,
+      "learning_rate": 0.00019632157755024649,
+      "loss": 16.148355102539064,
+      "step": 680
+    },
+    {
+      "epoch": 0.7856532877882152,
+      "grad_norm": 7.316576957702637,
+      "learning_rate": 0.00019626740343463894,
+      "loss": 16.328166198730468,
+      "step": 690
+    },
+    {
+      "epoch": 0.7970395673213777,
+      "grad_norm": 6.502153396606445,
+      "learning_rate": 0.00019621322931903136,
+      "loss": 15.988864135742187,
+      "step": 700
+    },
+    {
+      "epoch": 0.8084258468545403,
+      "grad_norm": 7.165337085723877,
+      "learning_rate": 0.0001961590552034238,
+      "loss": 16.477142333984375,
+      "step": 710
+    },
+    {
+      "epoch": 0.8198121263877028,
+      "grad_norm": 7.06484317779541,
+      "learning_rate": 0.00019610488108781626,
+      "loss": 16.042941284179687,
+      "step": 720
+    },
+    {
+      "epoch": 0.8311984059208654,
+      "grad_norm": 7.751636505126953,
+      "learning_rate": 0.0001960507069722087,
+      "loss": 16.278134155273438,
+      "step": 730
+    },
+    {
+      "epoch": 0.8425846854540279,
+      "grad_norm": 6.873840808868408,
+      "learning_rate": 0.00019599653285660114,
+      "loss": 15.732162475585938,
+      "step": 740
+    },
+    {
+      "epoch": 0.8539709649871904,
+      "grad_norm": 6.263609886169434,
+      "learning_rate": 0.00019594235874099356,
+      "loss": 16.18165283203125,
+      "step": 750
+    },
+    {
+      "epoch": 0.865357244520353,
+      "grad_norm": 6.69379186630249,
+      "learning_rate": 0.00019588818462538602,
+      "loss": 16.150958251953124,
+      "step": 760
+    },
+    {
+      "epoch": 0.8767435240535155,
+      "grad_norm": 6.563896656036377,
+      "learning_rate": 0.00019583401050977844,
+      "loss": 15.676901245117188,
+      "step": 770
+    },
+    {
+      "epoch": 0.888129803586678,
+      "grad_norm": 6.714715003967285,
+      "learning_rate": 0.0001957798363941709,
+      "loss": 15.763505554199218,
+      "step": 780
+    },
+    {
+      "epoch": 0.8995160831198405,
+      "grad_norm": 6.818701267242432,
+      "learning_rate": 0.00019572566227856332,
+      "loss": 16.16502685546875,
+      "step": 790
+    },
+    {
+      "epoch": 0.9109023626530032,
+      "grad_norm": 7.140949249267578,
+      "learning_rate": 0.00019567148816295574,
+      "loss": 15.515142822265625,
+      "step": 800
+    },
+    {
+      "epoch": 0.9222886421861657,
+      "grad_norm": 6.7389020919799805,
+      "learning_rate": 0.00019561731404734817,
+      "loss": 15.634983825683594,
+      "step": 810
+    },
+    {
+      "epoch": 0.9336749217193282,
+      "grad_norm": 6.847448825836182,
+      "learning_rate": 0.00019556313993174062,
+      "loss": 15.689366149902344,
+      "step": 820
+    },
+    {
+      "epoch": 0.9450612012524907,
+      "grad_norm": 7.26194429397583,
+      "learning_rate": 0.00019550896581613307,
+      "loss": 15.806666564941406,
+      "step": 830
+    },
+    {
+      "epoch": 0.9564474807856533,
+      "grad_norm": 6.119577407836914,
+      "learning_rate": 0.0001954547917005255,
+      "loss": 15.300843811035156,
+      "step": 840
+    },
+    {
+      "epoch": 0.9678337603188158,
+      "grad_norm": 6.605442047119141,
+      "learning_rate": 0.00019540061758491794,
+      "loss": 15.186727905273438,
+      "step": 850
+    },
+    {
+      "epoch": 0.9792200398519784,
+      "grad_norm": 6.060615062713623,
+      "learning_rate": 0.00019534644346931037,
+      "loss": 15.392152404785156,
+      "step": 860
+    },
+    {
+      "epoch": 0.9906063193851409,
+      "grad_norm": 6.676738262176514,
+      "learning_rate": 0.00019529226935370282,
+      "loss": 15.1415283203125,
+      "step": 870
+    },
+    {
+      "epoch": 1.0011386279533163,
+      "grad_norm": 6.309657096862793,
+      "learning_rate": 0.00019523809523809525,
+      "loss": 14.388438415527343,
+      "step": 880
+    },
+    {
+      "epoch": 1.0125249074864788,
+      "grad_norm": 6.854212760925293,
+      "learning_rate": 0.0001951839211224877,
+      "loss": 14.986418151855469,
+      "step": 890
+    },
+    {
+      "epoch": 1.0239111870196413,
+      "grad_norm": 6.602004051208496,
+      "learning_rate": 0.00019512974700688012,
+      "loss": 14.524969482421875,
+      "step": 900
+    },
+    {
+      "epoch": 1.0352974665528039,
+      "grad_norm": 7.170373439788818,
+      "learning_rate": 0.00019507557289127257,
+      "loss": 15.004403686523437,
+      "step": 910
+    },
+    {
+      "epoch": 1.0466837460859664,
+      "grad_norm": 6.895066738128662,
+      "learning_rate": 0.000195021398775665,
+      "loss": 14.763569641113282,
+      "step": 920
+    },
+    {
+      "epoch": 1.0580700256191289,
+      "grad_norm": 6.917144775390625,
+      "learning_rate": 0.00019496722466005742,
+      "loss": 14.740924072265624,
+      "step": 930
+    },
+    {
+      "epoch": 1.0694563051522914,
+      "grad_norm": 6.31508731842041,
+      "learning_rate": 0.00019491305054444987,
+      "loss": 14.591043090820312,
+      "step": 940
+    },
+    {
+      "epoch": 1.080842584685454,
+      "grad_norm": 6.562560558319092,
+      "learning_rate": 0.0001948588764288423,
+      "loss": 14.91121826171875,
+      "step": 950
+    },
+    {
+      "epoch": 1.0922288642186166,
+      "grad_norm": 6.635254383087158,
+      "learning_rate": 0.00019480470231323475,
+      "loss": 15.324360656738282,
+      "step": 960
+    },
+    {
+      "epoch": 1.1036151437517792,
+      "grad_norm": 6.704561233520508,
+      "learning_rate": 0.00019475052819762717,
+      "loss": 15.346974182128907,
+      "step": 970
+    },
+    {
+      "epoch": 1.1150014232849417,
+      "grad_norm": 6.2377190589904785,
+      "learning_rate": 0.00019469635408201963,
+      "loss": 15.083432006835938,
+      "step": 980
+    },
+    {
+      "epoch": 1.1263877028181042,
+      "grad_norm": 6.517440319061279,
+      "learning_rate": 0.00019464217996641205,
+      "loss": 14.828807067871093,
+      "step": 990
+    },
+    {
+      "epoch": 1.1377739823512667,
+      "grad_norm": 6.731694221496582,
+      "learning_rate": 0.0001945880058508045,
+      "loss": 14.4470703125,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1491602618844292,
+      "grad_norm": 6.918997287750244,
+      "learning_rate": 0.00019453383173519693,
+      "loss": 14.706845092773438,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1605465414175917,
+      "grad_norm": 6.228609561920166,
+      "learning_rate": 0.00019447965761958938,
+      "loss": 14.568064880371093,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1719328209507545,
+      "grad_norm": 7.079123497009277,
+      "learning_rate": 0.0001944254835039818,
+      "loss": 14.400852966308594,
+      "step": 1030
+    },
+    {
+      "epoch": 1.183319100483917,
+      "grad_norm": 6.567544937133789,
+      "learning_rate": 0.00019437130938837425,
+      "loss": 14.280006408691406,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1947053800170795,
+      "grad_norm": 7.218766689300537,
+      "learning_rate": 0.00019431713527276668,
+      "loss": 14.390548706054688,
+      "step": 1050
+    },
+    {
+      "epoch": 1.206091659550242,
+      "grad_norm": 6.728633880615234,
+      "learning_rate": 0.00019426296115715913,
+      "loss": 14.768504333496093,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2174779390834045,
+      "grad_norm": 6.698533535003662,
+      "learning_rate": 0.00019420878704155155,
+      "loss": 14.518492126464844,
+      "step": 1070
+    },
+    {
+      "epoch": 1.228864218616567,
+      "grad_norm": 6.782904148101807,
+      "learning_rate": 0.00019415461292594398,
+      "loss": 14.52556915283203,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2402504981497295,
+      "grad_norm": 6.522706508636475,
+      "learning_rate": 0.00019410043881033643,
+      "loss": 14.64172821044922,
+      "step": 1090
+    },
+    {
+      "epoch": 1.251636777682892,
+      "grad_norm": 6.942655563354492,
+      "learning_rate": 0.00019404626469472885,
+      "loss": 14.770219421386718,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2630230572160546,
+      "grad_norm": 6.034182071685791,
+      "learning_rate": 0.0001939920905791213,
+      "loss": 14.227700805664062,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2744093367492173,
+      "grad_norm": 6.776942253112793,
+      "learning_rate": 0.00019393791646351373,
+      "loss": 14.423974609375,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2857956162823796,
+      "grad_norm": 6.075601100921631,
+      "learning_rate": 0.00019388374234790618,
+      "loss": 14.799586486816406,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2971818958155423,
+      "grad_norm": 6.346046447753906,
+      "learning_rate": 0.0001938295682322986,
+      "loss": 14.888410949707032,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3085681753487048,
+      "grad_norm": 6.099461078643799,
+      "learning_rate": 0.00019377539411669106,
+      "loss": 14.434004211425782,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3199544548818674,
+      "grad_norm": 6.728332042694092,
+      "learning_rate": 0.00019372122000108348,
+      "loss": 14.108644104003906,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3313407344150299,
+      "grad_norm": 6.106889247894287,
+      "learning_rate": 0.00019366704588547593,
+      "loss": 13.984442138671875,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3427270139481924,
+      "grad_norm": 6.419578552246094,
+      "learning_rate": 0.00019361287176986839,
+      "loss": 14.819793701171875,
+      "step": 1180
+    },
+    {
+      "epoch": 1.354113293481355,
+      "grad_norm": 6.706035614013672,
+      "learning_rate": 0.0001935586976542608,
+      "loss": 14.234246826171875,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3654995730145174,
+      "grad_norm": 5.745364665985107,
+      "learning_rate": 0.00019350452353865324,
+      "loss": 14.144187927246094,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3768858525476801,
+      "grad_norm": 6.465015411376953,
+      "learning_rate": 0.0001934503494230457,
+      "loss": 14.569776916503907,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3882721320808427,
+      "grad_norm": 5.789831638336182,
+      "learning_rate": 0.0001933961753074381,
+      "loss": 14.204434204101563,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3996584116140052,
+      "grad_norm": 6.589231014251709,
+      "learning_rate": 0.00019334200119183054,
+      "loss": 13.60784912109375,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4110446911471677,
+      "grad_norm": 6.706899166107178,
+      "learning_rate": 0.000193287827076223,
+      "loss": 14.148931884765625,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4224309706803302,
+      "grad_norm": 6.473396301269531,
+      "learning_rate": 0.0001932336529606154,
+      "loss": 14.05177001953125,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4338172502134927,
+      "grad_norm": 6.002316474914551,
+      "learning_rate": 0.00019317947884500786,
+      "loss": 13.77730712890625,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4452035297466552,
+      "grad_norm": 5.781302452087402,
+      "learning_rate": 0.0001931253047294003,
+      "loss": 14.126388549804688,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4565898092798177,
+      "grad_norm": 6.653365612030029,
+      "learning_rate": 0.00019307113061379274,
+      "loss": 14.278581237792968,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4679760888129803,
+      "grad_norm": 6.22074556350708,
+      "learning_rate": 0.0001930169564981852,
+      "loss": 13.808645629882813,
+      "step": 1290
+    },
+    {
+      "epoch": 1.479362368346143,
+      "grad_norm": 6.5593647956848145,
+      "learning_rate": 0.00019296278238257762,
+      "loss": 13.794139099121093,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4907486478793055,
+      "grad_norm": 5.953279972076416,
+      "learning_rate": 0.00019290860826697007,
+      "loss": 13.860592651367188,
+      "step": 1310
+    },
+    {
+      "epoch": 1.502134927412468,
+      "grad_norm": 5.616598606109619,
+      "learning_rate": 0.0001928544341513625,
+      "loss": 13.759190368652344,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5135212069456305,
+      "grad_norm": 5.759921073913574,
+      "learning_rate": 0.00019280026003575494,
+      "loss": 13.61251220703125,
+      "step": 1330
+    },
+    {
+      "epoch": 1.524907486478793,
+      "grad_norm": 5.853238105773926,
+      "learning_rate": 0.00019274608592014737,
+      "loss": 13.766014099121094,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5362937660119556,
+      "grad_norm": 6.38026762008667,
+      "learning_rate": 0.0001926919118045398,
+      "loss": 13.8654296875,
+      "step": 1350
+    },
+    {
+      "epoch": 1.547680045545118,
+      "grad_norm": 6.176024436950684,
+      "learning_rate": 0.00019263773768893222,
+      "loss": 13.856707763671874,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5590663250782808,
+      "grad_norm": 5.767195701599121,
+      "learning_rate": 0.00019258356357332467,
+      "loss": 13.602825927734376,
+      "step": 1370
+    },
+    {
+      "epoch": 1.570452604611443,
+      "grad_norm": 6.558640480041504,
+      "learning_rate": 0.0001925293894577171,
+      "loss": 13.450588989257813,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5818388841446058,
+      "grad_norm": 5.716969966888428,
+      "learning_rate": 0.00019247521534210954,
+      "loss": 13.9936279296875,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5932251636777683,
+      "grad_norm": 6.217525005340576,
+      "learning_rate": 0.000192421041226502,
+      "loss": 13.740388488769531,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6046114432109309,
+      "grad_norm": 6.646007061004639,
+      "learning_rate": 0.00019236686711089442,
+      "loss": 13.809278869628907,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6159977227440934,
+      "grad_norm": 5.91151237487793,
+      "learning_rate": 0.00019231269299528687,
+      "loss": 13.775947570800781,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6273840022772559,
+      "grad_norm": 5.8763957023620605,
+      "learning_rate": 0.0001922585188796793,
+      "loss": 13.332572937011719,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6387702818104184,
+      "grad_norm": 5.781065940856934,
+      "learning_rate": 0.00019220434476407175,
+      "loss": 13.051596069335938,
+      "step": 1440
+    },
+    {
+      "epoch": 1.650156561343581,
+      "grad_norm": 5.9461445808410645,
+      "learning_rate": 0.00019215017064846417,
+      "loss": 13.740011596679688,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6615428408767436,
+      "grad_norm": 5.9059224128723145,
+      "learning_rate": 0.00019209599653285662,
+      "loss": 13.289398193359375,
+      "step": 1460
+    },
+    {
+      "epoch": 1.672929120409906,
+      "grad_norm": 6.076962947845459,
+      "learning_rate": 0.00019204182241724905,
+      "loss": 13.65223388671875,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6843153999430687,
+      "grad_norm": 6.304300308227539,
+      "learning_rate": 0.0001919876483016415,
+      "loss": 13.400830078125,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6957016794762312,
+      "grad_norm": 6.031702041625977,
+      "learning_rate": 0.00019193347418603392,
+      "loss": 13.493438720703125,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7070879590093937,
+      "grad_norm": 6.418713092803955,
+      "learning_rate": 0.00019187930007042635,
+      "loss": 13.696240234375,
+      "step": 1500
+    },
+    {
+      "epoch": 1.7184742385425562,
+      "grad_norm": 5.954122543334961,
+      "learning_rate": 0.0001918251259548188,
+      "loss": 13.905665588378906,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7298605180757187,
+      "grad_norm": 6.412307262420654,
+      "learning_rate": 0.00019177095183921122,
+      "loss": 13.017045593261718,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7412467976088815,
+      "grad_norm": 6.024705410003662,
+      "learning_rate": 0.00019171677772360368,
+      "loss": 13.734344482421875,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7526330771420437,
+      "grad_norm": 6.235986709594727,
+      "learning_rate": 0.0001916626036079961,
+      "loss": 13.35974578857422,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7640193566752065,
+      "grad_norm": 6.963014602661133,
+      "learning_rate": 0.00019160842949238855,
+      "loss": 13.176609802246094,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7754056362083688,
+      "grad_norm": 5.929773807525635,
+      "learning_rate": 0.00019155425537678098,
+      "loss": 13.372627258300781,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7867919157415315,
+      "grad_norm": 6.559890270233154,
+      "learning_rate": 0.00019150008126117343,
+      "loss": 13.397409057617187,
+      "step": 1570
+    },
+    {
+      "epoch": 1.798178195274694,
+      "grad_norm": 6.073424816131592,
+      "learning_rate": 0.00019144590714556585,
+      "loss": 13.51752166748047,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8095644748078565,
+      "grad_norm": 6.313562393188477,
+      "learning_rate": 0.0001913917330299583,
+      "loss": 13.768846130371093,
+      "step": 1590
+    },
+    {
+      "epoch": 1.820950754341019,
+      "grad_norm": 6.578205108642578,
+      "learning_rate": 0.00019133755891435073,
+      "loss": 13.342623901367187,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8323370338741816,
+      "grad_norm": 6.062254428863525,
+      "learning_rate": 0.00019128338479874318,
+      "loss": 13.149093627929688,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8437233134073443,
+      "grad_norm": 6.041051864624023,
+      "learning_rate": 0.0001912292106831356,
+      "loss": 12.984222412109375,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8551095929405066,
+      "grad_norm": 6.353360652923584,
+      "learning_rate": 0.00019117503656752803,
+      "loss": 13.421148681640625,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8664958724736693,
+      "grad_norm": 6.004988670349121,
+      "learning_rate": 0.00019112086245192048,
+      "loss": 13.050341796875,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8778821520068316,
+      "grad_norm": 5.7523932456970215,
+      "learning_rate": 0.0001910666883363129,
+      "loss": 12.99239501953125,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8892684315399944,
+      "grad_norm": 6.328149795532227,
+      "learning_rate": 0.00019101251422070536,
+      "loss": 12.906375122070312,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9006547110731569,
+      "grad_norm": 5.817428112030029,
+      "learning_rate": 0.00019095834010509778,
+      "loss": 12.926182556152344,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9120409906063194,
+      "grad_norm": 6.4034247398376465,
+      "learning_rate": 0.00019090416598949023,
+      "loss": 12.87743682861328,
+      "step": 1680
+    },
+    {
+      "epoch": 1.923427270139482,
+      "grad_norm": 5.860933303833008,
+      "learning_rate": 0.00019084999187388266,
+      "loss": 13.04234619140625,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9348135496726444,
+      "grad_norm": 5.991933345794678,
+      "learning_rate": 0.0001907958177582751,
+      "loss": 12.981626892089844,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9461998292058071,
+      "grad_norm": 5.810291767120361,
+      "learning_rate": 0.00019074164364266753,
+      "loss": 13.029983520507812,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9575861087389694,
+      "grad_norm": 6.07528829574585,
+      "learning_rate": 0.00019068746952705999,
+      "loss": 13.375914001464844,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9689723882721322,
+      "grad_norm": 6.209125995635986,
+      "learning_rate": 0.0001906332954114524,
+      "loss": 13.055149841308594,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9803586678052945,
+      "grad_norm": 5.829214572906494,
+      "learning_rate": 0.00019057912129584486,
+      "loss": 13.251846313476562,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9917449473384572,
+      "grad_norm": 5.681432723999023,
+      "learning_rate": 0.0001905249471802373,
+      "loss": 13.179347229003906,
+      "step": 1750
+    },
+    {
+      "epoch": 2.0022772559066326,
+      "grad_norm": 6.714110374450684,
+      "learning_rate": 0.00019047077306462974,
+      "loss": 12.078095245361329,
+      "step": 1760
+    },
+    {
+      "epoch": 2.013663535439795,
+      "grad_norm": 5.950106620788574,
+      "learning_rate": 0.00019041659894902216,
+      "loss": 12.684372711181641,
+      "step": 1770
+    },
+    {
+      "epoch": 2.0250498149729577,
+      "grad_norm": 6.075564861297607,
+      "learning_rate": 0.00019036242483341459,
+      "loss": 12.492652893066406,
+      "step": 1780
+    },
+    {
+      "epoch": 2.03643609450612,
+      "grad_norm": 5.315488338470459,
+      "learning_rate": 0.00019030825071780704,
+      "loss": 12.558822631835938,
+      "step": 1790
+    },
+    {
+      "epoch": 2.0478223740392827,
+      "grad_norm": 6.444157600402832,
+      "learning_rate": 0.00019025407660219946,
+      "loss": 12.013819122314453,
+      "step": 1800
+    },
+    {
+      "epoch": 2.059208653572445,
+      "grad_norm": 7.443634510040283,
+      "learning_rate": 0.00019019990248659191,
+      "loss": 12.401322174072266,
+      "step": 1810
+    },
+    {
+      "epoch": 2.0705949331056077,
+      "grad_norm": 6.039581298828125,
+      "learning_rate": 0.00019014572837098434,
+      "loss": 11.99999542236328,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0819812126387705,
+      "grad_norm": 6.211398601531982,
+      "learning_rate": 0.0001900915542553768,
+      "loss": 12.471622467041016,
+      "step": 1830
+    },
+    {
+      "epoch": 2.0933674921719327,
+      "grad_norm": 5.827359199523926,
+      "learning_rate": 0.00019003738013976921,
+      "loss": 12.675922393798828,
+      "step": 1840
+    },
+    {
+      "epoch": 2.1047537717050955,
+      "grad_norm": 6.171777248382568,
+      "learning_rate": 0.00018998320602416167,
+      "loss": 12.37955093383789,
+      "step": 1850
+    },
+    {
+      "epoch": 2.1161400512382578,
+      "grad_norm": 5.889030456542969,
+      "learning_rate": 0.00018992903190855412,
+      "loss": 12.620635986328125,
+      "step": 1860
+    },
+    {
+      "epoch": 2.1275263307714205,
+      "grad_norm": 5.960679531097412,
+      "learning_rate": 0.00018987485779294654,
+      "loss": 12.005726623535157,
+      "step": 1870
+    },
+    {
+      "epoch": 2.138912610304583,
+      "grad_norm": 6.680525302886963,
+      "learning_rate": 0.000189820683677339,
+      "loss": 12.5527099609375,
+      "step": 1880
+    },
+    {
+      "epoch": 2.1502988898377455,
+      "grad_norm": 6.306058883666992,
+      "learning_rate": 0.00018976650956173142,
+      "loss": 12.299282836914063,
+      "step": 1890
+    },
+    {
+      "epoch": 2.161685169370908,
+      "grad_norm": 6.001065254211426,
+      "learning_rate": 0.00018971233544612387,
+      "loss": 12.561842346191407,
+      "step": 1900
+    },
+    {
+      "epoch": 2.1730714489040706,
+      "grad_norm": 5.667927265167236,
+      "learning_rate": 0.0001896581613305163,
+      "loss": 12.342430877685548,
+      "step": 1910
+    },
+    {
+      "epoch": 2.1844577284372333,
+      "grad_norm": 5.749220848083496,
+      "learning_rate": 0.00018960398721490872,
+      "loss": 11.777164459228516,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1958440079703956,
+      "grad_norm": 5.756120681762695,
+      "learning_rate": 0.00018954981309930114,
+      "loss": 12.499293518066406,
+      "step": 1930
+    },
+    {
+      "epoch": 2.2072302875035583,
+      "grad_norm": 5.476713180541992,
+      "learning_rate": 0.0001894956389836936,
+      "loss": 12.119635009765625,
+      "step": 1940
+    },
+    {
+      "epoch": 2.2186165670367206,
+      "grad_norm": 6.148690700531006,
+      "learning_rate": 0.00018944146486808602,
+      "loss": 12.262448120117188,
+      "step": 1950
+    },
+    {
+      "epoch": 2.2300028465698833,
+      "grad_norm": 6.188674449920654,
+      "learning_rate": 0.00018938729075247847,
+      "loss": 11.836429595947266,
+      "step": 1960
+    },
+    {
+      "epoch": 2.2413891261030456,
+      "grad_norm": 5.806178569793701,
+      "learning_rate": 0.00018933311663687092,
+      "loss": 12.008864593505859,
+      "step": 1970
+    },
+    {
+      "epoch": 2.2527754056362084,
+      "grad_norm": 6.355106830596924,
+      "learning_rate": 0.00018927894252126335,
+      "loss": 12.07802505493164,
+      "step": 1980
+    },
+    {
+      "epoch": 2.2641616851693707,
+      "grad_norm": 5.5258612632751465,
+      "learning_rate": 0.0001892247684056558,
+      "loss": 12.283776092529298,
+      "step": 1990
+    },
+    {
+      "epoch": 2.2755479647025334,
+      "grad_norm": 5.87457275390625,
+      "learning_rate": 0.00018917059429004822,
+      "loss": 12.212881469726563,
+      "step": 2000
+    },
+    {
+      "epoch": 2.286934244235696,
+      "grad_norm": 5.784579277038574,
+      "learning_rate": 0.00018911642017444067,
+      "loss": 12.175121307373047,
+      "step": 2010
+    },
+    {
+      "epoch": 2.2983205237688584,
+      "grad_norm": 5.964444637298584,
+      "learning_rate": 0.00018906766347039387,
+      "loss": 12.410752868652343,
+      "step": 2020
+    },
+    {
+      "epoch": 2.309706803302021,
+      "grad_norm": 5.863014221191406,
+      "learning_rate": 0.00018901890676634707,
+      "loss": 12.259366607666015,
+      "step": 2030
+    },
+    {
+      "epoch": 2.3210930828351835,
+      "grad_norm": 6.4362382888793945,
+      "learning_rate": 0.0001889647326507395,
+      "loss": 12.253427124023437,
+      "step": 2040
+    },
+    {
+      "epoch": 2.332479362368346,
+      "grad_norm": 5.864607810974121,
+      "learning_rate": 0.00018891055853513194,
+      "loss": 12.033533477783203,
+      "step": 2050
+    },
+    {
+      "epoch": 2.343865641901509,
+      "grad_norm": 6.0086822509765625,
+      "learning_rate": 0.00018885638441952437,
+      "loss": 12.2366943359375,
+      "step": 2060
+    },
+    {
+      "epoch": 2.355251921434671,
+      "grad_norm": 5.782958507537842,
+      "learning_rate": 0.00018880221030391682,
+      "loss": 12.448239135742188,
+      "step": 2070
+    },
+    {
+      "epoch": 2.366638200967834,
+      "grad_norm": 5.685044288635254,
+      "learning_rate": 0.00018874803618830924,
+      "loss": 11.95639419555664,
+      "step": 2080
+    },
+    {
+      "epoch": 2.3780244805009962,
+      "grad_norm": 5.853184700012207,
+      "learning_rate": 0.00018869386207270167,
+      "loss": 12.157659912109375,
+      "step": 2090
+    },
+    {
+      "epoch": 2.389410760034159,
+      "grad_norm": 5.814517974853516,
+      "learning_rate": 0.0001886396879570941,
+      "loss": 11.77392349243164,
+      "step": 2100
+    },
+    {
+      "epoch": 2.4007970395673213,
+      "grad_norm": 6.793750286102295,
+      "learning_rate": 0.00018858551384148654,
+      "loss": 12.14357681274414,
+      "step": 2110
+    },
+    {
+      "epoch": 2.412183319100484,
+      "grad_norm": 5.957283020019531,
+      "learning_rate": 0.00018853133972587897,
+      "loss": 11.823822021484375,
+      "step": 2120
+    },
+    {
+      "epoch": 2.4235695986336463,
+      "grad_norm": 6.891757488250732,
+      "learning_rate": 0.00018847716561027142,
+      "loss": 12.402496337890625,
+      "step": 2130
+    },
+    {
+      "epoch": 2.434955878166809,
+      "grad_norm": 5.9682135581970215,
+      "learning_rate": 0.00018842299149466387,
+      "loss": 12.376535797119141,
+      "step": 2140
+    },
+    {
+      "epoch": 2.4463421576999718,
+      "grad_norm": 6.419319152832031,
+      "learning_rate": 0.0001883688173790563,
+      "loss": 11.78542709350586,
+      "step": 2150
+    },
+    {
+      "epoch": 2.457728437233134,
+      "grad_norm": 5.599404335021973,
+      "learning_rate": 0.00018831464326344875,
+      "loss": 11.65877685546875,
+      "step": 2160
+    },
+    {
+      "epoch": 2.469114716766297,
+      "grad_norm": 6.27384090423584,
+      "learning_rate": 0.00018826046914784117,
+      "loss": 12.384040832519531,
+      "step": 2170
+    },
+    {
+      "epoch": 2.480500996299459,
+      "grad_norm": 6.210123062133789,
+      "learning_rate": 0.00018820629503223362,
+      "loss": 11.9489990234375,
+      "step": 2180
+    },
+    {
+      "epoch": 2.491887275832622,
+      "grad_norm": 6.209619998931885,
+      "learning_rate": 0.00018815212091662605,
+      "loss": 12.142713928222657,
+      "step": 2190
+    },
+    {
+      "epoch": 2.503273555365784,
+      "grad_norm": 5.776041030883789,
+      "learning_rate": 0.0001880979468010185,
+      "loss": 12.368475341796875,
+      "step": 2200
+    },
+    {
+      "epoch": 2.514659834898947,
+      "grad_norm": 5.5831217765808105,
+      "learning_rate": 0.00018804377268541092,
+      "loss": 12.049860382080078,
+      "step": 2210
+    },
+    {
+      "epoch": 2.526046114432109,
+      "grad_norm": 6.643979072570801,
+      "learning_rate": 0.00018798959856980335,
+      "loss": 12.57002944946289,
+      "step": 2220
+    },
+    {
+      "epoch": 2.537432393965272,
+      "grad_norm": 5.437014102935791,
+      "learning_rate": 0.00018793542445419577,
+      "loss": 12.165242767333984,
+      "step": 2230
+    },
+    {
+      "epoch": 2.5488186734984346,
+      "grad_norm": 6.001914024353027,
+      "learning_rate": 0.00018788125033858822,
+      "loss": 11.841060638427734,
+      "step": 2240
+    },
+    {
+      "epoch": 2.560204953031597,
+      "grad_norm": 6.4120259284973145,
+      "learning_rate": 0.00018782707622298065,
+      "loss": 11.942637634277343,
+      "step": 2250
+    },
+    {
+      "epoch": 2.571591232564759,
+      "grad_norm": 6.355876922607422,
+      "learning_rate": 0.0001877729021073731,
+      "loss": 11.278862762451173,
+      "step": 2260
+    },
+    {
+      "epoch": 2.582977512097922,
+      "grad_norm": 5.9582600593566895,
+      "learning_rate": 0.00018771872799176555,
+      "loss": 12.05511932373047,
+      "step": 2270
+    },
+    {
+      "epoch": 2.5943637916310847,
+      "grad_norm": 6.194489479064941,
+      "learning_rate": 0.00018766455387615798,
+      "loss": 11.693316650390624,
+      "step": 2280
+    },
+    {
+      "epoch": 2.605750071164247,
+      "grad_norm": 6.129605770111084,
+      "learning_rate": 0.00018761037976055043,
+      "loss": 11.912431335449218,
+      "step": 2290
+    },
+    {
+      "epoch": 2.6171363506974097,
+      "grad_norm": 5.7922444343566895,
+      "learning_rate": 0.00018755620564494285,
+      "loss": 11.996822357177734,
+      "step": 2300
+    },
+    {
+      "epoch": 2.628522630230572,
+      "grad_norm": 5.562745094299316,
+      "learning_rate": 0.0001875020315293353,
+      "loss": 12.01687240600586,
+      "step": 2310
+    },
+    {
+      "epoch": 2.6399089097637347,
+      "grad_norm": 5.735344886779785,
+      "learning_rate": 0.00018744785741372773,
+      "loss": 12.01505355834961,
+      "step": 2320
+    },
+    {
+      "epoch": 2.6512951892968974,
+      "grad_norm": 5.924304485321045,
+      "learning_rate": 0.00018739368329812018,
+      "loss": 11.80968017578125,
+      "step": 2330
+    },
+    {
+      "epoch": 2.6626814688300597,
+      "grad_norm": 5.843868732452393,
+      "learning_rate": 0.0001873395091825126,
+      "loss": 12.081328582763671,
+      "step": 2340
+    },
+    {
+      "epoch": 2.6740677483632225,
+      "grad_norm": 5.781383514404297,
+      "learning_rate": 0.00018728533506690506,
+      "loss": 12.280255889892578,
+      "step": 2350
+    },
+    {
+      "epoch": 2.6854540278963848,
+      "grad_norm": 5.876210689544678,
+      "learning_rate": 0.00018723116095129748,
+      "loss": 11.941732788085938,
+      "step": 2360
+    },
+    {
+      "epoch": 2.6968403074295475,
+      "grad_norm": 5.828925609588623,
+      "learning_rate": 0.0001871769868356899,
+      "loss": 11.539690399169922,
+      "step": 2370
+    },
+    {
+      "epoch": 2.70822658696271,
+      "grad_norm": 6.443966388702393,
+      "learning_rate": 0.00018712281272008236,
+      "loss": 11.438835144042969,
+      "step": 2380
+    },
+    {
+      "epoch": 2.7196128664958725,
+      "grad_norm": 5.669212818145752,
+      "learning_rate": 0.00018706863860447478,
+      "loss": 12.036208343505859,
+      "step": 2390
+    },
+    {
+      "epoch": 2.730999146029035,
+      "grad_norm": 5.616700649261475,
+      "learning_rate": 0.00018701446448886723,
+      "loss": 11.880667114257813,
+      "step": 2400
+    },
+    {
+      "epoch": 2.7423854255621976,
+      "grad_norm": 5.890092372894287,
+      "learning_rate": 0.00018696029037325966,
+      "loss": 11.707859802246094,
+      "step": 2410
+    },
+    {
+      "epoch": 2.7537717050953603,
+      "grad_norm": 5.797362804412842,
+      "learning_rate": 0.0001869061162576521,
+      "loss": 11.643144989013672,
+      "step": 2420
+    },
+    {
+      "epoch": 2.7651579846285226,
+      "grad_norm": 5.619494438171387,
+      "learning_rate": 0.00018685194214204453,
+      "loss": 11.876329803466797,
+      "step": 2430
+    },
+    {
+      "epoch": 2.7765442641616853,
+      "grad_norm": 5.690425395965576,
+      "learning_rate": 0.00018679776802643698,
+      "loss": 11.896712493896484,
+      "step": 2440
+    },
+    {
+      "epoch": 2.7879305436948476,
+      "grad_norm": 5.649746894836426,
+      "learning_rate": 0.0001867435939108294,
+      "loss": 11.640943908691407,
+      "step": 2450
+    },
+    {
+      "epoch": 2.7993168232280103,
+      "grad_norm": 5.736121654510498,
+      "learning_rate": 0.00018668941979522186,
+      "loss": 11.80575942993164,
+      "step": 2460
+    },
+    {
+      "epoch": 2.8107031027611726,
+      "grad_norm": 5.586820125579834,
+      "learning_rate": 0.00018663524567961428,
+      "loss": 11.54848861694336,
+      "step": 2470
+    },
+    {
+      "epoch": 2.8220893822943354,
+      "grad_norm": 5.552623271942139,
+      "learning_rate": 0.00018658107156400674,
+      "loss": 11.699449157714843,
+      "step": 2480
+    },
+    {
+      "epoch": 2.8334756618274977,
+      "grad_norm": 5.449943542480469,
+      "learning_rate": 0.00018652689744839916,
+      "loss": 12.021916961669922,
+      "step": 2490
+    },
+    {
+      "epoch": 2.8448619413606604,
+      "grad_norm": 6.037112712860107,
+      "learning_rate": 0.0001864727233327916,
+      "loss": 11.778411865234375,
+      "step": 2500
+    },
+    {
+      "epoch": 2.856248220893823,
+      "grad_norm": 5.5578813552856445,
+      "learning_rate": 0.00018641854921718404,
+      "loss": 11.259034729003906,
+      "step": 2510
+    },
+    {
+      "epoch": 2.8676345004269854,
+      "grad_norm": 5.7229719161987305,
+      "learning_rate": 0.00018636437510157646,
+      "loss": 11.734458923339844,
+      "step": 2520
+    },
+    {
+      "epoch": 2.879020779960148,
+      "grad_norm": 5.95925235748291,
+      "learning_rate": 0.0001863102009859689,
+      "loss": 11.601738739013673,
+      "step": 2530
+    },
+    {
+      "epoch": 2.8904070594933104,
+      "grad_norm": 5.8496479988098145,
+      "learning_rate": 0.00018625602687036134,
+      "loss": 11.096221923828125,
+      "step": 2540
+    },
+    {
+      "epoch": 2.901793339026473,
+      "grad_norm": 6.491213321685791,
+      "learning_rate": 0.0001862018527547538,
+      "loss": 11.531390380859374,
+      "step": 2550
+    },
+    {
+      "epoch": 2.9131796185596355,
+      "grad_norm": 5.992333889007568,
+      "learning_rate": 0.0001861476786391462,
+      "loss": 11.45031509399414,
+      "step": 2560
+    },
+    {
+      "epoch": 2.924565898092798,
+      "grad_norm": 5.646914005279541,
+      "learning_rate": 0.00018609350452353866,
+      "loss": 11.133451080322265,
+      "step": 2570
+    },
+    {
+      "epoch": 2.9359521776259605,
+      "grad_norm": 5.3638176918029785,
+      "learning_rate": 0.0001860393304079311,
+      "loss": 11.416927337646484,
+      "step": 2580
+    },
+    {
+      "epoch": 2.9473384571591232,
+      "grad_norm": 5.817018985748291,
+      "learning_rate": 0.00018598515629232354,
+      "loss": 11.489157104492188,
+      "step": 2590
+    },
+    {
+      "epoch": 2.958724736692286,
+      "grad_norm": 6.094698905944824,
+      "learning_rate": 0.000185930982176716,
+      "loss": 11.472473907470704,
+      "step": 2600
+    },
+    {
+      "epoch": 2.9701110162254483,
+      "grad_norm": 6.355233669281006,
+      "learning_rate": 0.00018587680806110842,
+      "loss": 11.384600067138672,
+      "step": 2610
+    },
+    {
+      "epoch": 2.981497295758611,
+      "grad_norm": 5.65998649597168,
+      "learning_rate": 0.00018582263394550087,
+      "loss": 11.349325561523438,
+      "step": 2620
+    },
+    {
+      "epoch": 2.9928835752917733,
+      "grad_norm": 6.039316654205322,
+      "learning_rate": 0.0001857684598298933,
+      "loss": 11.65872802734375,
+      "step": 2630
+    },
+    {
+      "epoch": 3.0034158838599487,
+      "grad_norm": 5.6023993492126465,
+      "learning_rate": 0.00018571428571428572,
+      "loss": 10.584882354736328,
+      "step": 2640
+    },
+    {
+      "epoch": 3.0148021633931115,
+      "grad_norm": 6.057605743408203,
+      "learning_rate": 0.00018566011159867814,
+      "loss": 11.056887817382812,
+      "step": 2650
+    },
+    {
+      "epoch": 3.0261884429262738,
+      "grad_norm": 5.620913028717041,
+      "learning_rate": 0.0001856059374830706,
+      "loss": 11.04276351928711,
+      "step": 2660
+    },
+    {
+      "epoch": 3.0375747224594365,
+      "grad_norm": 5.724585056304932,
+      "learning_rate": 0.00018555176336746302,
+      "loss": 10.584162139892578,
+      "step": 2670
+    },
+    {
+      "epoch": 3.048961001992599,
+      "grad_norm": 6.663942337036133,
+      "learning_rate": 0.00018549758925185547,
+      "loss": 10.474478149414063,
+      "step": 2680
+    },
+    {
+      "epoch": 3.0603472815257615,
+      "grad_norm": 5.8548688888549805,
+      "learning_rate": 0.0001854434151362479,
+      "loss": 10.782363891601562,
+      "step": 2690
+    },
+    {
+      "epoch": 3.071733561058924,
+      "grad_norm": 5.4828314781188965,
+      "learning_rate": 0.00018538924102064035,
+      "loss": 10.86609115600586,
+      "step": 2700
+    },
+    {
+      "epoch": 3.0831198405920865,
+      "grad_norm": 5.855891704559326,
+      "learning_rate": 0.0001853350669050328,
+      "loss": 10.718875885009766,
+      "step": 2710
+    },
+    {
+      "epoch": 3.0945061201252493,
+      "grad_norm": 5.583263874053955,
+      "learning_rate": 0.00018528089278942522,
+      "loss": 10.526648712158202,
+      "step": 2720
+    },
+    {
+      "epoch": 3.1058923996584116,
+      "grad_norm": 5.6130690574646,
+      "learning_rate": 0.00018522671867381767,
+      "loss": 10.729558563232422,
+      "step": 2730
+    },
+    {
+      "epoch": 3.1172786791915743,
+      "grad_norm": 6.017326354980469,
+      "learning_rate": 0.0001851725445582101,
+      "loss": 10.696547698974609,
+      "step": 2740
+    },
+    {
+      "epoch": 3.1286649587247366,
+      "grad_norm": 5.918975353240967,
+      "learning_rate": 0.00018511837044260255,
+      "loss": 10.813334655761718,
+      "step": 2750
+    },
+    {
+      "epoch": 3.1400512382578993,
+      "grad_norm": 5.893662929534912,
+      "learning_rate": 0.00018506419632699497,
+      "loss": 10.956719970703125,
+      "step": 2760
+    },
+    {
+      "epoch": 3.1514375177910616,
+      "grad_norm": 6.190542221069336,
+      "learning_rate": 0.00018501002221138743,
+      "loss": 10.913722229003906,
+      "step": 2770
+    },
+    {
+      "epoch": 3.1628237973242244,
+      "grad_norm": 6.2601237297058105,
+      "learning_rate": 0.00018495584809577985,
+      "loss": 10.506605529785157,
+      "step": 2780
+    },
+    {
+      "epoch": 3.1742100768573867,
+      "grad_norm": 5.85802698135376,
+      "learning_rate": 0.00018490167398017227,
+      "loss": 11.01781005859375,
+      "step": 2790
+    },
+    {
+      "epoch": 3.1855963563905494,
+      "grad_norm": 5.59824800491333,
+      "learning_rate": 0.0001848474998645647,
+      "loss": 10.554005432128907,
+      "step": 2800
+    },
+    {
+      "epoch": 3.196982635923712,
+      "grad_norm": 5.9762282371521,
+      "learning_rate": 0.00018479332574895715,
+      "loss": 11.067266082763672,
+      "step": 2810
+    },
+    {
+      "epoch": 3.2083689154568744,
+      "grad_norm": 5.892134189605713,
+      "learning_rate": 0.0001847391516333496,
+      "loss": 10.800299072265625,
+      "step": 2820
+    },
+    {
+      "epoch": 3.219755194990037,
+      "grad_norm": 6.411027431488037,
+      "learning_rate": 0.00018468497751774203,
+      "loss": 10.963347625732421,
+      "step": 2830
+    },
+    {
+      "epoch": 3.2311414745231994,
+      "grad_norm": 5.7685394287109375,
+      "learning_rate": 0.00018463080340213448,
+      "loss": 10.73892822265625,
+      "step": 2840
+    },
+    {
+      "epoch": 3.242527754056362,
+      "grad_norm": 5.961293697357178,
+      "learning_rate": 0.0001845766292865269,
+      "loss": 10.619090270996093,
+      "step": 2850
+    },
+    {
+      "epoch": 3.2539140335895245,
+      "grad_norm": 7.078546524047852,
+      "learning_rate": 0.00018452245517091935,
+      "loss": 10.756937408447266,
+      "step": 2860
+    },
+    {
+      "epoch": 3.265300313122687,
+      "grad_norm": 6.000650405883789,
+      "learning_rate": 0.00018446828105531178,
+      "loss": 11.32503662109375,
+      "step": 2870
+    },
+    {
+      "epoch": 3.2766865926558495,
+      "grad_norm": 5.751965045928955,
+      "learning_rate": 0.00018441410693970423,
+      "loss": 11.059173583984375,
+      "step": 2880
+    },
+    {
+      "epoch": 3.2880728721890122,
+      "grad_norm": 6.229584693908691,
+      "learning_rate": 0.00018435993282409665,
+      "loss": 10.77396469116211,
+      "step": 2890
+    },
+    {
+      "epoch": 3.299459151722175,
+      "grad_norm": 5.8026123046875,
+      "learning_rate": 0.0001843057587084891,
+      "loss": 10.68900146484375,
+      "step": 2900
+    },
+    {
+      "epoch": 3.3108454312553373,
+      "grad_norm": 5.8648529052734375,
+      "learning_rate": 0.00018425158459288153,
+      "loss": 10.619082641601562,
+      "step": 2910
+    },
+    {
+      "epoch": 3.3222317107885,
+      "grad_norm": 6.061291217803955,
+      "learning_rate": 0.00018419741047727395,
+      "loss": 11.001821899414063,
+      "step": 2920
+    },
+    {
+      "epoch": 3.3336179903216623,
+      "grad_norm": 6.154610633850098,
+      "learning_rate": 0.0001841432363616664,
+      "loss": 10.811206817626953,
+      "step": 2930
+    },
+    {
+      "epoch": 3.345004269854825,
+      "grad_norm": 5.7560133934021,
+      "learning_rate": 0.00018408906224605883,
+      "loss": 11.004631805419923,
+      "step": 2940
+    },
+    {
+      "epoch": 3.3563905493879873,
+      "grad_norm": 5.845217227935791,
+      "learning_rate": 0.00018403488813045128,
+      "loss": 10.858744049072266,
+      "step": 2950
+    },
+    {
+      "epoch": 3.36777682892115,
+      "grad_norm": 5.666683197021484,
+      "learning_rate": 0.0001839807140148437,
+      "loss": 11.276985931396485,
+      "step": 2960
+    },
+    {
+      "epoch": 3.3791631084543123,
+      "grad_norm": 6.077807426452637,
+      "learning_rate": 0.00018392653989923616,
+      "loss": 10.676035308837891,
+      "step": 2970
+    },
+    {
+      "epoch": 3.390549387987475,
+      "grad_norm": 5.791286468505859,
+      "learning_rate": 0.00018387236578362858,
+      "loss": 10.966339874267579,
+      "step": 2980
+    },
+    {
+      "epoch": 3.401935667520638,
+      "grad_norm": 6.040390968322754,
+      "learning_rate": 0.00018381819166802103,
+      "loss": 10.893710327148437,
+      "step": 2990
+    },
+    {
+      "epoch": 3.4133219470538,
+      "grad_norm": 5.881897926330566,
+      "learning_rate": 0.00018376401755241346,
+      "loss": 10.969651031494141,
+      "step": 3000
+    },
+    {
+      "epoch": 3.424708226586963,
+      "grad_norm": 6.096680641174316,
+      "learning_rate": 0.0001837098434368059,
+      "loss": 10.903501892089844,
+      "step": 3010
+    },
+    {
+      "epoch": 3.436094506120125,
+      "grad_norm": 6.454861164093018,
+      "learning_rate": 0.00018365566932119834,
+      "loss": 10.728018188476563,
+      "step": 3020
+    },
+    {
+      "epoch": 3.447480785653288,
+      "grad_norm": 5.956232070922852,
+      "learning_rate": 0.0001836014952055908,
+      "loss": 10.590874481201173,
+      "step": 3030
+    },
+    {
+      "epoch": 3.45886706518645,
+      "grad_norm": 5.892461776733398,
+      "learning_rate": 0.0001835473210899832,
+      "loss": 10.788093566894531,
+      "step": 3040
+    },
+    {
+      "epoch": 3.470253344719613,
+      "grad_norm": 5.8851423263549805,
+      "learning_rate": 0.00018349314697437566,
+      "loss": 10.701885223388672,
+      "step": 3050
+    },
+    {
+      "epoch": 3.481639624252775,
+      "grad_norm": 6.0876078605651855,
+      "learning_rate": 0.0001834389728587681,
+      "loss": 10.101341247558594,
+      "step": 3060
+    },
+    {
+      "epoch": 3.493025903785938,
+      "grad_norm": 6.961325168609619,
+      "learning_rate": 0.0001833847987431605,
+      "loss": 10.46026153564453,
+      "step": 3070
+    },
+    {
+      "epoch": 3.5044121833191006,
+      "grad_norm": 6.137760639190674,
+      "learning_rate": 0.00018333062462755296,
+      "loss": 10.805878448486329,
+      "step": 3080
+    },
+    {
+      "epoch": 3.515798462852263,
+      "grad_norm": 5.97675895690918,
+      "learning_rate": 0.0001832764505119454,
+      "loss": 11.289669799804688,
+      "step": 3090
+    },
+    {
+      "epoch": 3.5271847423854257,
+      "grad_norm": 6.564485549926758,
+      "learning_rate": 0.00018322227639633784,
+      "loss": 10.845318603515626,
+      "step": 3100
+    },
+    {
+      "epoch": 3.538571021918588,
+      "grad_norm": 6.3509979248046875,
+      "learning_rate": 0.00018316810228073026,
+      "loss": 10.126212310791015,
+      "step": 3110
+    },
+    {
+      "epoch": 3.5499573014517507,
+      "grad_norm": 5.984918117523193,
+      "learning_rate": 0.00018311392816512272,
+      "loss": 10.31814956665039,
+      "step": 3120
+    },
+    {
+      "epoch": 3.5613435809849134,
+      "grad_norm": 5.558696746826172,
+      "learning_rate": 0.00018305975404951514,
+      "loss": 10.726148223876953,
+      "step": 3130
+    },
+    {
+      "epoch": 3.5727298605180757,
+      "grad_norm": 6.394120216369629,
+      "learning_rate": 0.0001830055799339076,
+      "loss": 10.786263275146485,
+      "step": 3140
+    },
+    {
+      "epoch": 3.584116140051238,
+      "grad_norm": 5.847959041595459,
+      "learning_rate": 0.00018295140581830002,
+      "loss": 10.560782623291015,
+      "step": 3150
+    },
+    {
+      "epoch": 3.5955024195844008,
+      "grad_norm": 5.2607879638671875,
+      "learning_rate": 0.00018289723170269247,
+      "loss": 10.637435913085938,
+      "step": 3160
+    },
+    {
+      "epoch": 3.6068886991175635,
+      "grad_norm": 6.327804088592529,
+      "learning_rate": 0.00018284305758708492,
+      "loss": 11.141863250732422,
+      "step": 3170
+    },
+    {
+      "epoch": 3.618274978650726,
+      "grad_norm": 5.784237861633301,
+      "learning_rate": 0.00018278888347147734,
+      "loss": 10.513392639160156,
+      "step": 3180
+    },
+    {
+      "epoch": 3.6296612581838885,
+      "grad_norm": 5.775552272796631,
+      "learning_rate": 0.0001827347093558698,
+      "loss": 10.784326934814453,
+      "step": 3190
+    },
+    {
+      "epoch": 3.641047537717051,
+      "grad_norm": 5.3834075927734375,
+      "learning_rate": 0.00018268053524026222,
+      "loss": 11.03073959350586,
+      "step": 3200
+    },
+    {
+      "epoch": 3.6524338172502135,
+      "grad_norm": 5.9715800285339355,
+      "learning_rate": 0.00018262636112465464,
+      "loss": 11.21976318359375,
+      "step": 3210
+    },
+    {
+      "epoch": 3.6638200967833763,
+      "grad_norm": 5.637823104858398,
+      "learning_rate": 0.00018257218700904707,
+      "loss": 10.567466735839844,
+      "step": 3220
+    },
+    {
+      "epoch": 3.6752063763165386,
+      "grad_norm": 6.525228500366211,
+      "learning_rate": 0.00018251801289343952,
+      "loss": 10.669526672363281,
+      "step": 3230
+    },
+    {
+      "epoch": 3.686592655849701,
+      "grad_norm": 5.675905704498291,
+      "learning_rate": 0.00018246383877783194,
+      "loss": 10.238113403320312,
+      "step": 3240
+    },
+    {
+      "epoch": 3.6979789353828636,
+      "grad_norm": 6.053264617919922,
+      "learning_rate": 0.0001824096646622244,
+      "loss": 10.952618408203126,
+      "step": 3250
+    },
+    {
+      "epoch": 3.7093652149160263,
+      "grad_norm": 6.067233562469482,
+      "learning_rate": 0.00018235549054661682,
+      "loss": 10.320521545410156,
+      "step": 3260
+    },
+    {
+      "epoch": 3.7207514944491886,
+      "grad_norm": 6.502700328826904,
+      "learning_rate": 0.00018230131643100927,
+      "loss": 10.383914947509766,
+      "step": 3270
+    },
+    {
+      "epoch": 3.7321377739823514,
+      "grad_norm": 5.968015670776367,
+      "learning_rate": 0.00018224714231540172,
+      "loss": 10.5248779296875,
+      "step": 3280
+    },
+    {
+      "epoch": 3.7435240535155136,
+      "grad_norm": 5.4662628173828125,
+      "learning_rate": 0.00018219296819979415,
+      "loss": 10.531697082519532,
+      "step": 3290
+    },
+    {
+      "epoch": 3.7549103330486764,
+      "grad_norm": 5.9595255851745605,
+      "learning_rate": 0.0001821387940841866,
+      "loss": 10.61429443359375,
+      "step": 3300
+    },
+    {
+      "epoch": 3.766296612581839,
+      "grad_norm": 6.310218334197998,
+      "learning_rate": 0.00018208461996857902,
+      "loss": 10.1953125,
+      "step": 3310
+    },
+    {
+      "epoch": 3.7776828921150014,
+      "grad_norm": 5.952932834625244,
+      "learning_rate": 0.00018203044585297148,
+      "loss": 10.674753570556641,
+      "step": 3320
+    },
+    {
+      "epoch": 3.7890691716481637,
+      "grad_norm": 6.101319313049316,
+      "learning_rate": 0.0001819762717373639,
+      "loss": 10.45705337524414,
+      "step": 3330
+    },
+    {
+      "epoch": 3.8004554511813264,
+      "grad_norm": 5.611819744110107,
+      "learning_rate": 0.00018192209762175632,
+      "loss": 10.72674560546875,
+      "step": 3340
+    },
+    {
+      "epoch": 3.811841730714489,
+      "grad_norm": 5.927121162414551,
+      "learning_rate": 0.00018186792350614878,
+      "loss": 10.702400970458985,
+      "step": 3350
+    },
+    {
+      "epoch": 3.8232280102476515,
+      "grad_norm": 6.374614715576172,
+      "learning_rate": 0.0001818137493905412,
+      "loss": 10.934248352050782,
+      "step": 3360
+    },
+    {
+      "epoch": 3.834614289780814,
+      "grad_norm": 6.032021522521973,
+      "learning_rate": 0.00018175957527493363,
+      "loss": 10.430912017822266,
+      "step": 3370
+    },
+    {
+      "epoch": 3.8460005693139765,
+      "grad_norm": 5.958329677581787,
+      "learning_rate": 0.00018170540115932608,
+      "loss": 10.657380676269531,
+      "step": 3380
+    },
+    {
+      "epoch": 3.8573868488471392,
+      "grad_norm": 5.539936542510986,
+      "learning_rate": 0.00018165122704371853,
+      "loss": 10.864521026611328,
+      "step": 3390
+    },
+    {
+      "epoch": 3.868773128380302,
+      "grad_norm": 5.501955986022949,
+      "learning_rate": 0.00018159705292811095,
+      "loss": 10.671641540527343,
+      "step": 3400
+    },
+    {
+      "epoch": 3.8801594079134643,
+      "grad_norm": 6.021368503570557,
+      "learning_rate": 0.0001815428788125034,
+      "loss": 10.651559448242187,
+      "step": 3410
+    },
+    {
+      "epoch": 3.8915456874466265,
+      "grad_norm": 5.438447952270508,
+      "learning_rate": 0.00018148870469689583,
+      "loss": 10.541700744628907,
+      "step": 3420
+    },
+    {
+      "epoch": 3.9029319669797893,
+      "grad_norm": 5.7153215408325195,
+      "learning_rate": 0.00018143453058128828,
+      "loss": 10.635433959960938,
+      "step": 3430
+    },
+    {
+      "epoch": 3.914318246512952,
+      "grad_norm": 5.716714382171631,
+      "learning_rate": 0.0001813803564656807,
+      "loss": 10.716056060791015,
+      "step": 3440
+    },
+    {
+      "epoch": 3.9257045260461143,
+      "grad_norm": 5.737730979919434,
+      "learning_rate": 0.00018132618235007316,
+      "loss": 10.449008178710937,
+      "step": 3450
+    },
+    {
+      "epoch": 3.937090805579277,
+      "grad_norm": 5.603618621826172,
+      "learning_rate": 0.00018127200823446558,
+      "loss": 10.800203704833985,
+      "step": 3460
+    },
+    {
+      "epoch": 3.9484770851124393,
+      "grad_norm": 5.745213508605957,
+      "learning_rate": 0.00018121783411885803,
+      "loss": 10.808299255371093,
+      "step": 3470
+    },
+    {
+      "epoch": 3.959863364645602,
+      "grad_norm": 5.7340850830078125,
+      "learning_rate": 0.00018116366000325046,
+      "loss": 10.769782257080077,
+      "step": 3480
+    },
+    {
+      "epoch": 3.971249644178765,
+      "grad_norm": 6.192749500274658,
+      "learning_rate": 0.00018110948588764288,
+      "loss": 10.350373840332031,
+      "step": 3490
+    },
+    {
+      "epoch": 3.982635923711927,
+      "grad_norm": 5.680601119995117,
+      "learning_rate": 0.00018105531177203533,
+      "loss": 10.832538604736328,
+      "step": 3500
+    },
+    {
+      "epoch": 3.99402220324509,
+      "grad_norm": 5.494757652282715,
+      "learning_rate": 0.00018100113765642776,
+      "loss": 10.541521453857422,
+      "step": 3510
+    },
+    {
+      "epoch": 4.004554511813265,
+      "grad_norm": 5.503895282745361,
+      "learning_rate": 0.0001809469635408202,
+      "loss": 9.482205200195313,
+      "step": 3520
+    },
+    {
+      "epoch": 4.015940791346427,
+      "grad_norm": 5.427241802215576,
+      "learning_rate": 0.00018089278942521263,
+      "loss": 9.544134521484375,
+      "step": 3530
+    },
+    {
+      "epoch": 4.02732707087959,
+      "grad_norm": 6.108213424682617,
+      "learning_rate": 0.00018083861530960509,
+      "loss": 10.208637237548828,
+      "step": 3540
+    },
+    {
+      "epoch": 4.038713350412753,
+      "grad_norm": 5.874690055847168,
+      "learning_rate": 0.0001807844411939975,
+      "loss": 9.760848236083984,
+      "step": 3550
+    },
+    {
+      "epoch": 4.050099629945915,
+      "grad_norm": 5.867772102355957,
+      "learning_rate": 0.00018073026707838996,
+      "loss": 9.961907958984375,
+      "step": 3560
+    },
+    {
+      "epoch": 4.061485909479078,
+      "grad_norm": 5.899345874786377,
+      "learning_rate": 0.00018067609296278239,
+      "loss": 9.273954010009765,
+      "step": 3570
+    },
+    {
+      "epoch": 4.07287218901224,
+      "grad_norm": 6.068745136260986,
+      "learning_rate": 0.00018062191884717484,
+      "loss": 10.06835708618164,
+      "step": 3580
+    },
+    {
+      "epoch": 4.084258468545403,
+      "grad_norm": 6.152563095092773,
+      "learning_rate": 0.00018056774473156726,
+      "loss": 9.612554931640625,
+      "step": 3590
+    },
+    {
+      "epoch": 4.095644748078565,
+      "grad_norm": 6.043203830718994,
+      "learning_rate": 0.0001805135706159597,
+      "loss": 9.750363159179688,
+      "step": 3600
+    },
+    {
+      "epoch": 4.107031027611728,
+      "grad_norm": 5.493103504180908,
+      "learning_rate": 0.00018045939650035214,
+      "loss": 9.853363800048829,
+      "step": 3610
+    },
+    {
+      "epoch": 4.11841730714489,
+      "grad_norm": 6.1369242668151855,
+      "learning_rate": 0.0001804052223847446,
+      "loss": 10.104590606689452,
+      "step": 3620
+    },
+    {
+      "epoch": 4.129803586678053,
+      "grad_norm": 6.410492897033691,
+      "learning_rate": 0.00018035104826913701,
+      "loss": 10.213319396972656,
+      "step": 3630
+    },
+    {
+      "epoch": 4.141189866211215,
+      "grad_norm": 5.654990196228027,
+      "learning_rate": 0.00018029687415352944,
+      "loss": 9.714891052246093,
+      "step": 3640
+    },
+    {
+      "epoch": 4.152576145744378,
+      "grad_norm": 6.079676151275635,
+      "learning_rate": 0.0001802427000379219,
+      "loss": 9.392752838134765,
+      "step": 3650
+    },
+    {
+      "epoch": 4.163962425277541,
+      "grad_norm": 6.427825927734375,
+      "learning_rate": 0.00018018852592231431,
+      "loss": 9.643639373779298,
+      "step": 3660
+    },
+    {
+      "epoch": 4.175348704810703,
+      "grad_norm": 5.963044166564941,
+      "learning_rate": 0.00018013435180670677,
+      "loss": 9.924317932128906,
+      "step": 3670
+    },
+    {
+      "epoch": 4.1867349843438655,
+      "grad_norm": 5.854789733886719,
+      "learning_rate": 0.0001800801776910992,
+      "loss": 9.63187255859375,
+      "step": 3680
+    },
+    {
+      "epoch": 4.198121263877028,
+      "grad_norm": 5.785211086273193,
+      "learning_rate": 0.00018002600357549164,
+      "loss": 9.579486846923828,
+      "step": 3690
+    },
+    {
+      "epoch": 4.209507543410191,
+      "grad_norm": 5.718067169189453,
+      "learning_rate": 0.00017997182945988407,
+      "loss": 9.896450805664063,
+      "step": 3700
+    },
+    {
+      "epoch": 4.220893822943353,
+      "grad_norm": 6.051468849182129,
+      "learning_rate": 0.00017991765534427652,
+      "loss": 9.21618194580078,
+      "step": 3710
+    },
+    {
+      "epoch": 4.2322801024765155,
+      "grad_norm": 5.758811950683594,
+      "learning_rate": 0.00017986348122866894,
+      "loss": 9.783828735351562,
+      "step": 3720
+    },
+    {
+      "epoch": 4.243666382009678,
+      "grad_norm": 6.194498062133789,
+      "learning_rate": 0.0001798093071130614,
+      "loss": 9.940241241455078,
+      "step": 3730
+    },
+    {
+      "epoch": 4.255052661542841,
+      "grad_norm": 5.6434125900268555,
+      "learning_rate": 0.00017975513299745385,
+      "loss": 9.683363342285157,
+      "step": 3740
+    },
+    {
+      "epoch": 4.266438941076004,
+      "grad_norm": 5.981060028076172,
+      "learning_rate": 0.00017970095888184627,
+      "loss": 10.015727996826172,
+      "step": 3750
+    },
+    {
+      "epoch": 4.277825220609166,
+      "grad_norm": 5.95534610748291,
+      "learning_rate": 0.0001796467847662387,
+      "loss": 10.008995819091798,
+      "step": 3760
+    },
+    {
+      "epoch": 4.289211500142328,
+      "grad_norm": 6.140223503112793,
+      "learning_rate": 0.00017959261065063112,
+      "loss": 9.99697265625,
+      "step": 3770
+    },
+    {
+      "epoch": 4.300597779675491,
+      "grad_norm": 5.725556373596191,
+      "learning_rate": 0.00017953843653502357,
+      "loss": 9.58441390991211,
+      "step": 3780
+    },
+    {
+      "epoch": 4.311984059208654,
+      "grad_norm": 6.234734058380127,
+      "learning_rate": 0.000179484262419416,
+      "loss": 10.283954620361328,
+      "step": 3790
+    },
+    {
+      "epoch": 4.323370338741816,
+      "grad_norm": 5.869581699371338,
+      "learning_rate": 0.00017943008830380845,
+      "loss": 9.939014434814453,
+      "step": 3800
+    },
+    {
+      "epoch": 4.334756618274978,
+      "grad_norm": 6.069183826446533,
+      "learning_rate": 0.00017937591418820087,
+      "loss": 9.780485534667969,
+      "step": 3810
+    },
+    {
+      "epoch": 4.346142897808141,
+      "grad_norm": 6.241194725036621,
+      "learning_rate": 0.00017932174007259332,
+      "loss": 9.680929565429688,
+      "step": 3820
+    },
+    {
+      "epoch": 4.357529177341304,
+      "grad_norm": 5.660580635070801,
+      "learning_rate": 0.00017926756595698575,
+      "loss": 9.670439147949219,
+      "step": 3830
+    },
+    {
+      "epoch": 4.368915456874467,
+      "grad_norm": 6.770995140075684,
+      "learning_rate": 0.0001792133918413782,
+      "loss": 10.093508911132812,
+      "step": 3840
+    },
+    {
+      "epoch": 4.380301736407628,
+      "grad_norm": 6.052023887634277,
+      "learning_rate": 0.00017915921772577065,
+      "loss": 10.153280639648438,
+      "step": 3850
+    },
+    {
+      "epoch": 4.391688015940791,
+      "grad_norm": 5.781612396240234,
+      "learning_rate": 0.00017910504361016308,
+      "loss": 9.563631439208985,
+      "step": 3860
+    },
+    {
+      "epoch": 4.403074295473954,
+      "grad_norm": 6.453577995300293,
+      "learning_rate": 0.00017905086949455553,
+      "loss": 10.253567504882813,
+      "step": 3870
+    },
+    {
+      "epoch": 4.414460575007117,
+      "grad_norm": 5.5569539070129395,
+      "learning_rate": 0.00017899669537894795,
+      "loss": 9.466841888427734,
+      "step": 3880
+    },
+    {
+      "epoch": 4.425846854540279,
+      "grad_norm": 6.755964756011963,
+      "learning_rate": 0.0001789425212633404,
+      "loss": 10.00699234008789,
+      "step": 3890
+    },
+    {
+      "epoch": 4.437233134073441,
+      "grad_norm": 5.5527472496032715,
+      "learning_rate": 0.00017888834714773283,
+      "loss": 9.851504516601562,
+      "step": 3900
+    },
+    {
+      "epoch": 4.448619413606604,
+      "grad_norm": 6.0654096603393555,
+      "learning_rate": 0.00017883417303212525,
+      "loss": 9.837632751464843,
+      "step": 3910
+    },
+    {
+      "epoch": 4.460005693139767,
+      "grad_norm": 5.762067794799805,
+      "learning_rate": 0.00017877999891651768,
+      "loss": 9.580419158935547,
+      "step": 3920
+    },
+    {
+      "epoch": 4.471391972672929,
+      "grad_norm": 6.3049139976501465,
+      "learning_rate": 0.00017872582480091013,
+      "loss": 10.155924987792968,
+      "step": 3930
+    },
+    {
+      "epoch": 4.482778252206091,
+      "grad_norm": 6.560263633728027,
+      "learning_rate": 0.00017867165068530255,
+      "loss": 9.540550231933594,
+      "step": 3940
+    },
+    {
+      "epoch": 4.494164531739254,
+      "grad_norm": 6.04788064956665,
+      "learning_rate": 0.000178617476569695,
+      "loss": 9.617048645019532,
+      "step": 3950
+    },
+    {
+      "epoch": 4.505550811272417,
+      "grad_norm": 6.14173698425293,
+      "learning_rate": 0.00017856330245408746,
+      "loss": 9.965501403808593,
+      "step": 3960
+    },
+    {
+      "epoch": 4.5169370908055795,
+      "grad_norm": 6.081863880157471,
+      "learning_rate": 0.00017850912833847988,
+      "loss": 10.01316146850586,
+      "step": 3970
+    },
+    {
+      "epoch": 4.528323370338741,
+      "grad_norm": 5.400860786437988,
+      "learning_rate": 0.00017845495422287233,
+      "loss": 10.162345886230469,
+      "step": 3980
+    },
+    {
+      "epoch": 4.539709649871904,
+      "grad_norm": 5.537269592285156,
+      "learning_rate": 0.00017840078010726476,
+      "loss": 9.989215850830078,
+      "step": 3990
+    },
+    {
+      "epoch": 4.551095929405067,
+      "grad_norm": 5.948838710784912,
+      "learning_rate": 0.0001783466059916572,
+      "loss": 10.042298126220704,
+      "step": 4000
+    },
+    {
+      "epoch": 4.5624822089382295,
+      "grad_norm": 5.892082691192627,
+      "learning_rate": 0.00017829243187604963,
+      "loss": 10.084636688232422,
+      "step": 4010
+    },
+    {
+      "epoch": 4.573868488471392,
+      "grad_norm": 6.188459396362305,
+      "learning_rate": 0.00017823825776044208,
+      "loss": 9.930858612060547,
+      "step": 4020
+    },
+    {
+      "epoch": 4.585254768004555,
+      "grad_norm": 6.035775661468506,
+      "learning_rate": 0.0001781840836448345,
+      "loss": 9.804845428466797,
+      "step": 4030
+    },
+    {
+      "epoch": 4.596641047537717,
+      "grad_norm": 6.26588249206543,
+      "learning_rate": 0.00017812990952922696,
+      "loss": 9.714436340332032,
+      "step": 4040
+    },
+    {
+      "epoch": 4.60802732707088,
+      "grad_norm": 5.697126865386963,
+      "learning_rate": 0.00017807573541361938,
+      "loss": 9.831003570556641,
+      "step": 4050
+    },
+    {
+      "epoch": 4.619413606604042,
+      "grad_norm": 5.572251796722412,
+      "learning_rate": 0.0001780215612980118,
+      "loss": 10.07955551147461,
+      "step": 4060
+    },
+    {
+      "epoch": 4.630799886137204,
+      "grad_norm": 5.784109592437744,
+      "learning_rate": 0.00017796738718240426,
+      "loss": 9.46136016845703,
+      "step": 4070
+    },
+    {
+      "epoch": 4.642186165670367,
+      "grad_norm": 6.043442726135254,
+      "learning_rate": 0.00017791321306679668,
+      "loss": 10.091639709472656,
+      "step": 4080
+    },
+    {
+      "epoch": 4.65357244520353,
+      "grad_norm": 5.721872329711914,
+      "learning_rate": 0.00017785903895118914,
+      "loss": 10.043199157714843,
+      "step": 4090
+    },
+    {
+      "epoch": 4.664958724736692,
+      "grad_norm": 5.753310203552246,
+      "learning_rate": 0.00017780486483558156,
+      "loss": 9.644969940185547,
+      "step": 4100
+    },
+    {
+      "epoch": 4.676345004269855,
+      "grad_norm": 6.087526321411133,
+      "learning_rate": 0.000177750690719974,
+      "loss": 9.953328704833984,
+      "step": 4110
+    },
+    {
+      "epoch": 4.687731283803018,
+      "grad_norm": 5.621133327484131,
+      "learning_rate": 0.00017769651660436644,
+      "loss": 9.69726104736328,
+      "step": 4120
+    },
+    {
+      "epoch": 4.69911756333618,
+      "grad_norm": 6.229346752166748,
+      "learning_rate": 0.0001776423424887589,
+      "loss": 9.628369140625,
+      "step": 4130
+    },
+    {
+      "epoch": 4.710503842869342,
+      "grad_norm": 5.986842155456543,
+      "learning_rate": 0.0001775881683731513,
+      "loss": 9.742918395996094,
+      "step": 4140
+    },
+    {
+      "epoch": 4.721890122402505,
+      "grad_norm": 5.907449245452881,
+      "learning_rate": 0.00017753399425754376,
+      "loss": 9.967567443847656,
+      "step": 4150
+    },
+    {
+      "epoch": 4.733276401935668,
+      "grad_norm": 6.126130104064941,
+      "learning_rate": 0.0001774798201419362,
+      "loss": 9.773100280761719,
+      "step": 4160
+    },
+    {
+      "epoch": 4.74466268146883,
+      "grad_norm": 5.871231555938721,
+      "learning_rate": 0.00017742564602632864,
+      "loss": 9.900487518310547,
+      "step": 4170
+    },
+    {
+      "epoch": 4.7560489610019925,
+      "grad_norm": 5.7376275062561035,
+      "learning_rate": 0.00017737147191072106,
+      "loss": 9.520377349853515,
+      "step": 4180
+    },
+    {
+      "epoch": 4.767435240535155,
+      "grad_norm": 5.596216201782227,
+      "learning_rate": 0.0001773172977951135,
+      "loss": 10.175640869140626,
+      "step": 4190
+    },
+    {
+      "epoch": 4.778821520068318,
+      "grad_norm": 6.0666351318359375,
+      "learning_rate": 0.00017726312367950594,
+      "loss": 9.647950744628906,
+      "step": 4200
+    },
+    {
+      "epoch": 4.790207799601481,
+      "grad_norm": 6.583392143249512,
+      "learning_rate": 0.00017720894956389837,
+      "loss": 9.790153503417969,
+      "step": 4210
+    },
+    {
+      "epoch": 4.8015940791346425,
+      "grad_norm": 5.613272190093994,
+      "learning_rate": 0.00017715477544829082,
+      "loss": 9.466513061523438,
+      "step": 4220
+    },
+    {
+      "epoch": 4.812980358667805,
+      "grad_norm": 5.842050552368164,
+      "learning_rate": 0.00017710060133268324,
+      "loss": 9.622956085205079,
+      "step": 4230
+    },
+    {
+      "epoch": 4.824366638200968,
+      "grad_norm": 5.759915828704834,
+      "learning_rate": 0.0001770464272170757,
+      "loss": 9.80495834350586,
+      "step": 4240
+    },
+    {
+      "epoch": 4.835752917734131,
+      "grad_norm": 5.608836650848389,
+      "learning_rate": 0.00017699225310146812,
+      "loss": 9.477723693847656,
+      "step": 4250
+    },
+    {
+      "epoch": 4.847139197267293,
+      "grad_norm": 6.463353157043457,
+      "learning_rate": 0.00017693807898586057,
+      "loss": 9.470064544677735,
+      "step": 4260
+    },
+    {
+      "epoch": 4.858525476800455,
+      "grad_norm": 5.962118148803711,
+      "learning_rate": 0.000176883904870253,
+      "loss": 9.367919921875,
+      "step": 4270
+    },
+    {
+      "epoch": 4.869911756333618,
+      "grad_norm": 5.565269470214844,
+      "learning_rate": 0.00017682973075464544,
+      "loss": 9.758961486816407,
+      "step": 4280
+    },
+    {
+      "epoch": 4.881298035866781,
+      "grad_norm": 5.9193010330200195,
+      "learning_rate": 0.0001767755566390379,
+      "loss": 9.685740661621093,
+      "step": 4290
+    },
+    {
+      "epoch": 4.8926843153999435,
+      "grad_norm": 5.730422019958496,
+      "learning_rate": 0.00017672138252343032,
+      "loss": 9.70977554321289,
+      "step": 4300
+    },
+    {
+      "epoch": 4.904070594933105,
+      "grad_norm": 5.72115421295166,
+      "learning_rate": 0.00017666720840782277,
+      "loss": 9.60604476928711,
+      "step": 4310
+    },
+    {
+      "epoch": 4.915456874466268,
+      "grad_norm": 6.489387035369873,
+      "learning_rate": 0.0001766130342922152,
+      "loss": 9.322960662841798,
+      "step": 4320
+    },
+    {
+      "epoch": 4.926843153999431,
+      "grad_norm": 5.830432415008545,
+      "learning_rate": 0.00017655886017660762,
+      "loss": 9.564017486572265,
+      "step": 4330
+    },
+    {
+      "epoch": 4.938229433532594,
+      "grad_norm": 6.302420616149902,
+      "learning_rate": 0.00017650468606100005,
+      "loss": 10.048651123046875,
+      "step": 4340
+    },
+    {
+      "epoch": 4.949615713065755,
+      "grad_norm": 6.4496660232543945,
+      "learning_rate": 0.0001764505119453925,
+      "loss": 9.775814819335938,
+      "step": 4350
+    },
+    {
+      "epoch": 4.961001992598918,
+      "grad_norm": 6.146448612213135,
+      "learning_rate": 0.00017639633782978492,
+      "loss": 9.664933013916016,
+      "step": 4360
+    },
+    {
+      "epoch": 4.972388272132081,
+      "grad_norm": 6.442460060119629,
+      "learning_rate": 0.00017634216371417737,
+      "loss": 9.787579345703126,
+      "step": 4370
+    },
+    {
+      "epoch": 4.983774551665244,
+      "grad_norm": 5.952507019042969,
+      "learning_rate": 0.0001762879895985698,
+      "loss": 9.962568664550782,
+      "step": 4380
+    },
+    {
+      "epoch": 4.995160831198406,
+      "grad_norm": 5.775949954986572,
+      "learning_rate": 0.00017623381548296225,
+      "loss": 9.72928466796875,
+      "step": 4390
+    },
+    {
+      "epoch": 5.005693139766581,
+      "grad_norm": 6.083469390869141,
+      "learning_rate": 0.00017617964136735467,
+      "loss": 8.628085327148437,
+      "step": 4400
+    },
+    {
+      "epoch": 5.017079419299744,
+      "grad_norm": 6.120087623596191,
+      "learning_rate": 0.00017612546725174713,
+      "loss": 8.888311004638672,
+      "step": 4410
+    },
+    {
+      "epoch": 5.028465698832906,
+      "grad_norm": 6.082287311553955,
+      "learning_rate": 0.00017607129313613958,
+      "loss": 9.107136535644532,
+      "step": 4420
+    },
+    {
+      "epoch": 5.039851978366069,
+      "grad_norm": 5.743897438049316,
+      "learning_rate": 0.000176017119020532,
+      "loss": 8.61147232055664,
+      "step": 4430
+    },
+    {
+      "epoch": 5.051238257899231,
+      "grad_norm": 5.427106857299805,
+      "learning_rate": 0.00017596294490492445,
+      "loss": 8.694246673583985,
+      "step": 4440
+    },
+    {
+      "epoch": 5.062624537432394,
+      "grad_norm": 6.282980442047119,
+      "learning_rate": 0.00017590877078931688,
+      "loss": 9.144784545898437,
+      "step": 4450
+    },
+    {
+      "epoch": 5.074010816965557,
+      "grad_norm": 6.395759105682373,
+      "learning_rate": 0.0001758545966737093,
+      "loss": 8.932855224609375,
+      "step": 4460
+    },
+    {
+      "epoch": 5.085397096498719,
+      "grad_norm": 6.1782917976379395,
+      "learning_rate": 0.00017580042255810175,
+      "loss": 9.115586853027343,
+      "step": 4470
+    },
+    {
+      "epoch": 5.0967833760318815,
+      "grad_norm": 6.45112943649292,
+      "learning_rate": 0.00017574624844249418,
+      "loss": 8.773223876953125,
+      "step": 4480
+    },
+    {
+      "epoch": 5.108169655565044,
+      "grad_norm": 5.9199371337890625,
+      "learning_rate": 0.0001756920743268866,
+      "loss": 8.678252410888671,
+      "step": 4490
+    },
+    {
+      "epoch": 5.119555935098207,
+      "grad_norm": 5.901211738586426,
+      "learning_rate": 0.00017563790021127905,
+      "loss": 9.167632293701171,
+      "step": 4500
+    },
+    {
+      "epoch": 5.130942214631369,
+      "grad_norm": 6.0059661865234375,
+      "learning_rate": 0.00017558372609567148,
+      "loss": 8.844408416748047,
+      "step": 4510
+    },
+    {
+      "epoch": 5.1423284941645315,
+      "grad_norm": 6.615947246551514,
+      "learning_rate": 0.00017552955198006393,
+      "loss": 8.853307342529297,
+      "step": 4520
+    },
+    {
+      "epoch": 5.153714773697694,
+      "grad_norm": 5.707233428955078,
+      "learning_rate": 0.00017547537786445638,
+      "loss": 8.580878448486327,
+      "step": 4530
+    },
+    {
+      "epoch": 5.165101053230857,
+      "grad_norm": 6.256187915802002,
+      "learning_rate": 0.0001754212037488488,
+      "loss": 8.889945983886719,
+      "step": 4540
+    },
+    {
+      "epoch": 5.17648733276402,
+      "grad_norm": 5.724560260772705,
+      "learning_rate": 0.00017536702963324126,
+      "loss": 9.283840942382813,
+      "step": 4550
+    },
+    {
+      "epoch": 5.187873612297182,
+      "grad_norm": 6.063653469085693,
+      "learning_rate": 0.00017531285551763368,
+      "loss": 9.45126953125,
+      "step": 4560
+    },
+    {
+      "epoch": 5.199259891830344,
+      "grad_norm": 6.490650177001953,
+      "learning_rate": 0.00017525868140202613,
+      "loss": 9.213851928710938,
+      "step": 4570
+    },
+    {
+      "epoch": 5.210646171363507,
+      "grad_norm": 6.077616214752197,
+      "learning_rate": 0.00017520450728641856,
+      "loss": 8.94876937866211,
+      "step": 4580
+    },
+    {
+      "epoch": 5.22203245089667,
+      "grad_norm": 6.140579700469971,
+      "learning_rate": 0.000175150333170811,
+      "loss": 9.061854553222656,
+      "step": 4590
+    },
+    {
+      "epoch": 5.233418730429832,
+      "grad_norm": 6.086933135986328,
+      "learning_rate": 0.00017509615905520343,
+      "loss": 8.97676773071289,
+      "step": 4600
+    },
+    {
+      "epoch": 5.244805009962994,
+      "grad_norm": 6.013683319091797,
+      "learning_rate": 0.00017504198493959586,
+      "loss": 9.112580871582031,
+      "step": 4610
+    },
+    {
+      "epoch": 5.256191289496157,
+      "grad_norm": 6.05435037612915,
+      "learning_rate": 0.00017498781082398828,
+      "loss": 8.846112823486328,
+      "step": 4620
+    },
+    {
+      "epoch": 5.26757756902932,
+      "grad_norm": 6.0196919441223145,
+      "learning_rate": 0.00017493363670838074,
+      "loss": 9.190632629394532,
+      "step": 4630
+    },
+    {
+      "epoch": 5.278963848562483,
+      "grad_norm": 5.778266906738281,
+      "learning_rate": 0.0001748794625927732,
+      "loss": 9.008673858642577,
+      "step": 4640
+    },
+    {
+      "epoch": 5.290350128095644,
+      "grad_norm": 6.087759971618652,
+      "learning_rate": 0.0001748252884771656,
+      "loss": 9.25833969116211,
+      "step": 4650
+    },
+    {
+      "epoch": 5.301736407628807,
+      "grad_norm": 5.978583812713623,
+      "learning_rate": 0.00017477111436155806,
+      "loss": 8.707977294921875,
+      "step": 4660
+    },
+    {
+      "epoch": 5.31312268716197,
+      "grad_norm": 5.761394500732422,
+      "learning_rate": 0.0001747169402459505,
+      "loss": 9.27506103515625,
+      "step": 4670
+    },
+    {
+      "epoch": 5.324508966695133,
+      "grad_norm": 6.357019901275635,
+      "learning_rate": 0.00017466276613034294,
+      "loss": 9.029557037353516,
+      "step": 4680
+    },
+    {
+      "epoch": 5.335895246228295,
+      "grad_norm": 6.016201496124268,
+      "learning_rate": 0.00017460859201473536,
+      "loss": 8.777602386474609,
+      "step": 4690
+    },
+    {
+      "epoch": 5.347281525761457,
+      "grad_norm": 6.20839786529541,
+      "learning_rate": 0.00017455441789912781,
+      "loss": 8.966375732421875,
+      "step": 4700
+    },
+    {
+      "epoch": 5.35866780529462,
+      "grad_norm": 6.455466270446777,
+      "learning_rate": 0.00017450024378352024,
+      "loss": 9.284576416015625,
+      "step": 4710
+    },
+    {
+      "epoch": 5.370054084827783,
+      "grad_norm": 6.2462663650512695,
+      "learning_rate": 0.0001744460696679127,
+      "loss": 8.876305389404298,
+      "step": 4720
+    },
+    {
+      "epoch": 5.381440364360945,
+      "grad_norm": 6.3350396156311035,
+      "learning_rate": 0.00017439189555230512,
+      "loss": 9.187533569335937,
+      "step": 4730
+    },
+    {
+      "epoch": 5.392826643894107,
+      "grad_norm": 6.246124267578125,
+      "learning_rate": 0.00017433772143669757,
+      "loss": 8.913845825195313,
+      "step": 4740
+    },
+    {
+      "epoch": 5.40421292342727,
+      "grad_norm": 5.948974609375,
+      "learning_rate": 0.00017428354732109,
+      "loss": 9.092901611328125,
+      "step": 4750
+    },
+    {
+      "epoch": 5.415599202960433,
+      "grad_norm": 6.238656520843506,
+      "learning_rate": 0.00017422937320548242,
+      "loss": 9.446102905273438,
+      "step": 4760
+    },
+    {
+      "epoch": 5.4269854824935955,
+      "grad_norm": 6.002922058105469,
+      "learning_rate": 0.00017417519908987487,
+      "loss": 9.022492980957031,
+      "step": 4770
+    },
+    {
+      "epoch": 5.438371762026758,
+      "grad_norm": 5.8916707038879395,
+      "learning_rate": 0.0001741210249742673,
+      "loss": 8.732567596435548,
+      "step": 4780
+    },
+    {
+      "epoch": 5.44975804155992,
+      "grad_norm": 6.253623008728027,
+      "learning_rate": 0.00017406685085865974,
+      "loss": 8.986526489257812,
+      "step": 4790
+    },
+    {
+      "epoch": 5.461144321093083,
+      "grad_norm": 6.794173717498779,
+      "learning_rate": 0.00017401267674305217,
+      "loss": 8.910897827148437,
+      "step": 4800
+    },
+    {
+      "epoch": 5.4725306006262455,
+      "grad_norm": 6.226437568664551,
+      "learning_rate": 0.00017395850262744462,
+      "loss": 9.053468322753906,
+      "step": 4810
+    },
+    {
+      "epoch": 5.483916880159408,
+      "grad_norm": 5.871084213256836,
+      "learning_rate": 0.00017390432851183704,
+      "loss": 8.758150482177735,
+      "step": 4820
+    },
+    {
+      "epoch": 5.49530315969257,
+      "grad_norm": 6.246805667877197,
+      "learning_rate": 0.0001738501543962295,
+      "loss": 9.195549774169923,
+      "step": 4830
+    },
+    {
+      "epoch": 5.506689439225733,
+      "grad_norm": 6.047657489776611,
+      "learning_rate": 0.00017379598028062192,
+      "loss": 9.35385513305664,
+      "step": 4840
+    },
+    {
+      "epoch": 5.518075718758896,
+      "grad_norm": 5.7672529220581055,
+      "learning_rate": 0.00017374180616501437,
+      "loss": 9.131436157226563,
+      "step": 4850
+    },
+    {
+      "epoch": 5.529461998292058,
+      "grad_norm": 6.151646614074707,
+      "learning_rate": 0.00017368763204940682,
+      "loss": 9.32352523803711,
+      "step": 4860
+    },
+    {
+      "epoch": 5.540848277825221,
+      "grad_norm": 6.154118537902832,
+      "learning_rate": 0.00017363345793379925,
+      "loss": 9.332242584228515,
+      "step": 4870
+    },
+    {
+      "epoch": 5.552234557358383,
+      "grad_norm": 5.870425224304199,
+      "learning_rate": 0.00017357928381819167,
+      "loss": 8.844430541992187,
+      "step": 4880
+    },
+    {
+      "epoch": 5.563620836891546,
+      "grad_norm": 6.215554237365723,
+      "learning_rate": 0.0001735251097025841,
+      "loss": 9.3614013671875,
+      "step": 4890
+    },
+    {
+      "epoch": 5.575007116424708,
+      "grad_norm": 6.865769386291504,
+      "learning_rate": 0.00017347093558697655,
+      "loss": 8.7501708984375,
+      "step": 4900
+    },
+    {
+      "epoch": 5.586393395957871,
+      "grad_norm": 5.617834091186523,
+      "learning_rate": 0.00017341676147136897,
+      "loss": 9.023835754394531,
+      "step": 4910
+    },
+    {
+      "epoch": 5.597779675491033,
+      "grad_norm": 6.217780590057373,
+      "learning_rate": 0.00017336258735576142,
+      "loss": 8.83472900390625,
+      "step": 4920
+    },
+    {
+      "epoch": 5.609165955024196,
+      "grad_norm": 6.248758316040039,
+      "learning_rate": 0.00017330841324015385,
+      "loss": 9.115097808837891,
+      "step": 4930
+    },
+    {
+      "epoch": 5.620552234557358,
+      "grad_norm": 5.923741817474365,
+      "learning_rate": 0.0001732542391245463,
+      "loss": 9.077340698242187,
+      "step": 4940
+    },
+    {
+      "epoch": 5.631938514090521,
+      "grad_norm": 6.346497058868408,
+      "learning_rate": 0.00017320006500893872,
+      "loss": 9.040455627441407,
+      "step": 4950
+    },
+    {
+      "epoch": 5.643324793623684,
+      "grad_norm": 6.081203937530518,
+      "learning_rate": 0.00017314589089333118,
+      "loss": 9.156749725341797,
+      "step": 4960
+    },
+    {
+      "epoch": 5.654711073156846,
+      "grad_norm": 5.836169719696045,
+      "learning_rate": 0.00017309171677772363,
+      "loss": 9.09630126953125,
+      "step": 4970
+    },
+    {
+      "epoch": 5.6660973526900085,
+      "grad_norm": 6.374985218048096,
+      "learning_rate": 0.00017303754266211605,
+      "loss": 9.138162231445312,
+      "step": 4980
+    },
+    {
+      "epoch": 5.677483632223171,
+      "grad_norm": 6.967176914215088,
+      "learning_rate": 0.0001729833685465085,
+      "loss": 9.05536346435547,
+      "step": 4990
+    },
+    {
+      "epoch": 5.688869911756334,
+      "grad_norm": 6.718258857727051,
+      "learning_rate": 0.00017292919443090093,
+      "loss": 8.888587951660156,
+      "step": 5000
+    },
+    {
+      "epoch": 5.700256191289496,
+      "grad_norm": 6.106222629547119,
+      "learning_rate": 0.00017287502031529338,
+      "loss": 9.413323974609375,
+      "step": 5010
+    },
+    {
+      "epoch": 5.7116424708226585,
+      "grad_norm": 5.6303229331970215,
+      "learning_rate": 0.0001728208461996858,
+      "loss": 9.086915588378906,
+      "step": 5020
+    },
+    {
+      "epoch": 5.723028750355821,
+      "grad_norm": 6.345915794372559,
+      "learning_rate": 0.00017276667208407823,
+      "loss": 8.981614685058593,
+      "step": 5030
+    },
+    {
+      "epoch": 5.734415029888984,
+      "grad_norm": 5.963151454925537,
+      "learning_rate": 0.00017271249796847065,
+      "loss": 8.846481323242188,
+      "step": 5040
+    },
+    {
+      "epoch": 5.745801309422147,
+      "grad_norm": 6.332173824310303,
+      "learning_rate": 0.0001726583238528631,
+      "loss": 9.01428451538086,
+      "step": 5050
+    },
+    {
+      "epoch": 5.757187588955309,
+      "grad_norm": 5.805280685424805,
+      "learning_rate": 0.00017260414973725553,
+      "loss": 8.98281021118164,
+      "step": 5060
+    },
+    {
+      "epoch": 5.768573868488471,
+      "grad_norm": 6.478013038635254,
+      "learning_rate": 0.00017254997562164798,
+      "loss": 9.026184844970704,
+      "step": 5070
+    },
+    {
+      "epoch": 5.779960148021634,
+      "grad_norm": 6.193087100982666,
+      "learning_rate": 0.00017249580150604043,
+      "loss": 8.499151611328125,
+      "step": 5080
+    },
+    {
+      "epoch": 5.791346427554797,
+      "grad_norm": 6.035495758056641,
+      "learning_rate": 0.00017244162739043286,
+      "loss": 9.09105224609375,
+      "step": 5090
+    },
+    {
+      "epoch": 5.802732707087959,
+      "grad_norm": 6.009443759918213,
+      "learning_rate": 0.0001723874532748253,
+      "loss": 9.080800628662109,
+      "step": 5100
+    },
+    {
+      "epoch": 5.814118986621121,
+      "grad_norm": 6.147922992706299,
+      "learning_rate": 0.00017233327915921773,
+      "loss": 9.082097625732422,
+      "step": 5110
+    },
+    {
+      "epoch": 5.825505266154284,
+      "grad_norm": 6.0678019523620605,
+      "learning_rate": 0.00017227910504361018,
+      "loss": 8.975556182861329,
+      "step": 5120
+    },
+    {
+      "epoch": 5.836891545687447,
+      "grad_norm": 5.851400852203369,
+      "learning_rate": 0.0001722249309280026,
+      "loss": 8.932150268554688,
+      "step": 5130
+    },
+    {
+      "epoch": 5.84827782522061,
+      "grad_norm": 5.831391334533691,
+      "learning_rate": 0.00017217075681239506,
+      "loss": 8.861358642578125,
+      "step": 5140
+    },
+    {
+      "epoch": 5.859664104753771,
+      "grad_norm": 6.788999557495117,
+      "learning_rate": 0.00017211658269678749,
+      "loss": 9.362552642822266,
+      "step": 5150
+    },
+    {
+      "epoch": 5.871050384286934,
+      "grad_norm": 6.117767333984375,
+      "learning_rate": 0.00017206240858117994,
+      "loss": 9.051403045654297,
+      "step": 5160
+    },
+    {
+      "epoch": 5.882436663820097,
+      "grad_norm": 5.815280914306641,
+      "learning_rate": 0.00017200823446557236,
+      "loss": 8.855480194091797,
+      "step": 5170
+    },
+    {
+      "epoch": 5.89382294335326,
+      "grad_norm": 6.476868152618408,
+      "learning_rate": 0.00017195406034996479,
+      "loss": 9.67321319580078,
+      "step": 5180
+    },
+    {
+      "epoch": 5.9052092228864215,
+      "grad_norm": 6.074749946594238,
+      "learning_rate": 0.0001718998862343572,
+      "loss": 8.859770965576171,
+      "step": 5190
+    },
+    {
+      "epoch": 5.916595502419584,
+      "grad_norm": 5.674811840057373,
+      "learning_rate": 0.00017184571211874966,
+      "loss": 9.397520446777344,
+      "step": 5200
+    },
+    {
+      "epoch": 5.927981781952747,
+      "grad_norm": 5.898608684539795,
+      "learning_rate": 0.0001717915380031421,
+      "loss": 9.098382568359375,
+      "step": 5210
+    },
+    {
+      "epoch": 5.93936806148591,
+      "grad_norm": 6.260279178619385,
+      "learning_rate": 0.00017173736388753454,
+      "loss": 9.085452270507812,
+      "step": 5220
+    },
+    {
+      "epoch": 5.950754341019072,
+      "grad_norm": 6.2707719802856445,
+      "learning_rate": 0.000171683189771927,
+      "loss": 8.759799194335937,
+      "step": 5230
+    },
+    {
+      "epoch": 5.962140620552234,
+      "grad_norm": 6.20477294921875,
+      "learning_rate": 0.00017162901565631941,
+      "loss": 9.199006652832031,
+      "step": 5240
+    },
+    {
+      "epoch": 5.973526900085397,
+      "grad_norm": 5.900763511657715,
+      "learning_rate": 0.00017157484154071187,
+      "loss": 9.182913970947265,
+      "step": 5250
+    },
+    {
+      "epoch": 5.98491317961856,
+      "grad_norm": 6.241118431091309,
+      "learning_rate": 0.0001715206674251043,
+      "loss": 8.96974868774414,
+      "step": 5260
+    },
+    {
+      "epoch": 5.9962994591517225,
+      "grad_norm": 5.996070384979248,
+      "learning_rate": 0.00017146649330949674,
+      "loss": 9.107511138916015,
+      "step": 5270
+    },
+    {
+      "epoch": 6.0068317677198975,
+      "grad_norm": 6.609316349029541,
+      "learning_rate": 0.00017141231919388917,
+      "loss": 7.848031616210937,
+      "step": 5280
+    },
+    {
+      "epoch": 6.01821804725306,
+      "grad_norm": 6.0844197273254395,
+      "learning_rate": 0.00017135814507828162,
+      "loss": 8.242960357666016,
+      "step": 5290
+    },
+    {
+      "epoch": 6.029604326786223,
+      "grad_norm": 5.959528923034668,
+      "learning_rate": 0.00017130397096267404,
+      "loss": 8.17882080078125,
+      "step": 5300
+    },
+    {
+      "epoch": 6.040990606319385,
+      "grad_norm": 5.812763690948486,
+      "learning_rate": 0.00017124979684706647,
+      "loss": 8.189881134033204,
+      "step": 5310
+    },
+    {
+      "epoch": 6.0523768858525475,
+      "grad_norm": 6.203407287597656,
+      "learning_rate": 0.00017119562273145892,
+      "loss": 7.9870758056640625,
+      "step": 5320
+    },
+    {
+      "epoch": 6.06376316538571,
+      "grad_norm": 6.108156681060791,
+      "learning_rate": 0.00017114144861585134,
+      "loss": 8.187033081054688,
+      "step": 5330
+    },
+    {
+      "epoch": 6.075149444918873,
+      "grad_norm": 6.818889617919922,
+      "learning_rate": 0.0001710872745002438,
+      "loss": 8.802757263183594,
+      "step": 5340
+    },
+    {
+      "epoch": 6.086535724452036,
+      "grad_norm": 6.080479145050049,
+      "learning_rate": 0.00017103310038463622,
+      "loss": 8.3635986328125,
+      "step": 5350
+    },
+    {
+      "epoch": 6.097922003985198,
+      "grad_norm": 6.009644508361816,
+      "learning_rate": 0.00017097892626902867,
+      "loss": 8.316793060302734,
+      "step": 5360
+    },
+    {
+      "epoch": 6.10930828351836,
+      "grad_norm": 5.928283214569092,
+      "learning_rate": 0.0001709247521534211,
+      "loss": 8.05916976928711,
+      "step": 5370
+    },
+    {
+      "epoch": 6.120694563051523,
+      "grad_norm": 6.18734073638916,
+      "learning_rate": 0.00017087057803781355,
+      "loss": 8.414893341064452,
+      "step": 5380
+    },
+    {
+      "epoch": 6.132080842584686,
+      "grad_norm": 5.8552021980285645,
+      "learning_rate": 0.00017081640392220597,
+      "loss": 8.328546905517578,
+      "step": 5390
+    },
+    {
+      "epoch": 6.143467122117848,
+      "grad_norm": 6.283035755157471,
+      "learning_rate": 0.00017076222980659842,
+      "loss": 8.294504547119141,
+      "step": 5400
+    },
+    {
+      "epoch": 6.15485340165101,
+      "grad_norm": 6.527463436126709,
+      "learning_rate": 0.00017070805569099085,
+      "loss": 8.205842590332031,
+      "step": 5410
+    },
+    {
+      "epoch": 6.166239681184173,
+      "grad_norm": 6.081474304199219,
+      "learning_rate": 0.0001706538815753833,
+      "loss": 8.215932464599609,
+      "step": 5420
+    },
+    {
+      "epoch": 6.177625960717336,
+      "grad_norm": 6.173142433166504,
+      "learning_rate": 0.00017059970745977575,
+      "loss": 8.617554473876954,
+      "step": 5430
+    },
+    {
+      "epoch": 6.189012240250499,
+      "grad_norm": 6.146962642669678,
+      "learning_rate": 0.00017054553334416817,
+      "loss": 8.21309814453125,
+      "step": 5440
+    },
+    {
+      "epoch": 6.20039851978366,
+      "grad_norm": 6.535096645355225,
+      "learning_rate": 0.0001704913592285606,
+      "loss": 8.393982696533204,
+      "step": 5450
+    },
+    {
+      "epoch": 6.211784799316823,
+      "grad_norm": 5.731675624847412,
+      "learning_rate": 0.00017043718511295302,
+      "loss": 8.488709259033204,
+      "step": 5460
+    },
+    {
+      "epoch": 6.223171078849986,
+      "grad_norm": 6.357123374938965,
+      "learning_rate": 0.00017038301099734547,
+      "loss": 8.502271270751953,
+      "step": 5470
+    },
+    {
+      "epoch": 6.234557358383149,
+      "grad_norm": 6.319135665893555,
+      "learning_rate": 0.0001703288368817379,
+      "loss": 8.702040100097657,
+      "step": 5480
+    },
+    {
+      "epoch": 6.2459436379163105,
+      "grad_norm": 6.496546268463135,
+      "learning_rate": 0.00017027466276613035,
+      "loss": 8.020443725585938,
+      "step": 5490
+    },
+    {
+      "epoch": 6.257329917449473,
+      "grad_norm": 6.284101963043213,
+      "learning_rate": 0.00017022048865052278,
+      "loss": 8.432199096679687,
+      "step": 5500
+    },
+    {
+      "epoch": 6.268716196982636,
+      "grad_norm": 6.48223876953125,
+      "learning_rate": 0.00017016631453491523,
+      "loss": 8.392076873779297,
+      "step": 5510
+    },
+    {
+      "epoch": 6.280102476515799,
+      "grad_norm": 6.497861862182617,
+      "learning_rate": 0.00017011214041930765,
+      "loss": 8.759424591064453,
+      "step": 5520
+    },
+    {
+      "epoch": 6.291488756048961,
+      "grad_norm": 6.510660648345947,
+      "learning_rate": 0.0001700579663037001,
+      "loss": 8.286143493652343,
+      "step": 5530
+    },
+    {
+      "epoch": 6.302875035582123,
+      "grad_norm": 5.99029541015625,
+      "learning_rate": 0.00017000379218809255,
+      "loss": 8.367719268798828,
+      "step": 5540
+    },
+    {
+      "epoch": 6.314261315115286,
+      "grad_norm": 6.750374794006348,
+      "learning_rate": 0.00016994961807248498,
+      "loss": 8.649713134765625,
+      "step": 5550
+    },
+    {
+      "epoch": 6.325647594648449,
+      "grad_norm": 5.9475202560424805,
+      "learning_rate": 0.00016989544395687743,
+      "loss": 8.837234497070312,
+      "step": 5560
+    },
+    {
+      "epoch": 6.3370338741816115,
+      "grad_norm": 6.30899715423584,
+      "learning_rate": 0.00016984126984126986,
+      "loss": 8.32588882446289,
+      "step": 5570
+    },
+    {
+      "epoch": 6.348420153714773,
+      "grad_norm": 6.342610836029053,
+      "learning_rate": 0.00016978709572566228,
+      "loss": 8.394316101074219,
+      "step": 5580
+    },
+    {
+      "epoch": 6.359806433247936,
+      "grad_norm": 6.095706462860107,
+      "learning_rate": 0.00016973292161005473,
+      "loss": 8.327243041992187,
+      "step": 5590
+    },
+    {
+      "epoch": 6.371192712781099,
+      "grad_norm": 6.3162336349487305,
+      "learning_rate": 0.00016967874749444716,
+      "loss": 8.443352508544923,
+      "step": 5600
+    },
+    {
+      "epoch": 6.3825789923142615,
+      "grad_norm": 6.393646240234375,
+      "learning_rate": 0.00016962457337883958,
+      "loss": 8.368497467041015,
+      "step": 5610
+    },
+    {
+      "epoch": 6.393965271847424,
+      "grad_norm": 6.893944263458252,
+      "learning_rate": 0.00016957039926323203,
+      "loss": 8.11514892578125,
+      "step": 5620
+    },
+    {
+      "epoch": 6.405351551380586,
+      "grad_norm": 6.172192573547363,
+      "learning_rate": 0.00016951622514762446,
+      "loss": 8.440129852294922,
+      "step": 5630
+    },
+    {
+      "epoch": 6.416737830913749,
+      "grad_norm": 6.662540435791016,
+      "learning_rate": 0.0001694620510320169,
+      "loss": 8.530294799804688,
+      "step": 5640
+    },
+    {
+      "epoch": 6.428124110446912,
+      "grad_norm": 6.44663143157959,
+      "learning_rate": 0.00016940787691640936,
+      "loss": 8.647301483154298,
+      "step": 5650
+    },
+    {
+      "epoch": 6.439510389980074,
+      "grad_norm": 7.118195056915283,
+      "learning_rate": 0.00016935370280080178,
+      "loss": 8.56711654663086,
+      "step": 5660
+    },
+    {
+      "epoch": 6.450896669513236,
+      "grad_norm": 5.5173540115356445,
+      "learning_rate": 0.00016929952868519424,
+      "loss": 8.647555541992187,
+      "step": 5670
+    },
+    {
+      "epoch": 6.462282949046399,
+      "grad_norm": 5.75939416885376,
+      "learning_rate": 0.00016924535456958666,
+      "loss": 8.223889923095703,
+      "step": 5680
+    },
+    {
+      "epoch": 6.473669228579562,
+      "grad_norm": 6.170014381408691,
+      "learning_rate": 0.0001691911804539791,
+      "loss": 7.9549812316894535,
+      "step": 5690
+    },
+    {
+      "epoch": 6.485055508112724,
+      "grad_norm": 6.4585490226745605,
+      "learning_rate": 0.00016913700633837154,
+      "loss": 8.623312377929688,
+      "step": 5700
+    },
+    {
+      "epoch": 6.496441787645887,
+      "grad_norm": 6.330554485321045,
+      "learning_rate": 0.000169082832222764,
+      "loss": 8.541600036621094,
+      "step": 5710
+    },
+    {
+      "epoch": 6.507828067179049,
+      "grad_norm": 6.698922634124756,
+      "learning_rate": 0.0001690286581071564,
+      "loss": 8.024588012695313,
+      "step": 5720
+    },
+    {
+      "epoch": 6.519214346712212,
+      "grad_norm": 6.28864049911499,
+      "learning_rate": 0.00016897448399154884,
+      "loss": 8.494007110595703,
+      "step": 5730
+    },
+    {
+      "epoch": 6.530600626245374,
+      "grad_norm": 6.495082378387451,
+      "learning_rate": 0.00016892030987594126,
+      "loss": 8.174886322021484,
+      "step": 5740
+    },
+    {
+      "epoch": 6.541986905778537,
+      "grad_norm": 6.506287097930908,
+      "learning_rate": 0.0001688661357603337,
+      "loss": 8.86504898071289,
+      "step": 5750
+    },
+    {
+      "epoch": 6.553373185311699,
+      "grad_norm": 6.550364971160889,
+      "learning_rate": 0.00016881196164472616,
+      "loss": 8.222732543945312,
+      "step": 5760
+    },
+    {
+      "epoch": 6.564759464844862,
+      "grad_norm": 6.110886096954346,
+      "learning_rate": 0.0001687577875291186,
+      "loss": 7.874899291992188,
+      "step": 5770
+    },
+    {
+      "epoch": 6.5761457443780245,
+      "grad_norm": 5.996333122253418,
+      "learning_rate": 0.00016870361341351104,
+      "loss": 8.523206329345703,
+      "step": 5780
+    },
+    {
+      "epoch": 6.587532023911187,
+      "grad_norm": 6.117859840393066,
+      "learning_rate": 0.00016864943929790346,
+      "loss": 8.307976531982423,
+      "step": 5790
+    },
+    {
+      "epoch": 6.59891830344435,
+      "grad_norm": 6.8809404373168945,
+      "learning_rate": 0.00016859526518229592,
+      "loss": 8.540397644042969,
+      "step": 5800
+    },
+    {
+      "epoch": 6.610304582977512,
+      "grad_norm": 6.469377040863037,
+      "learning_rate": 0.00016854109106668834,
+      "loss": 8.485488128662109,
+      "step": 5810
+    },
+    {
+      "epoch": 6.6216908625106745,
+      "grad_norm": 6.278227806091309,
+      "learning_rate": 0.0001684869169510808,
+      "loss": 8.363449096679688,
+      "step": 5820
+    },
+    {
+      "epoch": 6.633077142043837,
+      "grad_norm": 6.4080586433410645,
+      "learning_rate": 0.00016843274283547322,
+      "loss": 8.468556213378907,
+      "step": 5830
+    },
+    {
+      "epoch": 6.644463421577,
+      "grad_norm": 5.922115802764893,
+      "learning_rate": 0.00016837856871986567,
+      "loss": 8.429467010498048,
+      "step": 5840
+    },
+    {
+      "epoch": 6.655849701110162,
+      "grad_norm": 6.8086018562316895,
+      "learning_rate": 0.0001683243946042581,
+      "loss": 8.133071136474609,
+      "step": 5850
+    },
+    {
+      "epoch": 6.667235980643325,
+      "grad_norm": 6.547855377197266,
+      "learning_rate": 0.00016827022048865054,
+      "loss": 7.84344253540039,
+      "step": 5860
+    },
+    {
+      "epoch": 6.678622260176487,
+      "grad_norm": 6.792558193206787,
+      "learning_rate": 0.00016821604637304297,
+      "loss": 8.842105102539062,
+      "step": 5870
+    },
+    {
+      "epoch": 6.69000853970965,
+      "grad_norm": 6.492033004760742,
+      "learning_rate": 0.0001681618722574354,
+      "loss": 8.356997680664062,
+      "step": 5880
+    },
+    {
+      "epoch": 6.701394819242813,
+      "grad_norm": 6.224930763244629,
+      "learning_rate": 0.00016810769814182784,
+      "loss": 8.31610565185547,
+      "step": 5890
+    },
+    {
+      "epoch": 6.712781098775975,
+      "grad_norm": 6.407049655914307,
+      "learning_rate": 0.00016805352402622027,
+      "loss": 8.490274047851562,
+      "step": 5900
+    },
+    {
+      "epoch": 6.724167378309137,
+      "grad_norm": 6.361656665802002,
+      "learning_rate": 0.00016799934991061272,
+      "loss": 8.183870697021485,
+      "step": 5910
+    },
+    {
+      "epoch": 6.7355536578423,
+      "grad_norm": 6.558957576751709,
+      "learning_rate": 0.00016794517579500515,
+      "loss": 8.488135528564452,
+      "step": 5920
+    },
+    {
+      "epoch": 6.746939937375463,
+      "grad_norm": 7.5453925132751465,
+      "learning_rate": 0.0001678910016793976,
+      "loss": 8.479955291748047,
+      "step": 5930
+    },
+    {
+      "epoch": 6.758326216908625,
+      "grad_norm": 6.576931476593018,
+      "learning_rate": 0.00016783682756379002,
+      "loss": 8.354127502441406,
+      "step": 5940
+    },
+    {
+      "epoch": 6.769712496441787,
+      "grad_norm": 6.558088302612305,
+      "learning_rate": 0.00016778265344818247,
+      "loss": 8.517078399658203,
+      "step": 5950
+    },
+    {
+      "epoch": 6.78109877597495,
+      "grad_norm": 6.338796615600586,
+      "learning_rate": 0.0001677284793325749,
+      "loss": 8.582713317871093,
+      "step": 5960
+    },
+    {
+      "epoch": 6.792485055508113,
+      "grad_norm": 5.990917205810547,
+      "learning_rate": 0.00016767430521696735,
+      "loss": 8.49090347290039,
+      "step": 5970
+    },
+    {
+      "epoch": 6.803871335041276,
+      "grad_norm": 5.543584823608398,
+      "learning_rate": 0.00016762013110135977,
+      "loss": 8.585533905029298,
+      "step": 5980
+    },
+    {
+      "epoch": 6.8152576145744375,
+      "grad_norm": 6.125088214874268,
+      "learning_rate": 0.00016756595698575223,
+      "loss": 8.365689086914063,
+      "step": 5990
+    },
+    {
+      "epoch": 6.8266438941076,
+      "grad_norm": 6.520383358001709,
+      "learning_rate": 0.00016751178287014465,
+      "loss": 7.997092437744141,
+      "step": 6000
+    },
+    {
+      "epoch": 6.838030173640763,
+      "grad_norm": 6.5187530517578125,
+      "learning_rate": 0.00016745760875453707,
+      "loss": 8.529048919677734,
+      "step": 6010
+    },
+    {
+      "epoch": 6.849416453173926,
+      "grad_norm": 6.216691493988037,
+      "learning_rate": 0.00016740343463892953,
+      "loss": 8.508572387695313,
+      "step": 6020
+    },
+    {
+      "epoch": 6.8608027327070875,
+      "grad_norm": 6.541543960571289,
+      "learning_rate": 0.00016734926052332195,
+      "loss": 8.271363830566406,
+      "step": 6030
+    },
+    {
+      "epoch": 6.87218901224025,
+      "grad_norm": 6.519371509552002,
+      "learning_rate": 0.0001672950864077144,
+      "loss": 8.481680297851563,
+      "step": 6040
+    },
+    {
+      "epoch": 6.883575291773413,
+      "grad_norm": 6.069828510284424,
+      "learning_rate": 0.00016724091229210683,
+      "loss": 8.080921936035157,
+      "step": 6050
+    },
+    {
+      "epoch": 6.894961571306576,
+      "grad_norm": 6.756743907928467,
+      "learning_rate": 0.00016718673817649928,
+      "loss": 8.416082763671875,
+      "step": 6060
+    },
+    {
+      "epoch": 6.9063478508397385,
+      "grad_norm": 6.148792266845703,
+      "learning_rate": 0.0001671325640608917,
+      "loss": 8.236578369140625,
+      "step": 6070
+    },
+    {
+      "epoch": 6.9177341303729,
+      "grad_norm": 6.631039619445801,
+      "learning_rate": 0.00016707838994528415,
+      "loss": 8.468033599853516,
+      "step": 6080
+    },
+    {
+      "epoch": 6.929120409906063,
+      "grad_norm": 7.002105236053467,
+      "learning_rate": 0.00016702421582967658,
+      "loss": 8.596262359619141,
+      "step": 6090
+    },
+    {
+      "epoch": 6.940506689439226,
+      "grad_norm": 6.974485397338867,
+      "learning_rate": 0.00016697004171406903,
+      "loss": 8.201548767089843,
+      "step": 6100
+    },
+    {
+      "epoch": 6.9518929689723885,
+      "grad_norm": 6.358666896820068,
+      "learning_rate": 0.00016691586759846148,
+      "loss": 8.327249908447266,
+      "step": 6110
+    },
+    {
+      "epoch": 6.96327924850555,
+      "grad_norm": 6.620755195617676,
+      "learning_rate": 0.00016686711089441465,
+      "loss": 8.569712829589843,
+      "step": 6120
+    },
+    {
+      "epoch": 6.974665528038713,
+      "grad_norm": 6.0767059326171875,
+      "learning_rate": 0.0001668129367788071,
+      "loss": 8.352509307861329,
+      "step": 6130
+    },
+    {
+      "epoch": 6.986051807571876,
+      "grad_norm": 6.384547233581543,
+      "learning_rate": 0.00016675876266319955,
+      "loss": 8.438029479980468,
+      "step": 6140
+    },
+    {
+      "epoch": 6.997438087105039,
+      "grad_norm": 5.849452972412109,
+      "learning_rate": 0.00016670458854759198,
+      "loss": 8.576494598388672,
+      "step": 6150
+    },
+    {
+      "epoch": 7.007970395673214,
+      "grad_norm": 6.016312122344971,
+      "learning_rate": 0.0001666504144319844,
+      "loss": 7.190988922119141,
+      "step": 6160
+    },
+    {
+      "epoch": 7.019356675206376,
+      "grad_norm": 7.674123764038086,
+      "learning_rate": 0.00016659624031637683,
+      "loss": 7.777459716796875,
+      "step": 6170
+    },
+    {
+      "epoch": 7.030742954739539,
+      "grad_norm": 5.954489707946777,
+      "learning_rate": 0.00016654206620076928,
+      "loss": 8.038485717773437,
+      "step": 6180
+    },
+    {
+      "epoch": 7.042129234272702,
+      "grad_norm": 6.098645210266113,
+      "learning_rate": 0.0001664878920851617,
+      "loss": 7.921611022949219,
+      "step": 6190
+    },
+    {
+      "epoch": 7.053515513805864,
+      "grad_norm": 6.347640514373779,
+      "learning_rate": 0.00016643371796955415,
+      "loss": 7.475108337402344,
+      "step": 6200
+    },
+    {
+      "epoch": 7.064901793339026,
+      "grad_norm": 6.295144081115723,
+      "learning_rate": 0.00016637954385394658,
+      "loss": 7.842532348632813,
+      "step": 6210
+    },
+    {
+      "epoch": 7.076288072872189,
+      "grad_norm": 6.518668174743652,
+      "learning_rate": 0.00016632536973833903,
+      "loss": 8.195176696777343,
+      "step": 6220
+    },
+    {
+      "epoch": 7.087674352405352,
+      "grad_norm": 6.452752590179443,
+      "learning_rate": 0.00016627119562273145,
+      "loss": 7.876513671875,
+      "step": 6230
+    },
+    {
+      "epoch": 7.099060631938514,
+      "grad_norm": 6.639153003692627,
+      "learning_rate": 0.0001662170215071239,
+      "loss": 7.704139709472656,
+      "step": 6240
+    },
+    {
+      "epoch": 7.110446911471676,
+      "grad_norm": 6.341755390167236,
+      "learning_rate": 0.00016616284739151636,
+      "loss": 7.5064857482910154,
+      "step": 6250
+    },
+    {
+      "epoch": 7.121833191004839,
+      "grad_norm": 6.495119094848633,
+      "learning_rate": 0.00016610867327590878,
+      "loss": 7.757982635498047,
+      "step": 6260
+    },
+    {
+      "epoch": 7.133219470538002,
+      "grad_norm": 5.995544910430908,
+      "learning_rate": 0.00016605449916030123,
+      "loss": 7.433665466308594,
+      "step": 6270
+    },
+    {
+      "epoch": 7.144605750071165,
+      "grad_norm": 5.881967067718506,
+      "learning_rate": 0.00016600032504469366,
+      "loss": 7.579911804199218,
+      "step": 6280
+    },
+    {
+      "epoch": 7.1559920296043265,
+      "grad_norm": 6.570091724395752,
+      "learning_rate": 0.0001659461509290861,
+      "loss": 7.659120178222656,
+      "step": 6290
+    },
+    {
+      "epoch": 7.167378309137489,
+      "grad_norm": 6.527503490447998,
+      "learning_rate": 0.00016589197681347853,
+      "loss": 7.960893249511718,
+      "step": 6300
+    },
+    {
+      "epoch": 7.178764588670652,
+      "grad_norm": 6.29287576675415,
+      "learning_rate": 0.00016583780269787096,
+      "loss": 7.749033355712891,
+      "step": 6310
+    },
+    {
+      "epoch": 7.190150868203815,
+      "grad_norm": 6.531010627746582,
+      "learning_rate": 0.00016578362858226338,
+      "loss": 7.7865547180175785,
+      "step": 6320
+    },
+    {
+      "epoch": 7.2015371477369765,
+      "grad_norm": 6.445150852203369,
+      "learning_rate": 0.00016572945446665583,
+      "loss": 7.650438690185547,
+      "step": 6330
+    },
+    {
+      "epoch": 7.212923427270139,
+      "grad_norm": 6.2929816246032715,
+      "learning_rate": 0.00016567528035104826,
+      "loss": 7.579685974121094,
+      "step": 6340
+    },
+    {
+      "epoch": 7.224309706803302,
+      "grad_norm": 7.012646675109863,
+      "learning_rate": 0.0001656211062354407,
+      "loss": 7.769477844238281,
+      "step": 6350
+    },
+    {
+      "epoch": 7.235695986336465,
+      "grad_norm": 6.118347644805908,
+      "learning_rate": 0.00016556693211983316,
+      "loss": 7.643257904052734,
+      "step": 6360
+    },
+    {
+      "epoch": 7.2470822658696274,
+      "grad_norm": 5.888443946838379,
+      "learning_rate": 0.0001655127580042256,
+      "loss": 7.476705169677734,
+      "step": 6370
+    },
+    {
+      "epoch": 7.258468545402789,
+      "grad_norm": 6.040239334106445,
+      "learning_rate": 0.00016545858388861804,
+      "loss": 7.228121948242188,
+      "step": 6380
+    },
+    {
+      "epoch": 7.269854824935952,
+      "grad_norm": 6.494739055633545,
+      "learning_rate": 0.00016540440977301046,
+      "loss": 7.939952087402344,
+      "step": 6390
+    },
+    {
+      "epoch": 7.281241104469115,
+      "grad_norm": 6.026639461517334,
+      "learning_rate": 0.00016535023565740291,
+      "loss": 7.657688903808594,
+      "step": 6400
+    },
+    {
+      "epoch": 7.2926273840022775,
+      "grad_norm": 6.798668384552002,
+      "learning_rate": 0.00016529606154179534,
+      "loss": 8.28334732055664,
+      "step": 6410
+    },
+    {
+      "epoch": 7.304013663535439,
+      "grad_norm": 6.844184398651123,
+      "learning_rate": 0.0001652418874261878,
+      "loss": 7.631706237792969,
+      "step": 6420
+    },
+    {
+      "epoch": 7.315399943068602,
+      "grad_norm": 5.855307579040527,
+      "learning_rate": 0.00016518771331058022,
+      "loss": 7.990458679199219,
+      "step": 6430
+    },
+    {
+      "epoch": 7.326786222601765,
+      "grad_norm": 6.2944159507751465,
+      "learning_rate": 0.00016513353919497264,
+      "loss": 7.7971031188964846,
+      "step": 6440
+    },
+    {
+      "epoch": 7.3381725021349276,
+      "grad_norm": 5.8472900390625,
+      "learning_rate": 0.0001650793650793651,
+      "loss": 7.843349456787109,
+      "step": 6450
+    },
+    {
+      "epoch": 7.34955878166809,
+      "grad_norm": 6.461435317993164,
+      "learning_rate": 0.00016502519096375752,
+      "loss": 7.716208648681641,
+      "step": 6460
+    },
+    {
+      "epoch": 7.360945061201252,
+      "grad_norm": 7.011518955230713,
+      "learning_rate": 0.00016497101684814997,
+      "loss": 7.972675323486328,
+      "step": 6470
+    },
+    {
+      "epoch": 7.372331340734415,
+      "grad_norm": 6.545440196990967,
+      "learning_rate": 0.0001649168427325424,
+      "loss": 8.097943878173828,
+      "step": 6480
+    },
+    {
+      "epoch": 7.383717620267578,
+      "grad_norm": 6.5945820808410645,
+      "learning_rate": 0.00016486266861693484,
+      "loss": 8.236820220947266,
+      "step": 6490
+    },
+    {
+      "epoch": 7.39510389980074,
+      "grad_norm": 5.519364833831787,
+      "learning_rate": 0.00016480849450132727,
+      "loss": 7.836220550537109,
+      "step": 6500
+    },
+    {
+      "epoch": 7.406490179333902,
+      "grad_norm": 6.1039910316467285,
+      "learning_rate": 0.00016475432038571972,
+      "loss": 7.595920562744141,
+      "step": 6510
+    },
+    {
+      "epoch": 7.417876458867065,
+      "grad_norm": 6.2389020919799805,
+      "learning_rate": 0.00016470014627011214,
+      "loss": 7.607307434082031,
+      "step": 6520
+    },
+    {
+      "epoch": 7.429262738400228,
+      "grad_norm": 6.249361038208008,
+      "learning_rate": 0.0001646459721545046,
+      "loss": 7.690441131591797,
+      "step": 6530
+    },
+    {
+      "epoch": 7.44064901793339,
+      "grad_norm": 6.085043907165527,
+      "learning_rate": 0.00016459179803889702,
+      "loss": 7.78430404663086,
+      "step": 6540
+    },
+    {
+      "epoch": 7.452035297466553,
+      "grad_norm": 6.791484355926514,
+      "learning_rate": 0.00016453762392328947,
+      "loss": 8.073173522949219,
+      "step": 6550
+    },
+    {
+      "epoch": 7.463421576999715,
+      "grad_norm": 6.072941303253174,
+      "learning_rate": 0.0001644834498076819,
+      "loss": 7.5049797058105465,
+      "step": 6560
+    },
+    {
+      "epoch": 7.474807856532878,
+      "grad_norm": 6.538127422332764,
+      "learning_rate": 0.00016442927569207435,
+      "loss": 8.0398193359375,
+      "step": 6570
+    },
+    {
+      "epoch": 7.4861941360660404,
+      "grad_norm": 6.887174129486084,
+      "learning_rate": 0.00016437510157646677,
+      "loss": 7.953376007080078,
+      "step": 6580
+    },
+    {
+      "epoch": 7.497580415599203,
+      "grad_norm": 6.870800495147705,
+      "learning_rate": 0.0001643209274608592,
+      "loss": 7.675375366210938,
+      "step": 6590
+    },
+    {
+      "epoch": 7.508966695132365,
+      "grad_norm": 6.101712703704834,
+      "learning_rate": 0.00016426675334525165,
+      "loss": 7.8550971984863285,
+      "step": 6600
+    },
+    {
+      "epoch": 7.520352974665528,
+      "grad_norm": 6.480758190155029,
+      "learning_rate": 0.00016421257922964407,
+      "loss": 7.9089508056640625,
+      "step": 6610
+    },
+    {
+      "epoch": 7.5317392541986905,
+      "grad_norm": 6.663425445556641,
+      "learning_rate": 0.00016415840511403652,
+      "loss": 7.739738464355469,
+      "step": 6620
+    },
+    {
+      "epoch": 7.543125533731853,
+      "grad_norm": 6.832794666290283,
+      "learning_rate": 0.00016410423099842895,
+      "loss": 7.970989990234375,
+      "step": 6630
+    },
+    {
+      "epoch": 7.554511813265016,
+      "grad_norm": 6.2909770011901855,
+      "learning_rate": 0.0001640500568828214,
+      "loss": 8.1257080078125,
+      "step": 6640
+    },
+    {
+      "epoch": 7.565898092798178,
+      "grad_norm": 6.659548759460449,
+      "learning_rate": 0.00016399588276721382,
+      "loss": 7.914602661132813,
+      "step": 6650
+    },
+    {
+      "epoch": 7.577284372331341,
+      "grad_norm": 6.749256610870361,
+      "learning_rate": 0.00016394170865160628,
+      "loss": 7.901271057128906,
+      "step": 6660
+    },
+    {
+      "epoch": 7.588670651864503,
+      "grad_norm": 6.320638656616211,
+      "learning_rate": 0.0001638875345359987,
+      "loss": 7.549610137939453,
+      "step": 6670
+    },
+    {
+      "epoch": 7.600056931397666,
+      "grad_norm": 6.600435256958008,
+      "learning_rate": 0.00016383336042039115,
+      "loss": 7.900602722167969,
+      "step": 6680
+    },
+    {
+      "epoch": 7.611443210930828,
+      "grad_norm": 7.011569976806641,
+      "learning_rate": 0.00016377918630478358,
+      "loss": 7.672644805908203,
+      "step": 6690
+    },
+    {
+      "epoch": 7.622829490463991,
+      "grad_norm": 8.137212753295898,
+      "learning_rate": 0.00016372501218917603,
+      "loss": 7.771617126464844,
+      "step": 6700
+    },
+    {
+      "epoch": 7.634215769997153,
+      "grad_norm": 6.862710952758789,
+      "learning_rate": 0.00016367083807356848,
+      "loss": 8.049059295654297,
+      "step": 6710
+    },
+    {
+      "epoch": 7.645602049530316,
+      "grad_norm": 6.581108093261719,
+      "learning_rate": 0.0001636166639579609,
+      "loss": 7.8606727600097654,
+      "step": 6720
+    },
+    {
+      "epoch": 7.656988329063479,
+      "grad_norm": 6.7956647872924805,
+      "learning_rate": 0.00016356248984235333,
+      "loss": 8.168240356445313,
+      "step": 6730
+    },
+    {
+      "epoch": 7.668374608596641,
+      "grad_norm": 6.123491287231445,
+      "learning_rate": 0.00016350831572674575,
+      "loss": 7.896241760253906,
+      "step": 6740
+    },
+    {
+      "epoch": 7.679760888129803,
+      "grad_norm": 6.9219136238098145,
+      "learning_rate": 0.0001634541416111382,
+      "loss": 8.173204040527343,
+      "step": 6750
+    },
+    {
+      "epoch": 7.691147167662966,
+      "grad_norm": 6.664566993713379,
+      "learning_rate": 0.00016339996749553063,
+      "loss": 8.276904296875,
+      "step": 6760
+    },
+    {
+      "epoch": 7.702533447196129,
+      "grad_norm": 7.015839576721191,
+      "learning_rate": 0.00016334579337992308,
+      "loss": 7.637835693359375,
+      "step": 6770
+    },
+    {
+      "epoch": 7.713919726729291,
+      "grad_norm": 7.610071659088135,
+      "learning_rate": 0.0001632916192643155,
+      "loss": 7.590660095214844,
+      "step": 6780
+    },
+    {
+      "epoch": 7.7253060062624535,
+      "grad_norm": 6.450679779052734,
+      "learning_rate": 0.00016323744514870796,
+      "loss": 7.862370300292969,
+      "step": 6790
+    },
+    {
+      "epoch": 7.736692285795616,
+      "grad_norm": 6.770766735076904,
+      "learning_rate": 0.00016318327103310038,
+      "loss": 7.959809112548828,
+      "step": 6800
+    },
+    {
+      "epoch": 7.748078565328779,
+      "grad_norm": 6.543522357940674,
+      "learning_rate": 0.00016312909691749283,
+      "loss": 7.892201232910156,
+      "step": 6810
+    },
+    {
+      "epoch": 7.759464844861942,
+      "grad_norm": 6.228302955627441,
+      "learning_rate": 0.00016307492280188528,
+      "loss": 7.535732269287109,
+      "step": 6820
+    },
+    {
+      "epoch": 7.770851124395104,
+      "grad_norm": 6.860931396484375,
+      "learning_rate": 0.0001630207486862777,
+      "loss": 8.195712280273437,
+      "step": 6830
+    },
+    {
+      "epoch": 7.782237403928266,
+      "grad_norm": 6.871212482452393,
+      "learning_rate": 0.00016296657457067016,
+      "loss": 7.680415344238281,
+      "step": 6840
+    },
+    {
+      "epoch": 7.793623683461429,
+      "grad_norm": 6.185060977935791,
+      "learning_rate": 0.00016291240045506259,
+      "loss": 7.795463562011719,
+      "step": 6850
+    },
+    {
+      "epoch": 7.805009962994592,
+      "grad_norm": 6.6304826736450195,
+      "learning_rate": 0.000162858226339455,
+      "loss": 7.885947418212891,
+      "step": 6860
+    },
+    {
+      "epoch": 7.816396242527754,
+      "grad_norm": 5.791022777557373,
+      "learning_rate": 0.00016280405222384743,
+      "loss": 7.429241943359375,
+      "step": 6870
+    },
+    {
+      "epoch": 7.827782522060916,
+      "grad_norm": 6.679376602172852,
+      "learning_rate": 0.00016274987810823989,
+      "loss": 7.767063140869141,
+      "step": 6880
+    },
+    {
+      "epoch": 7.839168801594079,
+      "grad_norm": 6.607264518737793,
+      "learning_rate": 0.0001626957039926323,
+      "loss": 8.051114654541015,
+      "step": 6890
+    },
+    {
+      "epoch": 7.850555081127242,
+      "grad_norm": 6.548991680145264,
+      "learning_rate": 0.00016264152987702476,
+      "loss": 7.440624237060547,
+      "step": 6900
+    },
+    {
+      "epoch": 7.8619413606604045,
+      "grad_norm": 6.445249557495117,
+      "learning_rate": 0.00016258735576141719,
+      "loss": 7.984051513671875,
+      "step": 6910
+    },
+    {
+      "epoch": 7.873327640193567,
+      "grad_norm": 6.627695560455322,
+      "learning_rate": 0.00016253318164580964,
+      "loss": 7.8278968811035154,
+      "step": 6920
+    },
+    {
+      "epoch": 7.884713919726729,
+      "grad_norm": 6.76314640045166,
+      "learning_rate": 0.0001624790075302021,
+      "loss": 7.394937133789062,
+      "step": 6930
+    },
+    {
+      "epoch": 7.896100199259892,
+      "grad_norm": 6.663532257080078,
+      "learning_rate": 0.00016242483341459451,
+      "loss": 8.056930541992188,
+      "step": 6940
+    },
+    {
+      "epoch": 7.9074864787930546,
+      "grad_norm": 7.132768630981445,
+      "learning_rate": 0.00016237065929898697,
+      "loss": 7.636945343017578,
+      "step": 6950
+    },
+    {
+      "epoch": 7.918872758326217,
+      "grad_norm": 6.137714385986328,
+      "learning_rate": 0.0001623164851833794,
+      "loss": 8.028910827636718,
+      "step": 6960
+    },
+    {
+      "epoch": 7.930259037859379,
+      "grad_norm": 6.413361072540283,
+      "learning_rate": 0.00016226231106777184,
+      "loss": 7.650550842285156,
+      "step": 6970
+    },
+    {
+      "epoch": 7.941645317392542,
+      "grad_norm": 5.806910514831543,
+      "learning_rate": 0.00016220813695216427,
+      "loss": 7.590312194824219,
+      "step": 6980
+    },
+    {
+      "epoch": 7.953031596925705,
+      "grad_norm": 6.411592483520508,
+      "learning_rate": 0.00016215396283655672,
+      "loss": 8.053181457519532,
+      "step": 6990
+    },
+    {
+      "epoch": 7.964417876458867,
+      "grad_norm": 6.627279758453369,
+      "learning_rate": 0.00016209978872094914,
+      "loss": 7.878144073486328,
+      "step": 7000
+    },
+    {
+      "epoch": 7.97580415599203,
+      "grad_norm": 7.155358791351318,
+      "learning_rate": 0.00016204561460534157,
+      "loss": 7.987397003173828,
+      "step": 7010
+    },
+    {
+      "epoch": 7.987190435525192,
+      "grad_norm": 6.340551376342773,
+      "learning_rate": 0.000161991440489734,
+      "loss": 7.976909637451172,
+      "step": 7020
+    },
+    {
+      "epoch": 7.998576715058355,
+      "grad_norm": 6.796611309051514,
+      "learning_rate": 0.00016193726637412644,
+      "loss": 7.7033638000488285,
+      "step": 7030
+    },
+    {
+      "epoch": 8.00910902362653,
+      "grad_norm": 6.260311126708984,
+      "learning_rate": 0.0001618830922585189,
+      "loss": 6.682421112060547,
+      "step": 7040
+    },
+    {
+      "epoch": 8.020495303159693,
+      "grad_norm": 7.427417278289795,
+      "learning_rate": 0.00016182891814291132,
+      "loss": 7.5024055480957035,
+      "step": 7050
+    },
+    {
+      "epoch": 8.031881582692854,
+      "grad_norm": 6.3494486808776855,
+      "learning_rate": 0.00016177474402730377,
+      "loss": 7.2604820251464846,
+      "step": 7060
+    },
+    {
+      "epoch": 8.043267862226017,
+      "grad_norm": 6.655555725097656,
+      "learning_rate": 0.0001617205699116962,
+      "loss": 7.514097595214844,
+      "step": 7070
+    },
+    {
+      "epoch": 8.05465414175918,
+      "grad_norm": 6.684828758239746,
+      "learning_rate": 0.00016166639579608865,
+      "loss": 7.446005249023438,
+      "step": 7080
+    },
+    {
+      "epoch": 8.066040421292342,
+      "grad_norm": 6.890992164611816,
+      "learning_rate": 0.00016161222168048107,
+      "loss": 7.443931579589844,
+      "step": 7090
+    },
+    {
+      "epoch": 8.077426700825505,
+      "grad_norm": 6.673741340637207,
+      "learning_rate": 0.00016155804756487352,
+      "loss": 7.16052017211914,
+      "step": 7100
+    },
+    {
+      "epoch": 8.088812980358668,
+      "grad_norm": 6.923252105712891,
+      "learning_rate": 0.00016150387344926595,
+      "loss": 6.938526153564453,
+      "step": 7110
+    },
+    {
+      "epoch": 8.10019925989183,
+      "grad_norm": 6.302939414978027,
+      "learning_rate": 0.0001614496993336584,
+      "loss": 7.243843841552734,
+      "step": 7120
+    },
+    {
+      "epoch": 8.111585539424993,
+      "grad_norm": 6.899267673492432,
+      "learning_rate": 0.00016139552521805082,
+      "loss": 7.297703552246094,
+      "step": 7130
+    },
+    {
+      "epoch": 8.122971818958156,
+      "grad_norm": 7.268035888671875,
+      "learning_rate": 0.00016134135110244327,
+      "loss": 7.230997467041016,
+      "step": 7140
+    },
+    {
+      "epoch": 8.134358098491319,
+      "grad_norm": 7.285719871520996,
+      "learning_rate": 0.00016129259439839647,
+      "loss": 7.024531555175781,
+      "step": 7150
+    },
+    {
+      "epoch": 8.14574437802448,
+      "grad_norm": 6.868885040283203,
+      "learning_rate": 0.0001612384202827889,
+      "loss": 7.714209747314453,
+      "step": 7160
+    },
+    {
+      "epoch": 8.157130657557643,
+      "grad_norm": 6.319274425506592,
+      "learning_rate": 0.00016118424616718132,
+      "loss": 6.995166015625,
+      "step": 7170
+    },
+    {
+      "epoch": 8.168516937090805,
+      "grad_norm": 6.736001491546631,
+      "learning_rate": 0.00016113007205157374,
+      "loss": 7.084954833984375,
+      "step": 7180
+    },
+    {
+      "epoch": 8.179903216623968,
+      "grad_norm": 6.393606185913086,
+      "learning_rate": 0.0001610758979359662,
+      "loss": 7.56659164428711,
+      "step": 7190
+    },
+    {
+      "epoch": 8.19128949615713,
+      "grad_norm": 6.33527135848999,
+      "learning_rate": 0.00016102172382035865,
+      "loss": 7.253793334960937,
+      "step": 7200
+    },
+    {
+      "epoch": 8.202675775690293,
+      "grad_norm": 7.905284404754639,
+      "learning_rate": 0.00016096754970475107,
+      "loss": 7.590048217773438,
+      "step": 7210
+    },
+    {
+      "epoch": 8.214062055223456,
+      "grad_norm": 6.864686965942383,
+      "learning_rate": 0.00016091337558914352,
+      "loss": 7.1065208435058596,
+      "step": 7220
+    },
+    {
+      "epoch": 8.225448334756619,
+      "grad_norm": 6.355857849121094,
+      "learning_rate": 0.00016085920147353595,
+      "loss": 7.114351654052735,
+      "step": 7230
+    },
+    {
+      "epoch": 8.23683461428978,
+      "grad_norm": 7.217507839202881,
+      "learning_rate": 0.0001608050273579284,
+      "loss": 7.072686767578125,
+      "step": 7240
+    },
+    {
+      "epoch": 8.248220893822943,
+      "grad_norm": 6.846201419830322,
+      "learning_rate": 0.00016075085324232082,
+      "loss": 7.156787109375,
+      "step": 7250
+    },
+    {
+      "epoch": 8.259607173356105,
+      "grad_norm": 7.219080448150635,
+      "learning_rate": 0.00016069667912671327,
+      "loss": 7.439443969726563,
+      "step": 7260
+    },
+    {
+      "epoch": 8.270993452889268,
+      "grad_norm": 6.303773880004883,
+      "learning_rate": 0.0001606425050111057,
+      "loss": 7.361647033691407,
+      "step": 7270
+    },
+    {
+      "epoch": 8.28237973242243,
+      "grad_norm": 7.611039161682129,
+      "learning_rate": 0.00016058833089549815,
+      "loss": 7.37061767578125,
+      "step": 7280
+    },
+    {
+      "epoch": 8.293766011955594,
+      "grad_norm": 6.5654497146606445,
+      "learning_rate": 0.00016053415677989057,
+      "loss": 7.721343994140625,
+      "step": 7290
+    },
+    {
+      "epoch": 8.305152291488756,
+      "grad_norm": 5.921697616577148,
+      "learning_rate": 0.000160479982664283,
+      "loss": 7.14508285522461,
+      "step": 7300
+    },
+    {
+      "epoch": 8.316538571021919,
+      "grad_norm": 6.788980484008789,
+      "learning_rate": 0.00016042580854867545,
+      "loss": 7.477375793457031,
+      "step": 7310
+    },
+    {
+      "epoch": 8.327924850555082,
+      "grad_norm": 7.130054950714111,
+      "learning_rate": 0.00016037163443306788,
+      "loss": 6.889068603515625,
+      "step": 7320
+    },
+    {
+      "epoch": 8.339311130088245,
+      "grad_norm": 6.373105049133301,
+      "learning_rate": 0.00016031746031746033,
+      "loss": 7.234850311279297,
+      "step": 7330
+    },
+    {
+      "epoch": 8.350697409621405,
+      "grad_norm": 7.498371124267578,
+      "learning_rate": 0.00016026328620185275,
+      "loss": 7.064684295654297,
+      "step": 7340
+    },
+    {
+      "epoch": 8.362083689154568,
+      "grad_norm": 6.624664783477783,
+      "learning_rate": 0.0001602091120862452,
+      "loss": 7.807039642333985,
+      "step": 7350
+    },
+    {
+      "epoch": 8.373469968687731,
+      "grad_norm": 6.387814044952393,
+      "learning_rate": 0.00016015493797063763,
+      "loss": 7.460430908203125,
+      "step": 7360
+    },
+    {
+      "epoch": 8.384856248220894,
+      "grad_norm": 7.011965274810791,
+      "learning_rate": 0.00016010076385503008,
+      "loss": 7.333013916015625,
+      "step": 7370
+    },
+    {
+      "epoch": 8.396242527754056,
+      "grad_norm": 6.425861835479736,
+      "learning_rate": 0.0001600465897394225,
+      "loss": 7.393313598632813,
+      "step": 7380
+    },
+    {
+      "epoch": 8.40762880728722,
+      "grad_norm": 6.637088775634766,
+      "learning_rate": 0.00015999241562381496,
+      "loss": 7.426857757568359,
+      "step": 7390
+    },
+    {
+      "epoch": 8.419015086820382,
+      "grad_norm": 7.036196708679199,
+      "learning_rate": 0.00015993824150820738,
+      "loss": 7.142207336425781,
+      "step": 7400
+    },
+    {
+      "epoch": 8.430401366353545,
+      "grad_norm": 6.357840061187744,
+      "learning_rate": 0.00015988406739259983,
+      "loss": 7.445525360107422,
+      "step": 7410
+    },
+    {
+      "epoch": 8.441787645886706,
+      "grad_norm": 6.880340099334717,
+      "learning_rate": 0.00015982989327699228,
+      "loss": 7.163670349121094,
+      "step": 7420
+    },
+    {
+      "epoch": 8.453173925419868,
+      "grad_norm": 6.362428188323975,
+      "learning_rate": 0.0001597757191613847,
+      "loss": 7.434315490722656,
+      "step": 7430
+    },
+    {
+      "epoch": 8.464560204953031,
+      "grad_norm": 6.664401531219482,
+      "learning_rate": 0.00015972154504577713,
+      "loss": 7.144511413574219,
+      "step": 7440
+    },
+    {
+      "epoch": 8.475946484486194,
+      "grad_norm": 6.501242637634277,
+      "learning_rate": 0.00015966737093016956,
+      "loss": 7.108064270019531,
+      "step": 7450
+    },
+    {
+      "epoch": 8.487332764019357,
+      "grad_norm": 6.9143476486206055,
+      "learning_rate": 0.000159613196814562,
+      "loss": 7.1984611511230465,
+      "step": 7460
+    },
+    {
+      "epoch": 8.49871904355252,
+      "grad_norm": 6.557959079742432,
+      "learning_rate": 0.00015955902269895443,
+      "loss": 7.022095489501953,
+      "step": 7470
+    },
+    {
+      "epoch": 8.510105323085682,
+      "grad_norm": 6.557929992675781,
+      "learning_rate": 0.00015950484858334688,
+      "loss": 7.145917510986328,
+      "step": 7480
+    },
+    {
+      "epoch": 8.521491602618845,
+      "grad_norm": 6.603511810302734,
+      "learning_rate": 0.0001594506744677393,
+      "loss": 7.341641998291015,
+      "step": 7490
+    },
+    {
+      "epoch": 8.532877882152007,
+      "grad_norm": 7.191463470458984,
+      "learning_rate": 0.00015939650035213176,
+      "loss": 6.9570457458496096,
+      "step": 7500
+    },
+    {
+      "epoch": 8.54426416168517,
+      "grad_norm": 6.919197082519531,
+      "learning_rate": 0.00015934232623652418,
+      "loss": 7.235923004150391,
+      "step": 7510
+    },
+    {
+      "epoch": 8.555650441218331,
+      "grad_norm": 6.883152008056641,
+      "learning_rate": 0.00015928815212091664,
+      "loss": 7.444821166992187,
+      "step": 7520
+    },
+    {
+      "epoch": 8.567036720751494,
+      "grad_norm": 6.3207244873046875,
+      "learning_rate": 0.0001592339780053091,
+      "loss": 7.598611450195312,
+      "step": 7530
+    },
+    {
+      "epoch": 8.578423000284657,
+      "grad_norm": 6.418648719787598,
+      "learning_rate": 0.0001591798038897015,
+      "loss": 7.209278869628906,
+      "step": 7540
+    },
+    {
+      "epoch": 8.58980927981782,
+      "grad_norm": 6.839104175567627,
+      "learning_rate": 0.00015912562977409396,
+      "loss": 7.241474914550781,
+      "step": 7550
+    },
+    {
+      "epoch": 8.601195559350982,
+      "grad_norm": 6.884853363037109,
+      "learning_rate": 0.0001590714556584864,
+      "loss": 7.021162414550782,
+      "step": 7560
+    },
+    {
+      "epoch": 8.612581838884145,
+      "grad_norm": 6.1609787940979,
+      "learning_rate": 0.00015901728154287884,
+      "loss": 7.474005126953125,
+      "step": 7570
+    },
+    {
+      "epoch": 8.623968118417308,
+      "grad_norm": 6.195271968841553,
+      "learning_rate": 0.00015896310742727126,
+      "loss": 6.983114624023438,
+      "step": 7580
+    },
+    {
+      "epoch": 8.63535439795047,
+      "grad_norm": 6.70726203918457,
+      "learning_rate": 0.0001589089333116637,
+      "loss": 7.420458984375,
+      "step": 7590
+    },
+    {
+      "epoch": 8.646740677483631,
+      "grad_norm": 7.531630516052246,
+      "learning_rate": 0.0001588547591960561,
+      "loss": 7.718193817138672,
+      "step": 7600
+    },
+    {
+      "epoch": 8.658126957016794,
+      "grad_norm": 6.9179534912109375,
+      "learning_rate": 0.00015880058508044856,
+      "loss": 7.357187652587891,
+      "step": 7610
+    },
+    {
+      "epoch": 8.669513236549957,
+      "grad_norm": 6.675024509429932,
+      "learning_rate": 0.000158746410964841,
+      "loss": 7.010201263427734,
+      "step": 7620
+    },
+    {
+      "epoch": 8.68089951608312,
+      "grad_norm": 6.622598648071289,
+      "learning_rate": 0.00015869223684923344,
+      "loss": 7.162730407714844,
+      "step": 7630
+    },
+    {
+      "epoch": 8.692285795616282,
+      "grad_norm": 6.835996627807617,
+      "learning_rate": 0.0001586380627336259,
+      "loss": 7.303845977783203,
+      "step": 7640
+    },
+    {
+      "epoch": 8.703672075149445,
+      "grad_norm": 6.367660045623779,
+      "learning_rate": 0.00015858388861801832,
+      "loss": 7.625505065917968,
+      "step": 7650
+    },
+    {
+      "epoch": 8.715058354682608,
+      "grad_norm": 6.169827938079834,
+      "learning_rate": 0.00015852971450241077,
+      "loss": 7.09228744506836,
+      "step": 7660
+    },
+    {
+      "epoch": 8.72644463421577,
+      "grad_norm": 6.215500354766846,
+      "learning_rate": 0.0001584755403868032,
+      "loss": 7.060032653808594,
+      "step": 7670
+    },
+    {
+      "epoch": 8.737830913748933,
+      "grad_norm": 6.325623035430908,
+      "learning_rate": 0.00015842136627119564,
+      "loss": 7.31326904296875,
+      "step": 7680
+    },
+    {
+      "epoch": 8.749217193282096,
+      "grad_norm": 6.559711933135986,
+      "learning_rate": 0.00015836719215558807,
+      "loss": 7.409132385253907,
+      "step": 7690
+    },
+    {
+      "epoch": 8.760603472815257,
+      "grad_norm": 6.366078853607178,
+      "learning_rate": 0.00015831301803998052,
+      "loss": 6.915621948242188,
+      "step": 7700
+    },
+    {
+      "epoch": 8.77198975234842,
+      "grad_norm": 6.734594345092773,
+      "learning_rate": 0.00015825884392437294,
+      "loss": 7.271147155761719,
+      "step": 7710
+    },
+    {
+      "epoch": 8.783376031881582,
+      "grad_norm": 7.609946250915527,
+      "learning_rate": 0.00015820466980876537,
+      "loss": 7.408287048339844,
+      "step": 7720
+    },
+    {
+      "epoch": 8.794762311414745,
+      "grad_norm": 6.976548671722412,
+      "learning_rate": 0.00015815049569315782,
+      "loss": 7.269252777099609,
+      "step": 7730
+    },
+    {
+      "epoch": 8.806148590947908,
+      "grad_norm": 7.052664279937744,
+      "learning_rate": 0.00015809632157755025,
+      "loss": 7.457789611816406,
+      "step": 7740
+    },
+    {
+      "epoch": 8.81753487048107,
+      "grad_norm": 6.683445453643799,
+      "learning_rate": 0.0001580421474619427,
+      "loss": 7.239376831054687,
+      "step": 7750
+    },
+    {
+      "epoch": 8.828921150014233,
+      "grad_norm": 6.740503311157227,
+      "learning_rate": 0.00015798797334633512,
+      "loss": 7.014515686035156,
+      "step": 7760
+    },
+    {
+      "epoch": 8.840307429547396,
+      "grad_norm": 7.012024879455566,
+      "learning_rate": 0.00015793379923072757,
+      "loss": 7.423218536376953,
+      "step": 7770
+    },
+    {
+      "epoch": 8.851693709080559,
+      "grad_norm": 6.79911994934082,
+      "learning_rate": 0.00015787962511512,
+      "loss": 7.527141571044922,
+      "step": 7780
+    },
+    {
+      "epoch": 8.86307998861372,
+      "grad_norm": 6.133242130279541,
+      "learning_rate": 0.00015782545099951245,
+      "loss": 7.559020233154297,
+      "step": 7790
+    },
+    {
+      "epoch": 8.874466268146882,
+      "grad_norm": 6.954835891723633,
+      "learning_rate": 0.00015777127688390487,
+      "loss": 7.530062103271485,
+      "step": 7800
+    },
+    {
+      "epoch": 8.885852547680045,
+      "grad_norm": 6.71362829208374,
+      "learning_rate": 0.00015771710276829733,
+      "loss": 7.113396453857422,
+      "step": 7810
+    },
+    {
+      "epoch": 8.897238827213208,
+      "grad_norm": 6.9403862953186035,
+      "learning_rate": 0.00015766292865268975,
+      "loss": 7.578011322021484,
+      "step": 7820
+    },
+    {
+      "epoch": 8.90862510674637,
+      "grad_norm": 6.492265701293945,
+      "learning_rate": 0.0001576087545370822,
+      "loss": 7.283029937744141,
+      "step": 7830
+    },
+    {
+      "epoch": 8.920011386279533,
+      "grad_norm": 5.762649059295654,
+      "learning_rate": 0.00015755458042147463,
+      "loss": 7.412893676757813,
+      "step": 7840
+    },
+    {
+      "epoch": 8.931397665812696,
+      "grad_norm": 6.935257911682129,
+      "learning_rate": 0.00015750040630586708,
+      "loss": 7.5289451599121096,
+      "step": 7850
+    },
+    {
+      "epoch": 8.942783945345859,
+      "grad_norm": 6.323666095733643,
+      "learning_rate": 0.0001574462321902595,
+      "loss": 7.003956604003906,
+      "step": 7860
+    },
+    {
+      "epoch": 8.954170224879022,
+      "grad_norm": 5.6248779296875,
+      "learning_rate": 0.00015739205807465193,
+      "loss": 7.441354370117187,
+      "step": 7870
+    },
+    {
+      "epoch": 8.965556504412183,
+      "grad_norm": 6.482755661010742,
+      "learning_rate": 0.00015733788395904438,
+      "loss": 7.2617744445800785,
+      "step": 7880
+    },
+    {
+      "epoch": 8.976942783945345,
+      "grad_norm": 6.673778533935547,
+      "learning_rate": 0.0001572837098434368,
+      "loss": 7.212700653076172,
+      "step": 7890
+    },
+    {
+      "epoch": 8.988329063478508,
+      "grad_norm": 6.261040210723877,
+      "learning_rate": 0.00015722953572782925,
+      "loss": 7.402249145507812,
+      "step": 7900
+    },
+    {
+      "epoch": 8.99971534301167,
+      "grad_norm": 6.743113994598389,
+      "learning_rate": 0.00015717536161222168,
+      "loss": 7.82423095703125,
+      "step": 7910
+    },
+    {
+      "epoch": 9.010247651579846,
+      "grad_norm": 6.245288372039795,
+      "learning_rate": 0.00015712118749661413,
+      "loss": 6.120718765258789,
+      "step": 7920
+    },
+    {
+      "epoch": 9.021633931113008,
+      "grad_norm": 6.373956203460693,
+      "learning_rate": 0.00015706701338100655,
+      "loss": 6.5073394775390625,
+      "step": 7930
+    },
+    {
+      "epoch": 9.033020210646171,
+      "grad_norm": 6.782715320587158,
+      "learning_rate": 0.000157012839265399,
+      "loss": 6.719621276855468,
+      "step": 7940
+    },
+    {
+      "epoch": 9.044406490179334,
+      "grad_norm": 6.204227924346924,
+      "learning_rate": 0.00015695866514979143,
+      "loss": 6.846994781494141,
+      "step": 7950
+    },
+    {
+      "epoch": 9.055792769712497,
+      "grad_norm": 5.706500053405762,
+      "learning_rate": 0.00015690449103418388,
+      "loss": 6.545850372314453,
+      "step": 7960
+    },
+    {
+      "epoch": 9.06717904924566,
+      "grad_norm": 6.953368186950684,
+      "learning_rate": 0.0001568503169185763,
+      "loss": 6.754498291015625,
+      "step": 7970
+    },
+    {
+      "epoch": 9.078565328778822,
+      "grad_norm": 7.3118743896484375,
+      "learning_rate": 0.00015679614280296876,
+      "loss": 7.14571533203125,
+      "step": 7980
+    },
+    {
+      "epoch": 9.089951608311985,
+      "grad_norm": 7.408244609832764,
+      "learning_rate": 0.00015674196868736118,
+      "loss": 6.966554260253906,
+      "step": 7990
+    },
+    {
+      "epoch": 9.101337887845146,
+      "grad_norm": 6.550361156463623,
+      "learning_rate": 0.00015668779457175363,
+      "loss": 6.683185577392578,
+      "step": 8000
+    },
+    {
+      "epoch": 9.112724167378309,
+      "grad_norm": 7.23964262008667,
+      "learning_rate": 0.00015663362045614606,
+      "loss": 6.666042327880859,
+      "step": 8010
+    },
+    {
+      "epoch": 9.124110446911471,
+      "grad_norm": 6.447325706481934,
+      "learning_rate": 0.00015657944634053848,
+      "loss": 6.605396270751953,
+      "step": 8020
+    },
+    {
+      "epoch": 9.135496726444634,
+      "grad_norm": 6.252988338470459,
+      "learning_rate": 0.00015652527222493093,
+      "loss": 6.713479614257812,
+      "step": 8030
+    },
+    {
+      "epoch": 9.146883005977797,
+      "grad_norm": 7.02787971496582,
+      "learning_rate": 0.00015647109810932336,
+      "loss": 6.7254997253417965,
+      "step": 8040
+    },
+    {
+      "epoch": 9.15826928551096,
+      "grad_norm": 6.618356704711914,
+      "learning_rate": 0.0001564169239937158,
+      "loss": 6.731692504882813,
+      "step": 8050
+    },
+    {
+      "epoch": 9.169655565044122,
+      "grad_norm": 6.890060901641846,
+      "learning_rate": 0.00015636274987810824,
+      "loss": 6.741816711425781,
+      "step": 8060
+    },
+    {
+      "epoch": 9.181041844577285,
+      "grad_norm": 7.1875901222229,
+      "learning_rate": 0.0001563085757625007,
+      "loss": 6.789379119873047,
+      "step": 8070
+    },
+    {
+      "epoch": 9.192428124110448,
+      "grad_norm": 7.181351184844971,
+      "learning_rate": 0.0001562544016468931,
+      "loss": 6.818449401855469,
+      "step": 8080
+    },
+    {
+      "epoch": 9.203814403643609,
+      "grad_norm": 7.285808086395264,
+      "learning_rate": 0.00015620022753128556,
+      "loss": 6.539177703857422,
+      "step": 8090
+    },
+    {
+      "epoch": 9.215200683176771,
+      "grad_norm": 6.714638710021973,
+      "learning_rate": 0.00015614605341567801,
+      "loss": 6.656063842773437,
+      "step": 8100
+    },
+    {
+      "epoch": 9.226586962709934,
+      "grad_norm": 7.055619716644287,
+      "learning_rate": 0.00015609187930007044,
+      "loss": 6.620281982421875,
+      "step": 8110
+    },
+    {
+      "epoch": 9.237973242243097,
+      "grad_norm": 6.574080467224121,
+      "learning_rate": 0.0001560377051844629,
+      "loss": 6.747283172607422,
+      "step": 8120
+    },
+    {
+      "epoch": 9.24935952177626,
+      "grad_norm": 6.460296630859375,
+      "learning_rate": 0.00015598353106885531,
+      "loss": 6.993648529052734,
+      "step": 8130
+    },
+    {
+      "epoch": 9.260745801309422,
+      "grad_norm": 7.454887866973877,
+      "learning_rate": 0.00015592935695324774,
+      "loss": 6.948415374755859,
+      "step": 8140
+    },
+    {
+      "epoch": 9.272132080842585,
+      "grad_norm": 6.729802131652832,
+      "learning_rate": 0.00015587518283764016,
+      "loss": 6.89251937866211,
+      "step": 8150
+    },
+    {
+      "epoch": 9.283518360375748,
+      "grad_norm": 6.335530757904053,
+      "learning_rate": 0.00015582100872203262,
+      "loss": 6.703284454345703,
+      "step": 8160
+    },
+    {
+      "epoch": 9.29490463990891,
+      "grad_norm": 7.840336322784424,
+      "learning_rate": 0.00015576683460642504,
+      "loss": 7.09771957397461,
+      "step": 8170
+    },
+    {
+      "epoch": 9.306290919442072,
+      "grad_norm": 6.923723220825195,
+      "learning_rate": 0.0001557126604908175,
+      "loss": 6.888484954833984,
+      "step": 8180
+    },
+    {
+      "epoch": 9.317677198975234,
+      "grad_norm": 6.690946578979492,
+      "learning_rate": 0.00015565848637520992,
+      "loss": 7.289754486083984,
+      "step": 8190
+    },
+    {
+      "epoch": 9.329063478508397,
+      "grad_norm": 6.629517555236816,
+      "learning_rate": 0.00015560431225960237,
+      "loss": 7.009606170654297,
+      "step": 8200
+    },
+    {
+      "epoch": 9.34044975804156,
+      "grad_norm": 7.02864933013916,
+      "learning_rate": 0.00015555013814399482,
+      "loss": 6.7587532043457035,
+      "step": 8210
+    },
+    {
+      "epoch": 9.351836037574722,
+      "grad_norm": 6.852596282958984,
+      "learning_rate": 0.00015549596402838724,
+      "loss": 6.829173278808594,
+      "step": 8220
+    },
+    {
+      "epoch": 9.363222317107885,
+      "grad_norm": 6.9407243728637695,
+      "learning_rate": 0.0001554417899127797,
+      "loss": 6.788723754882812,
+      "step": 8230
+    },
+    {
+      "epoch": 9.374608596641048,
+      "grad_norm": 6.7164812088012695,
+      "learning_rate": 0.00015538761579717212,
+      "loss": 6.483773040771484,
+      "step": 8240
+    },
+    {
+      "epoch": 9.38599487617421,
+      "grad_norm": 6.983746528625488,
+      "learning_rate": 0.00015533344168156457,
+      "loss": 6.634527587890625,
+      "step": 8250
+    },
+    {
+      "epoch": 9.397381155707373,
+      "grad_norm": 7.202597618103027,
+      "learning_rate": 0.000155279267565957,
+      "loss": 6.860771942138672,
+      "step": 8260
+    },
+    {
+      "epoch": 9.408767435240534,
+      "grad_norm": 6.925684452056885,
+      "learning_rate": 0.00015522509345034945,
+      "loss": 7.078357696533203,
+      "step": 8270
+    },
+    {
+      "epoch": 9.420153714773697,
+      "grad_norm": 7.451114177703857,
+      "learning_rate": 0.00015517091933474187,
+      "loss": 6.643927764892578,
+      "step": 8280
+    },
+    {
+      "epoch": 9.43153999430686,
+      "grad_norm": 6.825454235076904,
+      "learning_rate": 0.0001551167452191343,
+      "loss": 6.680137634277344,
+      "step": 8290
+    },
+    {
+      "epoch": 9.442926273840023,
+      "grad_norm": 6.374398708343506,
+      "learning_rate": 0.00015506257110352672,
+      "loss": 6.5555778503417965,
+      "step": 8300
+    },
+    {
+      "epoch": 9.454312553373185,
+      "grad_norm": 7.115841388702393,
+      "learning_rate": 0.00015500839698791917,
+      "loss": 7.261045837402344,
+      "step": 8310
+    },
+    {
+      "epoch": 9.465698832906348,
+      "grad_norm": 6.24588680267334,
+      "learning_rate": 0.00015495422287231162,
+      "loss": 6.373281860351563,
+      "step": 8320
+    },
+    {
+      "epoch": 9.47708511243951,
+      "grad_norm": 6.433933258056641,
+      "learning_rate": 0.00015490004875670405,
+      "loss": 7.186137390136719,
+      "step": 8330
+    },
+    {
+      "epoch": 9.488471391972674,
+      "grad_norm": 6.373920440673828,
+      "learning_rate": 0.0001548458746410965,
+      "loss": 6.6659690856933596,
+      "step": 8340
+    },
+    {
+      "epoch": 9.499857671505836,
+      "grad_norm": 6.638841152191162,
+      "learning_rate": 0.00015479170052548892,
+      "loss": 6.709719848632813,
+      "step": 8350
+    },
+    {
+      "epoch": 9.511243951038997,
+      "grad_norm": 7.437965393066406,
+      "learning_rate": 0.00015473752640988138,
+      "loss": 6.974462890625,
+      "step": 8360
+    },
+    {
+      "epoch": 9.52263023057216,
+      "grad_norm": 7.176238059997559,
+      "learning_rate": 0.0001546833522942738,
+      "loss": 7.053614807128906,
+      "step": 8370
+    },
+    {
+      "epoch": 9.534016510105323,
+      "grad_norm": 6.885675430297852,
+      "learning_rate": 0.00015462917817866625,
+      "loss": 6.691897583007813,
+      "step": 8380
+    },
+    {
+      "epoch": 9.545402789638485,
+      "grad_norm": 5.81322717666626,
+      "learning_rate": 0.00015457500406305868,
+      "loss": 6.6090576171875,
+      "step": 8390
+    },
+    {
+      "epoch": 9.556789069171648,
+      "grad_norm": 6.862832069396973,
+      "learning_rate": 0.00015452082994745113,
+      "loss": 7.214694976806641,
+      "step": 8400
+    },
+    {
+      "epoch": 9.568175348704811,
+      "grad_norm": 7.299694538116455,
+      "learning_rate": 0.00015446665583184355,
+      "loss": 6.833685302734375,
+      "step": 8410
+    },
+    {
+      "epoch": 9.579561628237974,
+      "grad_norm": 6.393362522125244,
+      "learning_rate": 0.000154412481716236,
+      "loss": 7.033847045898438,
+      "step": 8420
+    },
+    {
+      "epoch": 9.590947907771136,
+      "grad_norm": 7.435060977935791,
+      "learning_rate": 0.00015435830760062843,
+      "loss": 6.689753723144531,
+      "step": 8430
+    },
+    {
+      "epoch": 9.602334187304299,
+      "grad_norm": 7.483737468719482,
+      "learning_rate": 0.00015430413348502085,
+      "loss": 7.141274261474609,
+      "step": 8440
+    },
+    {
+      "epoch": 9.61372046683746,
+      "grad_norm": 6.771705150604248,
+      "learning_rate": 0.0001542499593694133,
+      "loss": 6.743639373779297,
+      "step": 8450
+    },
+    {
+      "epoch": 9.625106746370623,
+      "grad_norm": 6.806461334228516,
+      "learning_rate": 0.00015419578525380573,
+      "loss": 6.662245178222657,
+      "step": 8460
+    },
+    {
+      "epoch": 9.636493025903786,
+      "grad_norm": 6.6654372215271,
+      "learning_rate": 0.00015414161113819818,
+      "loss": 6.932222747802735,
+      "step": 8470
+    },
+    {
+      "epoch": 9.647879305436948,
+      "grad_norm": 6.507349491119385,
+      "learning_rate": 0.0001540874370225906,
+      "loss": 7.092259216308594,
+      "step": 8480
+    },
+    {
+      "epoch": 9.659265584970111,
+      "grad_norm": 7.4099202156066895,
+      "learning_rate": 0.00015403326290698306,
+      "loss": 6.74954833984375,
+      "step": 8490
+    },
+    {
+      "epoch": 9.670651864503274,
+      "grad_norm": 6.947366714477539,
+      "learning_rate": 0.00015397908879137548,
+      "loss": 6.710967254638672,
+      "step": 8500
+    },
+    {
+      "epoch": 9.682038144036436,
+      "grad_norm": 7.391485214233398,
+      "learning_rate": 0.00015392491467576793,
+      "loss": 6.349737167358398,
+      "step": 8510
+    },
+    {
+      "epoch": 9.6934244235696,
+      "grad_norm": 6.834575653076172,
+      "learning_rate": 0.00015387074056016036,
+      "loss": 6.986430358886719,
+      "step": 8520
+    },
+    {
+      "epoch": 9.704810703102762,
+      "grad_norm": 6.834963321685791,
+      "learning_rate": 0.0001538165664445528,
+      "loss": 6.768232727050782,
+      "step": 8530
+    },
+    {
+      "epoch": 9.716196982635923,
+      "grad_norm": 6.545167446136475,
+      "learning_rate": 0.00015376239232894526,
+      "loss": 7.306761932373047,
+      "step": 8540
+    },
+    {
+      "epoch": 9.727583262169086,
+      "grad_norm": 6.159801959991455,
+      "learning_rate": 0.00015370821821333768,
+      "loss": 6.77575454711914,
+      "step": 8550
+    },
+    {
+      "epoch": 9.738969541702248,
+      "grad_norm": 6.571531772613525,
+      "learning_rate": 0.0001536540440977301,
+      "loss": 6.770185089111328,
+      "step": 8560
+    },
+    {
+      "epoch": 9.750355821235411,
+      "grad_norm": 7.542463779449463,
+      "learning_rate": 0.00015359986998212253,
+      "loss": 7.183005523681641,
+      "step": 8570
+    },
+    {
+      "epoch": 9.761742100768574,
+      "grad_norm": 6.521090984344482,
+      "learning_rate": 0.00015354569586651499,
+      "loss": 6.722921752929688,
+      "step": 8580
+    },
+    {
+      "epoch": 9.773128380301737,
+      "grad_norm": 6.779551029205322,
+      "learning_rate": 0.0001534915217509074,
+      "loss": 6.963375854492187,
+      "step": 8590
+    },
+    {
+      "epoch": 9.7845146598349,
+      "grad_norm": 7.098196029663086,
+      "learning_rate": 0.00015343734763529986,
+      "loss": 7.068116760253906,
+      "step": 8600
+    },
+    {
+      "epoch": 9.795900939368062,
+      "grad_norm": 6.990663528442383,
+      "learning_rate": 0.00015338317351969229,
+      "loss": 6.641743469238281,
+      "step": 8610
+    },
+    {
+      "epoch": 9.807287218901225,
+      "grad_norm": 6.753698348999023,
+      "learning_rate": 0.00015332899940408474,
+      "loss": 6.880230712890625,
+      "step": 8620
+    },
+    {
+      "epoch": 9.818673498434386,
+      "grad_norm": 7.067351341247559,
+      "learning_rate": 0.00015327482528847716,
+      "loss": 6.862619781494141,
+      "step": 8630
+    },
+    {
+      "epoch": 9.830059777967548,
+      "grad_norm": 7.574794292449951,
+      "learning_rate": 0.0001532206511728696,
+      "loss": 6.817996978759766,
+      "step": 8640
+    },
+    {
+      "epoch": 9.841446057500711,
+      "grad_norm": 6.906538963317871,
+      "learning_rate": 0.00015316647705726204,
+      "loss": 6.774374389648438,
+      "step": 8650
+    },
+    {
+      "epoch": 9.852832337033874,
+      "grad_norm": 6.665681838989258,
+      "learning_rate": 0.0001531123029416545,
+      "loss": 6.8718719482421875,
+      "step": 8660
+    },
+    {
+      "epoch": 9.864218616567037,
+      "grad_norm": 6.376792907714844,
+      "learning_rate": 0.00015305812882604694,
+      "loss": 6.672955322265625,
+      "step": 8670
+    },
+    {
+      "epoch": 9.8756048961002,
+      "grad_norm": 7.075804710388184,
+      "learning_rate": 0.00015300395471043937,
+      "loss": 7.0490264892578125,
+      "step": 8680
+    },
+    {
+      "epoch": 9.886991175633362,
+      "grad_norm": 6.787551403045654,
+      "learning_rate": 0.00015294978059483182,
+      "loss": 7.223753356933594,
+      "step": 8690
+    },
+    {
+      "epoch": 9.898377455166525,
+      "grad_norm": 6.694747447967529,
+      "learning_rate": 0.00015289560647922424,
+      "loss": 6.798531341552734,
+      "step": 8700
+    },
+    {
+      "epoch": 9.909763734699688,
+      "grad_norm": 6.917640686035156,
+      "learning_rate": 0.00015284143236361667,
+      "loss": 7.0906829833984375,
+      "step": 8710
+    },
+    {
+      "epoch": 9.921150014232849,
+      "grad_norm": 7.117720127105713,
+      "learning_rate": 0.0001527872582480091,
+      "loss": 7.004290771484375,
+      "step": 8720
+    },
+    {
+      "epoch": 9.932536293766011,
+      "grad_norm": 7.377773761749268,
+      "learning_rate": 0.00015273308413240154,
+      "loss": 6.975653839111328,
+      "step": 8730
+    },
+    {
+      "epoch": 9.943922573299174,
+      "grad_norm": 6.890756130218506,
+      "learning_rate": 0.00015267891001679397,
+      "loss": 6.900422668457031,
+      "step": 8740
+    },
+    {
+      "epoch": 9.955308852832337,
+      "grad_norm": 7.174078941345215,
+      "learning_rate": 0.00015262473590118642,
+      "loss": 7.064826202392578,
+      "step": 8750
+    },
+    {
+      "epoch": 9.9666951323655,
+      "grad_norm": 6.936395168304443,
+      "learning_rate": 0.00015257056178557884,
+      "loss": 6.7657325744628904,
+      "step": 8760
+    },
+    {
+      "epoch": 9.978081411898662,
+      "grad_norm": 6.216085433959961,
+      "learning_rate": 0.0001525163876699713,
+      "loss": 6.81783447265625,
+      "step": 8770
+    },
+    {
+      "epoch": 9.989467691431825,
+      "grad_norm": 6.8290839195251465,
+      "learning_rate": 0.00015246221355436375,
+      "loss": 6.718660736083985,
+      "step": 8780
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 5.303323745727539,
+      "learning_rate": 0.00015240803943875617,
+      "loss": 6.455754089355469,
+      "step": 8790
+    },
+    {
+      "epoch": 10.011386279533163,
+      "grad_norm": 6.515944480895996,
+      "learning_rate": 0.00015235386532314862,
+      "loss": 6.108221435546875,
+      "step": 8800
+    },
+    {
+      "epoch": 10.022772559066325,
+      "grad_norm": 6.763042449951172,
+      "learning_rate": 0.00015229969120754105,
+      "loss": 6.451548004150391,
+      "step": 8810
+    },
+    {
+      "epoch": 10.034158838599488,
+      "grad_norm": 7.289494037628174,
+      "learning_rate": 0.0001522455170919335,
+      "loss": 6.422400665283203,
+      "step": 8820
+    },
+    {
+      "epoch": 10.045545118132651,
+      "grad_norm": 6.849950790405273,
+      "learning_rate": 0.00015219134297632592,
+      "loss": 6.369514083862304,
+      "step": 8830
+    },
+    {
+      "epoch": 10.056931397665812,
+      "grad_norm": 6.839559555053711,
+      "learning_rate": 0.00015213716886071835,
+      "loss": 6.199007415771485,
+      "step": 8840
+    },
+    {
+      "epoch": 10.068317677198975,
+      "grad_norm": 7.444436550140381,
+      "learning_rate": 0.0001520829947451108,
+      "loss": 6.206901931762696,
+      "step": 8850
+    },
+    {
+      "epoch": 10.079703956732137,
+      "grad_norm": 7.2634782791137695,
+      "learning_rate": 0.00015202882062950322,
+      "loss": 6.411515808105468,
+      "step": 8860
+    },
+    {
+      "epoch": 10.0910902362653,
+      "grad_norm": 7.075826168060303,
+      "learning_rate": 0.00015197464651389565,
+      "loss": 6.321298980712891,
+      "step": 8870
+    },
+    {
+      "epoch": 10.102476515798463,
+      "grad_norm": 8.332730293273926,
+      "learning_rate": 0.0001519204723982881,
+      "loss": 6.658344268798828,
+      "step": 8880
+    },
+    {
+      "epoch": 10.113862795331626,
+      "grad_norm": 6.364902019500732,
+      "learning_rate": 0.00015186629828268055,
+      "loss": 6.109456634521484,
+      "step": 8890
+    },
+    {
+      "epoch": 10.125249074864788,
+      "grad_norm": 7.3898138999938965,
+      "learning_rate": 0.00015181212416707297,
+      "loss": 6.706474304199219,
+      "step": 8900
+    },
+    {
+      "epoch": 10.136635354397951,
+      "grad_norm": 7.5353264808654785,
+      "learning_rate": 0.00015175795005146543,
+      "loss": 6.191032791137696,
+      "step": 8910
+    },
+    {
+      "epoch": 10.148021633931114,
+      "grad_norm": 6.525182723999023,
+      "learning_rate": 0.00015170377593585785,
+      "loss": 6.050387191772461,
+      "step": 8920
+    },
+    {
+      "epoch": 10.159407913464275,
+      "grad_norm": 6.768884658813477,
+      "learning_rate": 0.0001516496018202503,
+      "loss": 6.448866271972657,
+      "step": 8930
+    },
+    {
+      "epoch": 10.170794192997437,
+      "grad_norm": 6.508273601531982,
+      "learning_rate": 0.00015159542770464273,
+      "loss": 6.556841278076172,
+      "step": 8940
+    },
+    {
+      "epoch": 10.1821804725306,
+      "grad_norm": 7.144713401794434,
+      "learning_rate": 0.00015154125358903518,
+      "loss": 6.7161407470703125,
+      "step": 8950
+    },
+    {
+      "epoch": 10.193566752063763,
+      "grad_norm": 6.919196128845215,
+      "learning_rate": 0.0001514870794734276,
+      "loss": 6.426815032958984,
+      "step": 8960
+    },
+    {
+      "epoch": 10.204953031596926,
+      "grad_norm": 6.691727161407471,
+      "learning_rate": 0.00015143290535782005,
+      "loss": 6.2255096435546875,
+      "step": 8970
+    },
+    {
+      "epoch": 10.216339311130088,
+      "grad_norm": 6.866335391998291,
+      "learning_rate": 0.00015137873124221248,
+      "loss": 6.1784015655517575,
+      "step": 8980
+    },
+    {
+      "epoch": 10.227725590663251,
+      "grad_norm": 6.891946792602539,
+      "learning_rate": 0.0001513245571266049,
+      "loss": 6.152831268310547,
+      "step": 8990
+    },
+    {
+      "epoch": 10.239111870196414,
+      "grad_norm": 6.802209854125977,
+      "learning_rate": 0.00015127038301099736,
+      "loss": 6.270240783691406,
+      "step": 9000
+    },
+    {
+      "epoch": 10.250498149729577,
+      "grad_norm": 7.624642372131348,
+      "learning_rate": 0.00015121620889538978,
+      "loss": 6.247557067871094,
+      "step": 9010
+    },
+    {
+      "epoch": 10.261884429262738,
+      "grad_norm": 7.880180835723877,
+      "learning_rate": 0.00015116203477978223,
+      "loss": 6.371080017089843,
+      "step": 9020
+    },
+    {
+      "epoch": 10.2732707087959,
+      "grad_norm": 6.9289116859436035,
+      "learning_rate": 0.00015110786066417466,
+      "loss": 6.2116645812988285,
+      "step": 9030
+    },
+    {
+      "epoch": 10.284656988329063,
+      "grad_norm": 7.059987545013428,
+      "learning_rate": 0.0001510536865485671,
+      "loss": 6.152032852172852,
+      "step": 9040
+    },
+    {
+      "epoch": 10.296043267862226,
+      "grad_norm": 7.274111270904541,
+      "learning_rate": 0.00015099951243295953,
+      "loss": 6.412078857421875,
+      "step": 9050
+    },
+    {
+      "epoch": 10.307429547395389,
+      "grad_norm": 6.609608173370361,
+      "learning_rate": 0.00015094533831735198,
+      "loss": 6.340757369995117,
+      "step": 9060
+    },
+    {
+      "epoch": 10.318815826928551,
+      "grad_norm": 6.39749002456665,
+      "learning_rate": 0.0001508911642017444,
+      "loss": 6.413553619384766,
+      "step": 9070
+    },
+    {
+      "epoch": 10.330202106461714,
+      "grad_norm": 6.884727478027344,
+      "learning_rate": 0.00015083699008613686,
+      "loss": 6.569889068603516,
+      "step": 9080
+    },
+    {
+      "epoch": 10.341588385994877,
+      "grad_norm": 7.041878700256348,
+      "learning_rate": 0.00015078281597052928,
+      "loss": 6.179785537719726,
+      "step": 9090
+    },
+    {
+      "epoch": 10.35297466552804,
+      "grad_norm": 6.370310306549072,
+      "learning_rate": 0.00015072864185492174,
+      "loss": 6.136286926269531,
+      "step": 9100
+    },
+    {
+      "epoch": 10.3643609450612,
+      "grad_norm": 7.227022647857666,
+      "learning_rate": 0.0001506744677393142,
+      "loss": 6.1931709289550785,
+      "step": 9110
+    },
+    {
+      "epoch": 10.375747224594363,
+      "grad_norm": 6.813343048095703,
+      "learning_rate": 0.0001506202936237066,
+      "loss": 6.568285369873047,
+      "step": 9120
+    },
+    {
+      "epoch": 10.387133504127526,
+      "grad_norm": 6.486794948577881,
+      "learning_rate": 0.00015056611950809904,
+      "loss": 6.357745361328125,
+      "step": 9130
+    },
+    {
+      "epoch": 10.398519783660689,
+      "grad_norm": 7.465872764587402,
+      "learning_rate": 0.00015051194539249146,
+      "loss": 6.26553955078125,
+      "step": 9140
+    },
+    {
+      "epoch": 10.409906063193851,
+      "grad_norm": 7.438695907592773,
+      "learning_rate": 0.0001504577712768839,
+      "loss": 6.602199554443359,
+      "step": 9150
+    },
+    {
+      "epoch": 10.421292342727014,
+      "grad_norm": 8.035082817077637,
+      "learning_rate": 0.00015040359716127634,
+      "loss": 6.462114715576172,
+      "step": 9160
+    },
+    {
+      "epoch": 10.432678622260177,
+      "grad_norm": 6.581029891967773,
+      "learning_rate": 0.0001503494230456688,
+      "loss": 6.531999206542968,
+      "step": 9170
+    },
+    {
+      "epoch": 10.44406490179334,
+      "grad_norm": 7.6987175941467285,
+      "learning_rate": 0.0001502952489300612,
+      "loss": 6.53165283203125,
+      "step": 9180
+    },
+    {
+      "epoch": 10.455451181326502,
+      "grad_norm": 5.949412822723389,
+      "learning_rate": 0.00015024107481445366,
+      "loss": 5.95707893371582,
+      "step": 9190
+    },
+    {
+      "epoch": 10.466837460859663,
+      "grad_norm": 7.387345314025879,
+      "learning_rate": 0.0001501869006988461,
+      "loss": 6.544392395019531,
+      "step": 9200
+    },
+    {
+      "epoch": 10.478223740392826,
+      "grad_norm": 7.590228080749512,
+      "learning_rate": 0.00015013272658323854,
+      "loss": 6.59503173828125,
+      "step": 9210
+    },
+    {
+      "epoch": 10.489610019925989,
+      "grad_norm": 6.827988147735596,
+      "learning_rate": 0.000150078552467631,
+      "loss": 6.5579277038574215,
+      "step": 9220
+    },
+    {
+      "epoch": 10.500996299459151,
+      "grad_norm": 6.989532470703125,
+      "learning_rate": 0.00015002437835202342,
+      "loss": 6.056536102294922,
+      "step": 9230
+    },
+    {
+      "epoch": 10.512382578992314,
+      "grad_norm": 6.390324592590332,
+      "learning_rate": 0.00014997020423641587,
+      "loss": 6.437796783447266,
+      "step": 9240
+    },
+    {
+      "epoch": 10.523768858525477,
+      "grad_norm": 7.057545185089111,
+      "learning_rate": 0.0001499160301208083,
+      "loss": 6.424283599853515,
+      "step": 9250
+    },
+    {
+      "epoch": 10.53515513805864,
+      "grad_norm": 7.236074924468994,
+      "learning_rate": 0.00014986185600520072,
+      "loss": 6.465802001953125,
+      "step": 9260
+    },
+    {
+      "epoch": 10.546541417591802,
+      "grad_norm": 7.380622863769531,
+      "learning_rate": 0.00014980768188959314,
+      "loss": 6.5994873046875,
+      "step": 9270
+    },
+    {
+      "epoch": 10.557927697124965,
+      "grad_norm": 6.613060474395752,
+      "learning_rate": 0.0001497535077739856,
+      "loss": 6.399759292602539,
+      "step": 9280
+    },
+    {
+      "epoch": 10.569313976658126,
+      "grad_norm": 7.060276508331299,
+      "learning_rate": 0.00014969933365837802,
+      "loss": 6.557682800292969,
+      "step": 9290
+    },
+    {
+      "epoch": 10.580700256191289,
+      "grad_norm": 6.692377090454102,
+      "learning_rate": 0.00014964515954277047,
+      "loss": 6.2362712860107425,
+      "step": 9300
+    },
+    {
+      "epoch": 10.592086535724452,
+      "grad_norm": 7.573443412780762,
+      "learning_rate": 0.0001495909854271629,
+      "loss": 6.369208908081054,
+      "step": 9310
+    },
+    {
+      "epoch": 10.603472815257614,
+      "grad_norm": 7.296620845794678,
+      "learning_rate": 0.00014953681131155534,
+      "loss": 6.499317932128906,
+      "step": 9320
+    },
+    {
+      "epoch": 10.614859094790777,
+      "grad_norm": 7.194809436798096,
+      "learning_rate": 0.00014948263719594777,
+      "loss": 6.53308334350586,
+      "step": 9330
+    },
+    {
+      "epoch": 10.62624537432394,
+      "grad_norm": 7.089186191558838,
+      "learning_rate": 0.00014942846308034022,
+      "loss": 6.357146835327148,
+      "step": 9340
+    },
+    {
+      "epoch": 10.637631653857103,
+      "grad_norm": 6.138644695281982,
+      "learning_rate": 0.00014937428896473267,
+      "loss": 6.987527465820312,
+      "step": 9350
+    },
+    {
+      "epoch": 10.649017933390265,
+      "grad_norm": 6.458044528961182,
+      "learning_rate": 0.0001493201148491251,
+      "loss": 6.447171783447265,
+      "step": 9360
+    },
+    {
+      "epoch": 10.660404212923428,
+      "grad_norm": 6.982960224151611,
+      "learning_rate": 0.00014926594073351755,
+      "loss": 6.367984390258789,
+      "step": 9370
+    },
+    {
+      "epoch": 10.67179049245659,
+      "grad_norm": 7.014771461486816,
+      "learning_rate": 0.00014921176661790997,
+      "loss": 6.597218322753906,
+      "step": 9380
+    },
+    {
+      "epoch": 10.683176771989752,
+      "grad_norm": 7.192128658294678,
+      "learning_rate": 0.00014915759250230242,
+      "loss": 6.468855285644532,
+      "step": 9390
+    },
+    {
+      "epoch": 10.694563051522914,
+      "grad_norm": 6.968006134033203,
+      "learning_rate": 0.00014910341838669485,
+      "loss": 6.514220428466797,
+      "step": 9400
+    },
+    {
+      "epoch": 10.705949331056077,
+      "grad_norm": 7.2908220291137695,
+      "learning_rate": 0.00014904924427108727,
+      "loss": 5.984500885009766,
+      "step": 9410
+    },
+    {
+      "epoch": 10.71733561058924,
+      "grad_norm": 7.383593559265137,
+      "learning_rate": 0.0001489950701554797,
+      "loss": 6.390668869018555,
+      "step": 9420
+    },
+    {
+      "epoch": 10.728721890122403,
+      "grad_norm": 7.428356170654297,
+      "learning_rate": 0.00014894089603987215,
+      "loss": 6.504000854492188,
+      "step": 9430
+    },
+    {
+      "epoch": 10.740108169655565,
+      "grad_norm": 6.935097694396973,
+      "learning_rate": 0.00014888672192426457,
+      "loss": 6.668525695800781,
+      "step": 9440
+    },
+    {
+      "epoch": 10.751494449188728,
+      "grad_norm": 7.8939595222473145,
+      "learning_rate": 0.00014883254780865703,
+      "loss": 6.7964630126953125,
+      "step": 9450
+    },
+    {
+      "epoch": 10.76288072872189,
+      "grad_norm": 7.261744976043701,
+      "learning_rate": 0.00014877837369304948,
+      "loss": 6.494899749755859,
+      "step": 9460
+    },
+    {
+      "epoch": 10.774267008255052,
+      "grad_norm": 7.5321044921875,
+      "learning_rate": 0.0001487241995774419,
+      "loss": 6.319461441040039,
+      "step": 9470
+    },
+    {
+      "epoch": 10.785653287788215,
+      "grad_norm": 7.152090072631836,
+      "learning_rate": 0.00014867002546183435,
+      "loss": 6.386759948730469,
+      "step": 9480
+    },
+    {
+      "epoch": 10.797039567321377,
+      "grad_norm": 6.318842887878418,
+      "learning_rate": 0.00014861585134622678,
+      "loss": 6.789165496826172,
+      "step": 9490
+    },
+    {
+      "epoch": 10.80842584685454,
+      "grad_norm": 7.488574981689453,
+      "learning_rate": 0.00014856167723061923,
+      "loss": 6.679385375976563,
+      "step": 9500
+    },
+    {
+      "epoch": 10.819812126387703,
+      "grad_norm": 7.495085716247559,
+      "learning_rate": 0.00014851292052657242,
+      "loss": 6.523779296875,
+      "step": 9510
+    },
+    {
+      "epoch": 10.831198405920865,
+      "grad_norm": 6.9265522956848145,
+      "learning_rate": 0.00014845874641096485,
+      "loss": 6.4424797058105465,
+      "step": 9520
+    },
+    {
+      "epoch": 10.842584685454028,
+      "grad_norm": 6.2375359535217285,
+      "learning_rate": 0.0001484045722953573,
+      "loss": 6.325106811523438,
+      "step": 9530
+    },
+    {
+      "epoch": 10.853970964987191,
+      "grad_norm": 7.461876392364502,
+      "learning_rate": 0.00014835039817974973,
+      "loss": 6.4818115234375,
+      "step": 9540
+    },
+    {
+      "epoch": 10.865357244520354,
+      "grad_norm": 6.406781196594238,
+      "learning_rate": 0.00014829622406414218,
+      "loss": 6.454816436767578,
+      "step": 9550
+    },
+    {
+      "epoch": 10.876743524053516,
+      "grad_norm": 6.492645263671875,
+      "learning_rate": 0.0001482420499485346,
+      "loss": 6.191008377075195,
+      "step": 9560
+    },
+    {
+      "epoch": 10.888129803586677,
+      "grad_norm": 7.3028645515441895,
+      "learning_rate": 0.00014818787583292703,
+      "loss": 6.245122909545898,
+      "step": 9570
+    },
+    {
+      "epoch": 10.89951608311984,
+      "grad_norm": 6.861218452453613,
+      "learning_rate": 0.00014813370171731945,
+      "loss": 6.476416778564453,
+      "step": 9580
+    },
+    {
+      "epoch": 10.910902362653003,
+      "grad_norm": 7.525650501251221,
+      "learning_rate": 0.0001480795276017119,
+      "loss": 6.472120666503907,
+      "step": 9590
+    },
+    {
+      "epoch": 10.922288642186166,
+      "grad_norm": 7.155680179595947,
+      "learning_rate": 0.00014802535348610435,
+      "loss": 6.589835357666016,
+      "step": 9600
+    },
+    {
+      "epoch": 10.933674921719328,
+      "grad_norm": 7.079594612121582,
+      "learning_rate": 0.00014797117937049678,
+      "loss": 6.457518005371094,
+      "step": 9610
+    },
+    {
+      "epoch": 10.945061201252491,
+      "grad_norm": 7.258676052093506,
+      "learning_rate": 0.00014791700525488923,
+      "loss": 6.361143112182617,
+      "step": 9620
+    },
+    {
+      "epoch": 10.956447480785654,
+      "grad_norm": 7.055770397186279,
+      "learning_rate": 0.00014786283113928165,
+      "loss": 6.50369873046875,
+      "step": 9630
+    },
+    {
+      "epoch": 10.967833760318817,
+      "grad_norm": 6.666393756866455,
+      "learning_rate": 0.0001478086570236741,
+      "loss": 6.464437866210938,
+      "step": 9640
+    },
+    {
+      "epoch": 10.979220039851977,
+      "grad_norm": 6.598408222198486,
+      "learning_rate": 0.00014775448290806653,
+      "loss": 6.123414993286133,
+      "step": 9650
+    },
+    {
+      "epoch": 10.99060631938514,
+      "grad_norm": 6.270415306091309,
+      "learning_rate": 0.00014770030879245898,
+      "loss": 6.181151580810547,
+      "step": 9660
+    },
+    {
+      "epoch": 11.001138627953317,
+      "grad_norm": 6.535286903381348,
+      "learning_rate": 0.0001476461346768514,
+      "loss": 5.540151977539063,
+      "step": 9670
+    },
+    {
+      "epoch": 11.012524907486478,
+      "grad_norm": 6.144158363342285,
+      "learning_rate": 0.00014759196056124386,
+      "loss": 5.872829818725586,
+      "step": 9680
+    },
+    {
+      "epoch": 11.02391118701964,
+      "grad_norm": 7.178324222564697,
+      "learning_rate": 0.00014753778644563628,
+      "loss": 6.154612350463867,
+      "step": 9690
+    },
+    {
+      "epoch": 11.035297466552803,
+      "grad_norm": 6.866833209991455,
+      "learning_rate": 0.0001474836123300287,
+      "loss": 5.634689712524414,
+      "step": 9700
+    },
+    {
+      "epoch": 11.046683746085966,
+      "grad_norm": 7.015516757965088,
+      "learning_rate": 0.00014742943821442116,
+      "loss": 5.930116271972656,
+      "step": 9710
+    },
+    {
+      "epoch": 11.058070025619129,
+      "grad_norm": 7.600584983825684,
+      "learning_rate": 0.00014737526409881358,
+      "loss": 5.9960792541503904,
+      "step": 9720
+    },
+    {
+      "epoch": 11.069456305152292,
+      "grad_norm": 7.366182804107666,
+      "learning_rate": 0.00014732108998320603,
+      "loss": 5.631900787353516,
+      "step": 9730
+    },
+    {
+      "epoch": 11.080842584685454,
+      "grad_norm": 7.046829700469971,
+      "learning_rate": 0.00014726691586759846,
+      "loss": 6.252207183837891,
+      "step": 9740
+    },
+    {
+      "epoch": 11.092228864218617,
+      "grad_norm": 6.92673397064209,
+      "learning_rate": 0.0001472127417519909,
+      "loss": 6.086066818237304,
+      "step": 9750
+    },
+    {
+      "epoch": 11.10361514375178,
+      "grad_norm": 9.307673454284668,
+      "learning_rate": 0.00014715856763638333,
+      "loss": 6.022590255737304,
+      "step": 9760
+    },
+    {
+      "epoch": 11.11500142328494,
+      "grad_norm": 7.187988758087158,
+      "learning_rate": 0.00014710439352077579,
+      "loss": 5.973920440673828,
+      "step": 9770
+    },
+    {
+      "epoch": 11.126387702818104,
+      "grad_norm": 7.109073638916016,
+      "learning_rate": 0.0001470502194051682,
+      "loss": 5.864291000366211,
+      "step": 9780
+    },
+    {
+      "epoch": 11.137773982351266,
+      "grad_norm": 7.3360276222229,
+      "learning_rate": 0.00014699604528956066,
+      "loss": 6.124382781982422,
+      "step": 9790
+    },
+    {
+      "epoch": 11.149160261884429,
+      "grad_norm": 7.7672343254089355,
+      "learning_rate": 0.0001469418711739531,
+      "loss": 6.148440170288086,
+      "step": 9800
+    },
+    {
+      "epoch": 11.160546541417592,
+      "grad_norm": 6.8055100440979,
+      "learning_rate": 0.00014688769705834554,
+      "loss": 5.931351852416992,
+      "step": 9810
+    },
+    {
+      "epoch": 11.171932820950754,
+      "grad_norm": 7.583737373352051,
+      "learning_rate": 0.000146833522942738,
+      "loss": 5.668278503417969,
+      "step": 9820
+    },
+    {
+      "epoch": 11.183319100483917,
+      "grad_norm": 7.5180463790893555,
+      "learning_rate": 0.00014677934882713041,
+      "loss": 6.020656585693359,
+      "step": 9830
+    },
+    {
+      "epoch": 11.19470538001708,
+      "grad_norm": 7.665243625640869,
+      "learning_rate": 0.00014672517471152284,
+      "loss": 5.651744842529297,
+      "step": 9840
+    },
+    {
+      "epoch": 11.206091659550243,
+      "grad_norm": 7.284232139587402,
+      "learning_rate": 0.00014667100059591526,
+      "loss": 5.927288818359375,
+      "step": 9850
+    },
+    {
+      "epoch": 11.217477939083404,
+      "grad_norm": 6.677280426025391,
+      "learning_rate": 0.00014661682648030772,
+      "loss": 5.800758743286133,
+      "step": 9860
+    },
+    {
+      "epoch": 11.228864218616566,
+      "grad_norm": 6.858931064605713,
+      "learning_rate": 0.00014656265236470014,
+      "loss": 5.876351928710937,
+      "step": 9870
+    },
+    {
+      "epoch": 11.240250498149729,
+      "grad_norm": 6.993981838226318,
+      "learning_rate": 0.0001465084782490926,
+      "loss": 5.805530166625976,
+      "step": 9880
+    },
+    {
+      "epoch": 11.251636777682892,
+      "grad_norm": 7.795031547546387,
+      "learning_rate": 0.00014645430413348502,
+      "loss": 6.2553356170654295,
+      "step": 9890
+    },
+    {
+      "epoch": 11.263023057216055,
+      "grad_norm": 7.493925094604492,
+      "learning_rate": 0.00014640013001787747,
+      "loss": 6.046601867675781,
+      "step": 9900
+    },
+    {
+      "epoch": 11.274409336749217,
+      "grad_norm": 7.555092811584473,
+      "learning_rate": 0.0001463459559022699,
+      "loss": 5.973406982421875,
+      "step": 9910
+    },
+    {
+      "epoch": 11.28579561628238,
+      "grad_norm": 6.385543346405029,
+      "learning_rate": 0.00014629178178666234,
+      "loss": 6.176846694946289,
+      "step": 9920
+    },
+    {
+      "epoch": 11.297181895815543,
+      "grad_norm": 7.017746448516846,
+      "learning_rate": 0.00014623760767105477,
+      "loss": 6.001755142211914,
+      "step": 9930
+    },
+    {
+      "epoch": 11.308568175348706,
+      "grad_norm": 7.1694416999816895,
+      "learning_rate": 0.00014618343355544722,
+      "loss": 6.172482299804687,
+      "step": 9940
+    },
+    {
+      "epoch": 11.319954454881866,
+      "grad_norm": 7.279458999633789,
+      "learning_rate": 0.00014612925943983967,
+      "loss": 6.369392013549804,
+      "step": 9950
+    },
+    {
+      "epoch": 11.33134073441503,
+      "grad_norm": 6.9396772384643555,
+      "learning_rate": 0.0001460750853242321,
+      "loss": 5.813999176025391,
+      "step": 9960
+    },
+    {
+      "epoch": 11.342727013948192,
+      "grad_norm": 7.31268310546875,
+      "learning_rate": 0.00014602091120862455,
+      "loss": 5.91674575805664,
+      "step": 9970
+    },
+    {
+      "epoch": 11.354113293481355,
+      "grad_norm": 6.88545560836792,
+      "learning_rate": 0.00014596673709301697,
+      "loss": 5.994848251342773,
+      "step": 9980
+    },
+    {
+      "epoch": 11.365499573014517,
+      "grad_norm": 7.708160400390625,
+      "learning_rate": 0.0001459125629774094,
+      "loss": 5.773896408081055,
+      "step": 9990
+    },
+    {
+      "epoch": 11.37688585254768,
+      "grad_norm": 7.004419803619385,
+      "learning_rate": 0.00014585838886180182,
+      "loss": 5.685453796386719,
+      "step": 10000
+    },
+    {
+      "epoch": 11.388272132080843,
+      "grad_norm": 6.840615749359131,
+      "learning_rate": 0.00014580421474619427,
+      "loss": 5.64319953918457,
+      "step": 10010
+    },
+    {
+      "epoch": 11.399658411614006,
+      "grad_norm": 7.056737899780273,
+      "learning_rate": 0.0001457500406305867,
+      "loss": 5.996237945556641,
+      "step": 10020
+    },
+    {
+      "epoch": 11.411044691147168,
+      "grad_norm": 7.051031589508057,
+      "learning_rate": 0.00014569586651497915,
+      "loss": 6.065059661865234,
+      "step": 10030
+    },
+    {
+      "epoch": 11.422430970680331,
+      "grad_norm": 6.999020576477051,
+      "learning_rate": 0.00014564169239937157,
+      "loss": 5.551327896118164,
+      "step": 10040
+    },
+    {
+      "epoch": 11.433817250213492,
+      "grad_norm": 7.3235650062561035,
+      "learning_rate": 0.00014558751828376402,
+      "loss": 5.727095031738282,
+      "step": 10050
+    },
+    {
+      "epoch": 11.445203529746655,
+      "grad_norm": 6.848404884338379,
+      "learning_rate": 0.00014553334416815648,
+      "loss": 6.206461334228516,
+      "step": 10060
+    },
+    {
+      "epoch": 11.456589809279818,
+      "grad_norm": 6.96482515335083,
+      "learning_rate": 0.0001454791700525489,
+      "loss": 6.19579963684082,
+      "step": 10070
+    },
+    {
+      "epoch": 11.46797608881298,
+      "grad_norm": 7.321053981781006,
+      "learning_rate": 0.00014542499593694135,
+      "loss": 6.439669799804688,
+      "step": 10080
+    },
+    {
+      "epoch": 11.479362368346143,
+      "grad_norm": 6.996761798858643,
+      "learning_rate": 0.00014537082182133378,
+      "loss": 5.96948356628418,
+      "step": 10090
+    },
+    {
+      "epoch": 11.490748647879306,
+      "grad_norm": 6.9859466552734375,
+      "learning_rate": 0.00014531664770572623,
+      "loss": 5.9932861328125,
+      "step": 10100
+    },
+    {
+      "epoch": 11.502134927412468,
+      "grad_norm": 6.946524620056152,
+      "learning_rate": 0.00014526247359011865,
+      "loss": 5.791952514648438,
+      "step": 10110
+    },
+    {
+      "epoch": 11.513521206945631,
+      "grad_norm": 7.48444938659668,
+      "learning_rate": 0.00014520829947451108,
+      "loss": 6.16157112121582,
+      "step": 10120
+    },
+    {
+      "epoch": 11.524907486478792,
+      "grad_norm": 7.355890274047852,
+      "learning_rate": 0.0001451541253589035,
+      "loss": 5.91258659362793,
+      "step": 10130
+    },
+    {
+      "epoch": 11.536293766011955,
+      "grad_norm": 7.450939178466797,
+      "learning_rate": 0.00014509995124329595,
+      "loss": 6.787471008300781,
+      "step": 10140
+    },
+    {
+      "epoch": 11.547680045545118,
+      "grad_norm": 7.768852710723877,
+      "learning_rate": 0.00014504577712768838,
+      "loss": 6.156349563598633,
+      "step": 10150
+    },
+    {
+      "epoch": 11.55906632507828,
+      "grad_norm": 7.6208953857421875,
+      "learning_rate": 0.00014499160301208083,
+      "loss": 6.0792900085449215,
+      "step": 10160
+    },
+    {
+      "epoch": 11.570452604611443,
+      "grad_norm": 6.2815632820129395,
+      "learning_rate": 0.00014493742889647328,
+      "loss": 5.962456130981446,
+      "step": 10170
+    },
+    {
+      "epoch": 11.581838884144606,
+      "grad_norm": 7.529908180236816,
+      "learning_rate": 0.0001448832547808657,
+      "loss": 6.132223510742188,
+      "step": 10180
+    },
+    {
+      "epoch": 11.593225163677769,
+      "grad_norm": 7.08761739730835,
+      "learning_rate": 0.00014482908066525816,
+      "loss": 6.127777099609375,
+      "step": 10190
+    },
+    {
+      "epoch": 11.604611443210931,
+      "grad_norm": 8.239002227783203,
+      "learning_rate": 0.00014477490654965058,
+      "loss": 5.923508834838867,
+      "step": 10200
+    },
+    {
+      "epoch": 11.615997722744094,
+      "grad_norm": 7.077022552490234,
+      "learning_rate": 0.00014472073243404303,
+      "loss": 5.75263786315918,
+      "step": 10210
+    },
+    {
+      "epoch": 11.627384002277257,
+      "grad_norm": 7.765398979187012,
+      "learning_rate": 0.00014466655831843546,
+      "loss": 6.193927001953125,
+      "step": 10220
+    },
+    {
+      "epoch": 11.638770281810418,
+      "grad_norm": 7.009711265563965,
+      "learning_rate": 0.0001446123842028279,
+      "loss": 5.800412368774414,
+      "step": 10230
+    },
+    {
+      "epoch": 11.65015656134358,
+      "grad_norm": 7.33494234085083,
+      "learning_rate": 0.00014455821008722033,
+      "loss": 5.910943222045899,
+      "step": 10240
+    },
+    {
+      "epoch": 11.661542840876743,
+      "grad_norm": 7.698012351989746,
+      "learning_rate": 0.00014450403597161278,
+      "loss": 6.299586486816406,
+      "step": 10250
+    },
+    {
+      "epoch": 11.672929120409906,
+      "grad_norm": 7.655066967010498,
+      "learning_rate": 0.0001444498618560052,
+      "loss": 6.288556671142578,
+      "step": 10260
+    },
+    {
+      "epoch": 11.684315399943069,
+      "grad_norm": 7.222828388214111,
+      "learning_rate": 0.00014439568774039763,
+      "loss": 5.758666610717773,
+      "step": 10270
+    },
+    {
+      "epoch": 11.695701679476231,
+      "grad_norm": 7.702454566955566,
+      "learning_rate": 0.00014434151362479008,
+      "loss": 6.061179733276367,
+      "step": 10280
+    },
+    {
+      "epoch": 11.707087959009394,
+      "grad_norm": 8.399511337280273,
+      "learning_rate": 0.0001442873395091825,
+      "loss": 6.088422012329102,
+      "step": 10290
+    },
+    {
+      "epoch": 11.718474238542557,
+      "grad_norm": 6.886438846588135,
+      "learning_rate": 0.00014423316539357496,
+      "loss": 6.215003585815429,
+      "step": 10300
+    },
+    {
+      "epoch": 11.729860518075718,
+      "grad_norm": 6.97210168838501,
+      "learning_rate": 0.00014417899127796739,
+      "loss": 5.770806121826172,
+      "step": 10310
+    },
+    {
+      "epoch": 11.74124679760888,
+      "grad_norm": 6.872792720794678,
+      "learning_rate": 0.00014412481716235984,
+      "loss": 6.1828559875488285,
+      "step": 10320
+    },
+    {
+      "epoch": 11.752633077142043,
+      "grad_norm": 7.779046535491943,
+      "learning_rate": 0.00014407064304675226,
+      "loss": 5.819805908203125,
+      "step": 10330
+    },
+    {
+      "epoch": 11.764019356675206,
+      "grad_norm": 6.890455722808838,
+      "learning_rate": 0.0001440164689311447,
+      "loss": 5.905824279785156,
+      "step": 10340
+    },
+    {
+      "epoch": 11.775405636208369,
+      "grad_norm": 7.692152976989746,
+      "learning_rate": 0.00014396229481553714,
+      "loss": 5.950515365600586,
+      "step": 10350
+    },
+    {
+      "epoch": 11.786791915741532,
+      "grad_norm": 7.703096866607666,
+      "learning_rate": 0.0001439081206999296,
+      "loss": 6.3083740234375,
+      "step": 10360
+    },
+    {
+      "epoch": 11.798178195274694,
+      "grad_norm": 7.19696044921875,
+      "learning_rate": 0.00014385394658432201,
+      "loss": 6.3550971984863285,
+      "step": 10370
+    },
+    {
+      "epoch": 11.809564474807857,
+      "grad_norm": 6.914282321929932,
+      "learning_rate": 0.00014379977246871447,
+      "loss": 5.792062759399414,
+      "step": 10380
+    },
+    {
+      "epoch": 11.82095075434102,
+      "grad_norm": 7.224524974822998,
+      "learning_rate": 0.0001437455983531069,
+      "loss": 5.989566802978516,
+      "step": 10390
+    },
+    {
+      "epoch": 11.832337033874182,
+      "grad_norm": 7.418369293212891,
+      "learning_rate": 0.00014369142423749934,
+      "loss": 5.995355224609375,
+      "step": 10400
+    },
+    {
+      "epoch": 11.843723313407343,
+      "grad_norm": 6.645310878753662,
+      "learning_rate": 0.00014363725012189177,
+      "loss": 5.791942596435547,
+      "step": 10410
+    },
+    {
+      "epoch": 11.855109592940506,
+      "grad_norm": 7.1590375900268555,
+      "learning_rate": 0.0001435830760062842,
+      "loss": 5.956428146362304,
+      "step": 10420
+    },
+    {
+      "epoch": 11.866495872473669,
+      "grad_norm": 7.662755489349365,
+      "learning_rate": 0.00014352890189067664,
+      "loss": 6.0646930694580075,
+      "step": 10430
+    },
+    {
+      "epoch": 11.877882152006832,
+      "grad_norm": 7.7084550857543945,
+      "learning_rate": 0.00014347472777506907,
+      "loss": 6.142111206054688,
+      "step": 10440
+    },
+    {
+      "epoch": 11.889268431539994,
+      "grad_norm": 6.89467191696167,
+      "learning_rate": 0.00014342055365946152,
+      "loss": 5.906317138671875,
+      "step": 10450
+    },
+    {
+      "epoch": 11.900654711073157,
+      "grad_norm": 7.120398044586182,
+      "learning_rate": 0.00014336637954385394,
+      "loss": 6.146725463867187,
+      "step": 10460
+    },
+    {
+      "epoch": 11.91204099060632,
+      "grad_norm": 7.415362358093262,
+      "learning_rate": 0.0001433122054282464,
+      "loss": 6.037253570556641,
+      "step": 10470
+    },
+    {
+      "epoch": 11.923427270139483,
+      "grad_norm": 7.9902262687683105,
+      "learning_rate": 0.00014325803131263882,
+      "loss": 6.28980598449707,
+      "step": 10480
+    },
+    {
+      "epoch": 11.934813549672645,
+      "grad_norm": 6.856546878814697,
+      "learning_rate": 0.00014320385719703127,
+      "loss": 6.058418273925781,
+      "step": 10490
+    },
+    {
+      "epoch": 11.946199829205806,
+      "grad_norm": 7.198825836181641,
+      "learning_rate": 0.00014314968308142372,
+      "loss": 6.076106262207031,
+      "step": 10500
+    },
+    {
+      "epoch": 11.957586108738969,
+      "grad_norm": 7.608924388885498,
+      "learning_rate": 0.00014309550896581615,
+      "loss": 5.793802261352539,
+      "step": 10510
+    },
+    {
+      "epoch": 11.968972388272132,
+      "grad_norm": 7.0075225830078125,
+      "learning_rate": 0.0001430413348502086,
+      "loss": 6.062651443481445,
+      "step": 10520
+    },
+    {
+      "epoch": 11.980358667805294,
+      "grad_norm": 6.545762062072754,
+      "learning_rate": 0.00014298716073460102,
+      "loss": 5.8560432434082035,
+      "step": 10530
+    },
+    {
+      "epoch": 11.991744947338457,
+      "grad_norm": 7.019240379333496,
+      "learning_rate": 0.00014293298661899345,
+      "loss": 6.309774398803711,
+      "step": 10540
+    },
+    {
+      "epoch": 12.002277255906632,
+      "grad_norm": 6.560851097106934,
+      "learning_rate": 0.00014287881250338587,
+      "loss": 5.244910812377929,
+      "step": 10550
+    },
+    {
+      "epoch": 12.013663535439795,
+      "grad_norm": 6.521633625030518,
+      "learning_rate": 0.00014282463838777832,
+      "loss": 5.584713363647461,
+      "step": 10560
+    },
+    {
+      "epoch": 12.025049814972958,
+      "grad_norm": 7.09952974319458,
+      "learning_rate": 0.00014277046427217075,
+      "loss": 5.414521789550781,
+      "step": 10570
+    },
+    {
+      "epoch": 12.03643609450612,
+      "grad_norm": 7.587541103363037,
+      "learning_rate": 0.0001427162901565632,
+      "loss": 5.422273635864258,
+      "step": 10580
+    },
+    {
+      "epoch": 12.047822374039283,
+      "grad_norm": 8.005366325378418,
+      "learning_rate": 0.00014266211604095562,
+      "loss": 5.438925552368164,
+      "step": 10590
+    },
+    {
+      "epoch": 12.059208653572446,
+      "grad_norm": 7.477734565734863,
+      "learning_rate": 0.00014260794192534807,
+      "loss": 5.222858810424805,
+      "step": 10600
+    },
+    {
+      "epoch": 12.070594933105609,
+      "grad_norm": 7.249375343322754,
+      "learning_rate": 0.00014255376780974053,
+      "loss": 5.38745231628418,
+      "step": 10610
+    },
+    {
+      "epoch": 12.08198121263877,
+      "grad_norm": 7.713613510131836,
+      "learning_rate": 0.00014249959369413295,
+      "loss": 5.812319946289063,
+      "step": 10620
+    },
+    {
+      "epoch": 12.093367492171932,
+      "grad_norm": 7.597423553466797,
+      "learning_rate": 0.0001424454195785254,
+      "loss": 5.536786270141602,
+      "step": 10630
+    },
+    {
+      "epoch": 12.104753771705095,
+      "grad_norm": 6.994418621063232,
+      "learning_rate": 0.00014239124546291783,
+      "loss": 5.595279312133789,
+      "step": 10640
+    },
+    {
+      "epoch": 12.116140051238258,
+      "grad_norm": 7.918735027313232,
+      "learning_rate": 0.00014233707134731028,
+      "loss": 5.551082611083984,
+      "step": 10650
+    },
+    {
+      "epoch": 12.12752633077142,
+      "grad_norm": 7.9564666748046875,
+      "learning_rate": 0.0001422828972317027,
+      "loss": 5.564613342285156,
+      "step": 10660
+    },
+    {
+      "epoch": 12.138912610304583,
+      "grad_norm": 6.786741733551025,
+      "learning_rate": 0.00014222872311609515,
+      "loss": 5.819846725463867,
+      "step": 10670
+    },
+    {
+      "epoch": 12.150298889837746,
+      "grad_norm": 6.291092872619629,
+      "learning_rate": 0.00014217454900048758,
+      "loss": 5.597542190551758,
+      "step": 10680
+    },
+    {
+      "epoch": 12.161685169370909,
+      "grad_norm": 8.643335342407227,
+      "learning_rate": 0.00014212037488488,
+      "loss": 5.279590225219726,
+      "step": 10690
+    },
+    {
+      "epoch": 12.173071448904071,
+      "grad_norm": 8.687835693359375,
+      "learning_rate": 0.00014206620076927243,
+      "loss": 5.638618087768554,
+      "step": 10700
+    },
+    {
+      "epoch": 12.184457728437232,
+      "grad_norm": 7.45721960067749,
+      "learning_rate": 0.00014201202665366488,
+      "loss": 5.268462371826172,
+      "step": 10710
+    },
+    {
+      "epoch": 12.195844007970395,
+      "grad_norm": 7.256382942199707,
+      "learning_rate": 0.0001419578525380573,
+      "loss": 5.795388412475586,
+      "step": 10720
+    },
+    {
+      "epoch": 12.207230287503558,
+      "grad_norm": 7.054190635681152,
+      "learning_rate": 0.00014190367842244976,
+      "loss": 5.79822006225586,
+      "step": 10730
+    },
+    {
+      "epoch": 12.21861656703672,
+      "grad_norm": 6.79879093170166,
+      "learning_rate": 0.0001418495043068422,
+      "loss": 5.52197036743164,
+      "step": 10740
+    },
+    {
+      "epoch": 12.230002846569883,
+      "grad_norm": 7.19872522354126,
+      "learning_rate": 0.00014179533019123463,
+      "loss": 5.513645935058594,
+      "step": 10750
+    },
+    {
+      "epoch": 12.241389126103046,
+      "grad_norm": 6.726789951324463,
+      "learning_rate": 0.00014174115607562708,
+      "loss": 5.6524711608886715,
+      "step": 10760
+    },
+    {
+      "epoch": 12.252775405636209,
+      "grad_norm": 7.368878364562988,
+      "learning_rate": 0.0001416869819600195,
+      "loss": 5.634532928466797,
+      "step": 10770
+    },
+    {
+      "epoch": 12.264161685169372,
+      "grad_norm": 7.050151348114014,
+      "learning_rate": 0.00014163280784441196,
+      "loss": 5.614359283447266,
+      "step": 10780
+    },
+    {
+      "epoch": 12.275547964702534,
+      "grad_norm": 7.243103981018066,
+      "learning_rate": 0.00014157863372880438,
+      "loss": 5.495789337158203,
+      "step": 10790
+    },
+    {
+      "epoch": 12.286934244235695,
+      "grad_norm": 7.616795063018799,
+      "learning_rate": 0.00014152445961319684,
+      "loss": 5.778979110717773,
+      "step": 10800
+    },
+    {
+      "epoch": 12.298320523768858,
+      "grad_norm": 7.822340488433838,
+      "learning_rate": 0.00014147028549758926,
+      "loss": 5.528537750244141,
+      "step": 10810
+    },
+    {
+      "epoch": 12.30970680330202,
+      "grad_norm": 7.852270603179932,
+      "learning_rate": 0.00014141611138198168,
+      "loss": 5.403163909912109,
+      "step": 10820
+    },
+    {
+      "epoch": 12.321093082835183,
+      "grad_norm": 7.205753326416016,
+      "learning_rate": 0.00014136193726637414,
+      "loss": 5.604093551635742,
+      "step": 10830
+    },
+    {
+      "epoch": 12.332479362368346,
+      "grad_norm": 6.92218017578125,
+      "learning_rate": 0.00014130776315076656,
+      "loss": 5.867033004760742,
+      "step": 10840
+    },
+    {
+      "epoch": 12.343865641901509,
+      "grad_norm": 7.30786657333374,
+      "learning_rate": 0.000141253589035159,
+      "loss": 5.671744537353516,
+      "step": 10850
+    },
+    {
+      "epoch": 12.355251921434672,
+      "grad_norm": 7.897021293640137,
+      "learning_rate": 0.00014119941491955144,
+      "loss": 5.346432495117187,
+      "step": 10860
+    },
+    {
+      "epoch": 12.366638200967834,
+      "grad_norm": 7.8339738845825195,
+      "learning_rate": 0.0001411452408039439,
+      "loss": 5.636806488037109,
+      "step": 10870
+    },
+    {
+      "epoch": 12.378024480500997,
+      "grad_norm": 7.503673553466797,
+      "learning_rate": 0.0001410910666883363,
+      "loss": 5.6904136657714846,
+      "step": 10880
+    },
+    {
+      "epoch": 12.389410760034158,
+      "grad_norm": 7.782256603240967,
+      "learning_rate": 0.00014103689257272876,
+      "loss": 5.786883544921875,
+      "step": 10890
+    },
+    {
+      "epoch": 12.40079703956732,
+      "grad_norm": 7.115951061248779,
+      "learning_rate": 0.0001409827184571212,
+      "loss": 5.459210968017578,
+      "step": 10900
+    },
+    {
+      "epoch": 12.412183319100484,
+      "grad_norm": 7.4705281257629395,
+      "learning_rate": 0.00014092854434151364,
+      "loss": 5.9279014587402346,
+      "step": 10910
+    },
+    {
+      "epoch": 12.423569598633646,
+      "grad_norm": 7.553955554962158,
+      "learning_rate": 0.00014087437022590606,
+      "loss": 5.823044204711914,
+      "step": 10920
+    },
+    {
+      "epoch": 12.434955878166809,
+      "grad_norm": 7.552109241485596,
+      "learning_rate": 0.00014082019611029852,
+      "loss": 5.539102172851562,
+      "step": 10930
+    },
+    {
+      "epoch": 12.446342157699972,
+      "grad_norm": 6.597175598144531,
+      "learning_rate": 0.00014076602199469094,
+      "loss": 6.030828857421875,
+      "step": 10940
+    },
+    {
+      "epoch": 12.457728437233134,
+      "grad_norm": 6.793281078338623,
+      "learning_rate": 0.0001407118478790834,
+      "loss": 5.719361114501953,
+      "step": 10950
+    },
+    {
+      "epoch": 12.469114716766297,
+      "grad_norm": 7.471744537353516,
+      "learning_rate": 0.00014065767376347582,
+      "loss": 5.419223403930664,
+      "step": 10960
+    },
+    {
+      "epoch": 12.48050099629946,
+      "grad_norm": 6.892999649047852,
+      "learning_rate": 0.00014060349964786824,
+      "loss": 5.7255298614501955,
+      "step": 10970
+    },
+    {
+      "epoch": 12.491887275832621,
+      "grad_norm": 7.7696533203125,
+      "learning_rate": 0.0001405493255322607,
+      "loss": 6.058211898803711,
+      "step": 10980
+    },
+    {
+      "epoch": 12.503273555365784,
+      "grad_norm": 7.769250869750977,
+      "learning_rate": 0.00014049515141665312,
+      "loss": 5.425541687011719,
+      "step": 10990
+    },
+    {
+      "epoch": 12.514659834898946,
+      "grad_norm": 7.765056610107422,
+      "learning_rate": 0.00014044097730104557,
+      "loss": 5.620434951782227,
+      "step": 11000
+    },
+    {
+      "epoch": 12.52604611443211,
+      "grad_norm": 7.338669300079346,
+      "learning_rate": 0.000140386803185438,
+      "loss": 5.756977081298828,
+      "step": 11010
+    },
+    {
+      "epoch": 12.537432393965272,
+      "grad_norm": 7.880843162536621,
+      "learning_rate": 0.00014033262906983044,
+      "loss": 5.7971046447753904,
+      "step": 11020
+    },
+    {
+      "epoch": 12.548818673498435,
+      "grad_norm": 7.511135101318359,
+      "learning_rate": 0.00014027845495422287,
+      "loss": 5.453466415405273,
+      "step": 11030
+    },
+    {
+      "epoch": 12.560204953031597,
+      "grad_norm": 7.005084037780762,
+      "learning_rate": 0.00014022428083861532,
+      "loss": 5.595069503784179,
+      "step": 11040
+    },
+    {
+      "epoch": 12.57159123256476,
+      "grad_norm": 7.802360534667969,
+      "learning_rate": 0.00014017010672300775,
+      "loss": 6.111663436889648,
+      "step": 11050
+    },
+    {
+      "epoch": 12.582977512097923,
+      "grad_norm": 7.821834564208984,
+      "learning_rate": 0.0001401159326074002,
+      "loss": 5.398255920410156,
+      "step": 11060
+    },
+    {
+      "epoch": 12.594363791631084,
+      "grad_norm": 7.565220832824707,
+      "learning_rate": 0.00014006175849179265,
+      "loss": 5.964350891113281,
+      "step": 11070
+    },
+    {
+      "epoch": 12.605750071164247,
+      "grad_norm": 7.540068626403809,
+      "learning_rate": 0.00014000758437618507,
+      "loss": 5.82092399597168,
+      "step": 11080
+    },
+    {
+      "epoch": 12.61713635069741,
+      "grad_norm": 7.710811138153076,
+      "learning_rate": 0.00013995341026057752,
+      "loss": 5.725724411010742,
+      "step": 11090
+    },
+    {
+      "epoch": 12.628522630230572,
+      "grad_norm": 6.454217433929443,
+      "learning_rate": 0.00013989923614496995,
+      "loss": 5.528153991699218,
+      "step": 11100
+    },
+    {
+      "epoch": 12.639908909763735,
+      "grad_norm": 6.6477837562561035,
+      "learning_rate": 0.00013984506202936237,
+      "loss": 5.676776885986328,
+      "step": 11110
+    },
+    {
+      "epoch": 12.651295189296897,
+      "grad_norm": 6.51104736328125,
+      "learning_rate": 0.0001397908879137548,
+      "loss": 5.7555488586425785,
+      "step": 11120
+    },
+    {
+      "epoch": 12.66268146883006,
+      "grad_norm": 7.661293029785156,
+      "learning_rate": 0.00013973671379814725,
+      "loss": 5.601126861572266,
+      "step": 11130
+    },
+    {
+      "epoch": 12.674067748363223,
+      "grad_norm": 7.333502292633057,
+      "learning_rate": 0.00013968253968253967,
+      "loss": 5.845965957641601,
+      "step": 11140
+    },
+    {
+      "epoch": 12.685454027896386,
+      "grad_norm": 8.663681030273438,
+      "learning_rate": 0.00013962836556693213,
+      "loss": 5.837159729003906,
+      "step": 11150
+    },
+    {
+      "epoch": 12.696840307429547,
+      "grad_norm": 6.259514808654785,
+      "learning_rate": 0.00013957419145132455,
+      "loss": 5.721532821655273,
+      "step": 11160
+    },
+    {
+      "epoch": 12.70822658696271,
+      "grad_norm": 6.932978630065918,
+      "learning_rate": 0.000139520017335717,
+      "loss": 5.9294178009033205,
+      "step": 11170
+    },
+    {
+      "epoch": 12.719612866495872,
+      "grad_norm": 7.781189441680908,
+      "learning_rate": 0.00013946584322010945,
+      "loss": 5.403998184204101,
+      "step": 11180
+    },
+    {
+      "epoch": 12.730999146029035,
+      "grad_norm": 7.4265336990356445,
+      "learning_rate": 0.00013941166910450188,
+      "loss": 5.511170196533203,
+      "step": 11190
+    },
+    {
+      "epoch": 12.742385425562198,
+      "grad_norm": 8.001691818237305,
+      "learning_rate": 0.00013935749498889433,
+      "loss": 5.863291931152344,
+      "step": 11200
+    },
+    {
+      "epoch": 12.75377170509536,
+      "grad_norm": 6.850292682647705,
+      "learning_rate": 0.00013930332087328675,
+      "loss": 5.665401458740234,
+      "step": 11210
+    },
+    {
+      "epoch": 12.765157984628523,
+      "grad_norm": 8.020200729370117,
+      "learning_rate": 0.0001392491467576792,
+      "loss": 5.8318031311035154,
+      "step": 11220
+    },
+    {
+      "epoch": 12.776544264161686,
+      "grad_norm": 7.251684188842773,
+      "learning_rate": 0.00013919497264207163,
+      "loss": 5.751366806030274,
+      "step": 11230
+    },
+    {
+      "epoch": 12.787930543694848,
+      "grad_norm": 8.099172592163086,
+      "learning_rate": 0.00013914079852646405,
+      "loss": 5.816202545166016,
+      "step": 11240
+    },
+    {
+      "epoch": 12.79931682322801,
+      "grad_norm": 7.1152024269104,
+      "learning_rate": 0.00013908662441085648,
+      "loss": 5.4952960968017575,
+      "step": 11250
+    },
+    {
+      "epoch": 12.810703102761172,
+      "grad_norm": 7.486706256866455,
+      "learning_rate": 0.00013903245029524893,
+      "loss": 5.637163162231445,
+      "step": 11260
+    },
+    {
+      "epoch": 12.822089382294335,
+      "grad_norm": 7.652496337890625,
+      "learning_rate": 0.00013897827617964135,
+      "loss": 5.713247680664063,
+      "step": 11270
+    },
+    {
+      "epoch": 12.833475661827498,
+      "grad_norm": 7.462140083312988,
+      "learning_rate": 0.0001389241020640338,
+      "loss": 5.825135040283203,
+      "step": 11280
+    },
+    {
+      "epoch": 12.84486194136066,
+      "grad_norm": 7.004937648773193,
+      "learning_rate": 0.00013886992794842626,
+      "loss": 5.827632904052734,
+      "step": 11290
+    },
+    {
+      "epoch": 12.856248220893823,
+      "grad_norm": 7.434920310974121,
+      "learning_rate": 0.00013881575383281868,
+      "loss": 5.4522960662841795,
+      "step": 11300
+    },
+    {
+      "epoch": 12.867634500426986,
+      "grad_norm": 6.466522693634033,
+      "learning_rate": 0.00013876157971721113,
+      "loss": 5.210875701904297,
+      "step": 11310
+    },
+    {
+      "epoch": 12.879020779960149,
+      "grad_norm": 7.1903510093688965,
+      "learning_rate": 0.00013870740560160356,
+      "loss": 5.339904022216797,
+      "step": 11320
+    },
+    {
+      "epoch": 12.890407059493311,
+      "grad_norm": 7.564988613128662,
+      "learning_rate": 0.000138653231485996,
+      "loss": 5.6416679382324215,
+      "step": 11330
+    },
+    {
+      "epoch": 12.901793339026472,
+      "grad_norm": 8.242656707763672,
+      "learning_rate": 0.00013859905737038843,
+      "loss": 5.6682594299316404,
+      "step": 11340
+    },
+    {
+      "epoch": 12.913179618559635,
+      "grad_norm": 7.32057523727417,
+      "learning_rate": 0.00013854488325478089,
+      "loss": 5.620843124389649,
+      "step": 11350
+    },
+    {
+      "epoch": 12.924565898092798,
+      "grad_norm": 7.37946891784668,
+      "learning_rate": 0.0001384907091391733,
+      "loss": 5.613275527954102,
+      "step": 11360
+    },
+    {
+      "epoch": 12.93595217762596,
+      "grad_norm": 7.773562431335449,
+      "learning_rate": 0.00013843653502356576,
+      "loss": 5.346442794799804,
+      "step": 11370
+    },
+    {
+      "epoch": 12.947338457159123,
+      "grad_norm": 7.526298999786377,
+      "learning_rate": 0.00013838236090795819,
+      "loss": 5.606759643554687,
+      "step": 11380
+    },
+    {
+      "epoch": 12.958724736692286,
+      "grad_norm": 7.332315921783447,
+      "learning_rate": 0.0001383281867923506,
+      "loss": 5.419976425170899,
+      "step": 11390
+    },
+    {
+      "epoch": 12.970111016225449,
+      "grad_norm": 7.382599830627441,
+      "learning_rate": 0.00013827401267674304,
+      "loss": 5.998534393310547,
+      "step": 11400
+    },
+    {
+      "epoch": 12.981497295758611,
+      "grad_norm": 7.097250461578369,
+      "learning_rate": 0.0001382198385611355,
+      "loss": 5.930072021484375,
+      "step": 11410
+    },
+    {
+      "epoch": 12.992883575291774,
+      "grad_norm": 6.900557041168213,
+      "learning_rate": 0.00013816566444552794,
+      "loss": 5.5991252899169925,
+      "step": 11420
+    },
+    {
+      "epoch": 13.00341588385995,
+      "grad_norm": 6.776562213897705,
+      "learning_rate": 0.00013811149032992036,
+      "loss": 4.849653244018555,
+      "step": 11430
+    },
+    {
+      "epoch": 13.014802163393112,
+      "grad_norm": 7.419433116912842,
+      "learning_rate": 0.00013805731621431281,
+      "loss": 5.020238876342773,
+      "step": 11440
+    },
+    {
+      "epoch": 13.026188442926275,
+      "grad_norm": 6.931766986846924,
+      "learning_rate": 0.00013800314209870524,
+      "loss": 5.281552124023437,
+      "step": 11450
+    },
+    {
+      "epoch": 13.037574722459436,
+      "grad_norm": 7.03234338760376,
+      "learning_rate": 0.0001379489679830977,
+      "loss": 5.1979625701904295,
+      "step": 11460
+    },
+    {
+      "epoch": 13.048961001992598,
+      "grad_norm": 7.978625774383545,
+      "learning_rate": 0.00013789479386749011,
+      "loss": 5.348671722412109,
+      "step": 11470
+    },
+    {
+      "epoch": 13.060347281525761,
+      "grad_norm": 7.255367755889893,
+      "learning_rate": 0.00013784061975188257,
+      "loss": 5.117027282714844,
+      "step": 11480
+    },
+    {
+      "epoch": 13.071733561058924,
+      "grad_norm": 7.438676834106445,
+      "learning_rate": 0.000137786445636275,
+      "loss": 5.207336044311523,
+      "step": 11490
+    },
+    {
+      "epoch": 13.083119840592087,
+      "grad_norm": 6.7722578048706055,
+      "learning_rate": 0.00013773227152066744,
+      "loss": 4.885826110839844,
+      "step": 11500
+    },
+    {
+      "epoch": 13.09450612012525,
+      "grad_norm": 6.830091953277588,
+      "learning_rate": 0.00013767809740505987,
+      "loss": 4.998377990722656,
+      "step": 11510
+    },
+    {
+      "epoch": 13.105892399658412,
+      "grad_norm": 7.623549461364746,
+      "learning_rate": 0.00013762392328945232,
+      "loss": 5.187086486816407,
+      "step": 11520
+    },
+    {
+      "epoch": 13.117278679191575,
+      "grad_norm": 8.049860954284668,
+      "learning_rate": 0.00013756974917384474,
+      "loss": 4.911079788208008,
+      "step": 11530
+    },
+    {
+      "epoch": 13.128664958724737,
+      "grad_norm": 7.040311336517334,
+      "learning_rate": 0.00013751557505823717,
+      "loss": 5.190642547607422,
+      "step": 11540
+    },
+    {
+      "epoch": 13.140051238257898,
+      "grad_norm": 7.5759806632995605,
+      "learning_rate": 0.00013746140094262962,
+      "loss": 5.331892776489258,
+      "step": 11550
+    },
+    {
+      "epoch": 13.151437517791061,
+      "grad_norm": 7.5590362548828125,
+      "learning_rate": 0.00013740722682702204,
+      "loss": 5.309208679199219,
+      "step": 11560
+    },
+    {
+      "epoch": 13.162823797324224,
+      "grad_norm": 7.6602559089660645,
+      "learning_rate": 0.0001373530527114145,
+      "loss": 5.0000556945800785,
+      "step": 11570
+    },
+    {
+      "epoch": 13.174210076857387,
+      "grad_norm": 6.687933921813965,
+      "learning_rate": 0.00013729887859580692,
+      "loss": 5.226931381225586,
+      "step": 11580
+    },
+    {
+      "epoch": 13.18559635639055,
+      "grad_norm": 7.443079471588135,
+      "learning_rate": 0.00013724470448019937,
+      "loss": 5.208231735229492,
+      "step": 11590
+    },
+    {
+      "epoch": 13.196982635923712,
+      "grad_norm": 7.559839725494385,
+      "learning_rate": 0.0001371905303645918,
+      "loss": 5.439892578125,
+      "step": 11600
+    },
+    {
+      "epoch": 13.208368915456875,
+      "grad_norm": 7.104702472686768,
+      "learning_rate": 0.00013713635624898425,
+      "loss": 5.115517807006836,
+      "step": 11610
+    },
+    {
+      "epoch": 13.219755194990038,
+      "grad_norm": NaN,
+      "learning_rate": 0.00013708218213337667,
+      "loss": 4.99559440612793,
+      "step": 11620
+    },
+    {
+      "epoch": 13.2311414745232,
+      "grad_norm": 7.759593486785889,
+      "learning_rate": 0.00013703342542932987,
+      "loss": 5.390767288208008,
+      "step": 11630
+    },
+    {
+      "epoch": 13.242527754056361,
+      "grad_norm": 7.8164286613464355,
+      "learning_rate": 0.00013697925131372232,
+      "loss": 5.311045455932617,
+      "step": 11640
+    },
+    {
+      "epoch": 13.253914033589524,
+      "grad_norm": 6.780811309814453,
+      "learning_rate": 0.00013692507719811474,
+      "loss": 5.088868713378906,
+      "step": 11650
+    },
+    {
+      "epoch": 13.265300313122687,
+      "grad_norm": 7.4553022384643555,
+      "learning_rate": 0.0001368709030825072,
+      "loss": 5.636979675292968,
+      "step": 11660
+    },
+    {
+      "epoch": 13.27668659265585,
+      "grad_norm": 7.379472255706787,
+      "learning_rate": 0.00013681672896689962,
+      "loss": 5.3572509765625,
+      "step": 11670
+    },
+    {
+      "epoch": 13.288072872189012,
+      "grad_norm": 7.313671588897705,
+      "learning_rate": 0.00013676255485129204,
+      "loss": 5.362029647827148,
+      "step": 11680
+    },
+    {
+      "epoch": 13.299459151722175,
+      "grad_norm": 7.911259174346924,
+      "learning_rate": 0.0001367083807356845,
+      "loss": 5.288186645507812,
+      "step": 11690
+    },
+    {
+      "epoch": 13.310845431255338,
+      "grad_norm": 7.032965183258057,
+      "learning_rate": 0.00013665420662007692,
+      "loss": 5.303681182861328,
+      "step": 11700
+    },
+    {
+      "epoch": 13.3222317107885,
+      "grad_norm": 8.510554313659668,
+      "learning_rate": 0.00013660003250446937,
+      "loss": 5.175273513793945,
+      "step": 11710
+    },
+    {
+      "epoch": 13.333617990321663,
+      "grad_norm": 7.989207744598389,
+      "learning_rate": 0.0001365458583888618,
+      "loss": 4.8652587890625,
+      "step": 11720
+    },
+    {
+      "epoch": 13.345004269854824,
+      "grad_norm": 8.093276977539062,
+      "learning_rate": 0.00013649168427325425,
+      "loss": 5.2994426727294925,
+      "step": 11730
+    },
+    {
+      "epoch": 13.356390549387987,
+      "grad_norm": 6.790765285491943,
+      "learning_rate": 0.00013643751015764667,
+      "loss": 5.194287109375,
+      "step": 11740
+    },
+    {
+      "epoch": 13.36777682892115,
+      "grad_norm": 7.429324626922607,
+      "learning_rate": 0.00013638333604203912,
+      "loss": 5.5465232849121096,
+      "step": 11750
+    },
+    {
+      "epoch": 13.379163108454312,
+      "grad_norm": 7.065948486328125,
+      "learning_rate": 0.00013632916192643155,
+      "loss": 5.191884613037109,
+      "step": 11760
+    },
+    {
+      "epoch": 13.390549387987475,
+      "grad_norm": 7.3118510246276855,
+      "learning_rate": 0.000136274987810824,
+      "loss": 5.531578826904297,
+      "step": 11770
+    },
+    {
+      "epoch": 13.401935667520638,
+      "grad_norm": 7.449267864227295,
+      "learning_rate": 0.00013622081369521645,
+      "loss": 5.3741718292236325,
+      "step": 11780
+    },
+    {
+      "epoch": 13.4133219470538,
+      "grad_norm": 7.919931411743164,
+      "learning_rate": 0.00013616663957960888,
+      "loss": 5.327036666870117,
+      "step": 11790
+    },
+    {
+      "epoch": 13.424708226586963,
+      "grad_norm": 7.673199653625488,
+      "learning_rate": 0.00013611246546400133,
+      "loss": 5.344607925415039,
+      "step": 11800
+    },
+    {
+      "epoch": 13.436094506120126,
+      "grad_norm": 7.837534427642822,
+      "learning_rate": 0.00013605829134839375,
+      "loss": 5.430044174194336,
+      "step": 11810
+    },
+    {
+      "epoch": 13.447480785653287,
+      "grad_norm": 7.384720325469971,
+      "learning_rate": 0.00013600411723278618,
+      "loss": 5.121067810058594,
+      "step": 11820
+    },
+    {
+      "epoch": 13.45886706518645,
+      "grad_norm": 7.407566070556641,
+      "learning_rate": 0.0001359499431171786,
+      "loss": 5.254208755493164,
+      "step": 11830
+    },
+    {
+      "epoch": 13.470253344719612,
+      "grad_norm": 7.422298431396484,
+      "learning_rate": 0.00013589576900157105,
+      "loss": 5.210581207275391,
+      "step": 11840
+    },
+    {
+      "epoch": 13.481639624252775,
+      "grad_norm": 7.2973737716674805,
+      "learning_rate": 0.00013584159488596348,
+      "loss": 5.4987037658691404,
+      "step": 11850
+    },
+    {
+      "epoch": 13.493025903785938,
+      "grad_norm": 7.023255348205566,
+      "learning_rate": 0.00013578742077035593,
+      "loss": 5.042602920532227,
+      "step": 11860
+    },
+    {
+      "epoch": 13.5044121833191,
+      "grad_norm": 7.912652492523193,
+      "learning_rate": 0.00013573324665474835,
+      "loss": 5.281571578979492,
+      "step": 11870
+    },
+    {
+      "epoch": 13.515798462852263,
+      "grad_norm": 7.739786148071289,
+      "learning_rate": 0.0001356790725391408,
+      "loss": 5.43682861328125,
+      "step": 11880
+    },
+    {
+      "epoch": 13.527184742385426,
+      "grad_norm": 7.652828216552734,
+      "learning_rate": 0.00013562489842353326,
+      "loss": 5.211148071289062,
+      "step": 11890
+    },
+    {
+      "epoch": 13.538571021918589,
+      "grad_norm": 8.158793449401855,
+      "learning_rate": 0.00013557072430792568,
+      "loss": 5.4678295135498045,
+      "step": 11900
+    },
+    {
+      "epoch": 13.54995730145175,
+      "grad_norm": 7.335878372192383,
+      "learning_rate": 0.00013551655019231813,
+      "loss": 5.306196975708008,
+      "step": 11910
+    },
+    {
+      "epoch": 13.561343580984913,
+      "grad_norm": 7.801717758178711,
+      "learning_rate": 0.00013546237607671056,
+      "loss": 5.475741195678711,
+      "step": 11920
+    },
+    {
+      "epoch": 13.572729860518075,
+      "grad_norm": 8.175219535827637,
+      "learning_rate": 0.000135408201961103,
+      "loss": 5.457925796508789,
+      "step": 11930
+    },
+    {
+      "epoch": 13.584116140051238,
+      "grad_norm": 6.884897232055664,
+      "learning_rate": 0.00013535402784549543,
+      "loss": 5.128502655029297,
+      "step": 11940
+    },
+    {
+      "epoch": 13.5955024195844,
+      "grad_norm": 6.6634521484375,
+      "learning_rate": 0.00013529985372988788,
+      "loss": 5.319187164306641,
+      "step": 11950
+    },
+    {
+      "epoch": 13.606888699117563,
+      "grad_norm": 8.205178260803223,
+      "learning_rate": 0.0001352456796142803,
+      "loss": 5.198386001586914,
+      "step": 11960
+    },
+    {
+      "epoch": 13.618274978650726,
+      "grad_norm": 6.997448444366455,
+      "learning_rate": 0.00013519150549867273,
+      "loss": 5.356737899780273,
+      "step": 11970
+    },
+    {
+      "epoch": 13.629661258183889,
+      "grad_norm": 7.976472854614258,
+      "learning_rate": 0.00013513733138306516,
+      "loss": 5.224060440063477,
+      "step": 11980
+    },
+    {
+      "epoch": 13.641047537717052,
+      "grad_norm": 7.6008477210998535,
+      "learning_rate": 0.0001350831572674576,
+      "loss": 5.205660629272461,
+      "step": 11990
+    },
+    {
+      "epoch": 13.652433817250213,
+      "grad_norm": 8.134635925292969,
+      "learning_rate": 0.00013502898315185003,
+      "loss": 5.284549331665039,
+      "step": 12000
+    },
+    {
+      "epoch": 13.663820096783375,
+      "grad_norm": 7.697743892669678,
+      "learning_rate": 0.00013497480903624249,
+      "loss": 5.365811157226562,
+      "step": 12010
+    },
+    {
+      "epoch": 13.675206376316538,
+      "grad_norm": 6.987123966217041,
+      "learning_rate": 0.00013492063492063494,
+      "loss": 5.420026779174805,
+      "step": 12020
+    },
+    {
+      "epoch": 13.6865926558497,
+      "grad_norm": 7.670924663543701,
+      "learning_rate": 0.00013486646080502736,
+      "loss": 5.44122543334961,
+      "step": 12030
+    },
+    {
+      "epoch": 13.697978935382864,
+      "grad_norm": 7.6146955490112305,
+      "learning_rate": 0.0001348122866894198,
+      "loss": 5.368587493896484,
+      "step": 12040
+    },
+    {
+      "epoch": 13.709365214916026,
+      "grad_norm": 7.290314674377441,
+      "learning_rate": 0.00013475811257381224,
+      "loss": 4.899549102783203,
+      "step": 12050
+    },
+    {
+      "epoch": 13.720751494449189,
+      "grad_norm": 7.6227569580078125,
+      "learning_rate": 0.0001347039384582047,
+      "loss": 5.226024627685547,
+      "step": 12060
+    },
+    {
+      "epoch": 13.732137773982352,
+      "grad_norm": 8.122727394104004,
+      "learning_rate": 0.0001346497643425971,
+      "loss": 5.342987060546875,
+      "step": 12070
+    },
+    {
+      "epoch": 13.743524053515515,
+      "grad_norm": 7.191249847412109,
+      "learning_rate": 0.00013459559022698956,
+      "loss": 4.999538421630859,
+      "step": 12080
+    },
+    {
+      "epoch": 13.754910333048677,
+      "grad_norm": 7.330814361572266,
+      "learning_rate": 0.000134541416111382,
+      "loss": 5.499901199340821,
+      "step": 12090
+    },
+    {
+      "epoch": 13.766296612581838,
+      "grad_norm": 7.275107383728027,
+      "learning_rate": 0.00013448724199577441,
+      "loss": 5.174909591674805,
+      "step": 12100
+    },
+    {
+      "epoch": 13.777682892115001,
+      "grad_norm": 7.071986675262451,
+      "learning_rate": 0.00013443306788016687,
+      "loss": 5.219038391113282,
+      "step": 12110
+    },
+    {
+      "epoch": 13.789069171648164,
+      "grad_norm": 6.817813396453857,
+      "learning_rate": 0.0001343788937645593,
+      "loss": 5.07507553100586,
+      "step": 12120
+    },
+    {
+      "epoch": 13.800455451181326,
+      "grad_norm": 7.419076442718506,
+      "learning_rate": 0.00013432471964895174,
+      "loss": 5.351544952392578,
+      "step": 12130
+    },
+    {
+      "epoch": 13.81184173071449,
+      "grad_norm": 7.717519283294678,
+      "learning_rate": 0.00013427054553334417,
+      "loss": 5.706594085693359,
+      "step": 12140
+    },
+    {
+      "epoch": 13.823228010247652,
+      "grad_norm": 7.404618263244629,
+      "learning_rate": 0.00013421637141773662,
+      "loss": 5.376433563232422,
+      "step": 12150
+    },
+    {
+      "epoch": 13.834614289780815,
+      "grad_norm": 7.158414363861084,
+      "learning_rate": 0.00013416219730212904,
+      "loss": 5.363436508178711,
+      "step": 12160
+    },
+    {
+      "epoch": 13.846000569313977,
+      "grad_norm": 7.699437618255615,
+      "learning_rate": 0.0001341080231865215,
+      "loss": 5.6122314453125,
+      "step": 12170
+    },
+    {
+      "epoch": 13.857386848847138,
+      "grad_norm": 7.237250804901123,
+      "learning_rate": 0.00013405384907091392,
+      "loss": 5.395546340942383,
+      "step": 12180
+    },
+    {
+      "epoch": 13.868773128380301,
+      "grad_norm": 7.333144187927246,
+      "learning_rate": 0.00013399967495530637,
+      "loss": 5.211500549316407,
+      "step": 12190
+    },
+    {
+      "epoch": 13.880159407913464,
+      "grad_norm": 6.954226493835449,
+      "learning_rate": 0.0001339455008396988,
+      "loss": 5.263496017456054,
+      "step": 12200
+    },
+    {
+      "epoch": 13.891545687446627,
+      "grad_norm": 7.646538734436035,
+      "learning_rate": 0.00013389132672409125,
+      "loss": 5.6256969451904295,
+      "step": 12210
+    },
+    {
+      "epoch": 13.90293196697979,
+      "grad_norm": 7.714359760284424,
+      "learning_rate": 0.00013383715260848367,
+      "loss": 5.669702529907227,
+      "step": 12220
+    },
+    {
+      "epoch": 13.914318246512952,
+      "grad_norm": 7.628199577331543,
+      "learning_rate": 0.00013378297849287612,
+      "loss": 5.399072265625,
+      "step": 12230
+    },
+    {
+      "epoch": 13.925704526046115,
+      "grad_norm": 8.523447036743164,
+      "learning_rate": 0.00013372880437726855,
+      "loss": 5.366733169555664,
+      "step": 12240
+    },
+    {
+      "epoch": 13.937090805579277,
+      "grad_norm": 7.897779941558838,
+      "learning_rate": 0.00013367463026166097,
+      "loss": 5.20464973449707,
+      "step": 12250
+    },
+    {
+      "epoch": 13.94847708511244,
+      "grad_norm": 7.602762699127197,
+      "learning_rate": 0.00013362045614605342,
+      "loss": 5.602539443969727,
+      "step": 12260
+    },
+    {
+      "epoch": 13.959863364645603,
+      "grad_norm": 7.468645095825195,
+      "learning_rate": 0.00013356628203044585,
+      "loss": 5.513700485229492,
+      "step": 12270
+    },
+    {
+      "epoch": 13.971249644178764,
+      "grad_norm": 7.841520309448242,
+      "learning_rate": 0.0001335121079148383,
+      "loss": 5.51887321472168,
+      "step": 12280
+    },
+    {
+      "epoch": 13.982635923711927,
+      "grad_norm": 7.047943115234375,
+      "learning_rate": 0.00013345793379923072,
+      "loss": 5.111991500854492,
+      "step": 12290
+    },
+    {
+      "epoch": 13.99402220324509,
+      "grad_norm": 6.829543113708496,
+      "learning_rate": 0.00013340375968362317,
+      "loss": 5.058186340332031,
+      "step": 12300
+    },
+    {
+      "epoch": 14.004554511813264,
+      "grad_norm": 7.048587799072266,
+      "learning_rate": 0.0001333495855680156,
+      "loss": 4.963529586791992,
+      "step": 12310
+    },
+    {
+      "epoch": 14.015940791346427,
+      "grad_norm": 6.852132797241211,
+      "learning_rate": 0.00013329541145240805,
+      "loss": 4.623985290527344,
+      "step": 12320
+    },
+    {
+      "epoch": 14.02732707087959,
+      "grad_norm": 7.333976745605469,
+      "learning_rate": 0.00013324123733680047,
+      "loss": 5.1697029113769535,
+      "step": 12330
+    },
+    {
+      "epoch": 14.038713350412753,
+      "grad_norm": 8.014049530029297,
+      "learning_rate": 0.00013318706322119293,
+      "loss": 4.859414672851562,
+      "step": 12340
+    },
+    {
+      "epoch": 14.050099629945915,
+      "grad_norm": 7.355411529541016,
+      "learning_rate": 0.00013313288910558538,
+      "loss": 4.995968627929687,
+      "step": 12350
+    },
+    {
+      "epoch": 14.061485909479078,
+      "grad_norm": 7.705524921417236,
+      "learning_rate": 0.0001330787149899778,
+      "loss": 5.119239807128906,
+      "step": 12360
+    },
+    {
+      "epoch": 14.07287218901224,
+      "grad_norm": 7.468197822570801,
+      "learning_rate": 0.00013302454087437023,
+      "loss": 5.107388305664062,
+      "step": 12370
+    },
+    {
+      "epoch": 14.084258468545404,
+      "grad_norm": 6.574582099914551,
+      "learning_rate": 0.00013297036675876268,
+      "loss": 4.7201496124267575,
+      "step": 12380
+    },
+    {
+      "epoch": 14.095644748078564,
+      "grad_norm": 6.222922325134277,
+      "learning_rate": 0.0001329161926431551,
+      "loss": 4.612395477294922,
+      "step": 12390
+    },
+    {
+      "epoch": 14.107031027611727,
+      "grad_norm": 7.6845011711120605,
+      "learning_rate": 0.00013286201852754753,
+      "loss": 5.049299621582032,
+      "step": 12400
+    },
+    {
+      "epoch": 14.11841730714489,
+      "grad_norm": 7.420602798461914,
+      "learning_rate": 0.00013280784441193998,
+      "loss": 4.884346389770508,
+      "step": 12410
+    },
+    {
+      "epoch": 14.129803586678053,
+      "grad_norm": 7.716628074645996,
+      "learning_rate": 0.0001327536702963324,
+      "loss": 4.882117080688476,
+      "step": 12420
+    },
+    {
+      "epoch": 14.141189866211215,
+      "grad_norm": 7.8805646896362305,
+      "learning_rate": 0.00013269949618072486,
+      "loss": 4.686184692382812,
+      "step": 12430
+    },
+    {
+      "epoch": 14.152576145744378,
+      "grad_norm": 7.1464948654174805,
+      "learning_rate": 0.00013264532206511728,
+      "loss": 4.810683822631836,
+      "step": 12440
+    },
+    {
+      "epoch": 14.163962425277541,
+      "grad_norm": 7.473194599151611,
+      "learning_rate": 0.00013259114794950973,
+      "loss": 4.700112533569336,
+      "step": 12450
+    },
+    {
+      "epoch": 14.175348704810704,
+      "grad_norm": 6.588601589202881,
+      "learning_rate": 0.00013253697383390218,
+      "loss": 4.817057037353516,
+      "step": 12460
+    },
+    {
+      "epoch": 14.186734984343866,
+      "grad_norm": 7.659865856170654,
+      "learning_rate": 0.0001324827997182946,
+      "loss": 4.8424022674560545,
+      "step": 12470
+    },
+    {
+      "epoch": 14.198121263877027,
+      "grad_norm": 7.2345194816589355,
+      "learning_rate": 0.00013242862560268706,
+      "loss": 4.720790100097656,
+      "step": 12480
+    },
+    {
+      "epoch": 14.20950754341019,
+      "grad_norm": 7.407914161682129,
+      "learning_rate": 0.00013237445148707948,
+      "loss": 4.918620681762695,
+      "step": 12490
+    },
+    {
+      "epoch": 14.220893822943353,
+      "grad_norm": 8.086915969848633,
+      "learning_rate": 0.00013232027737147193,
+      "loss": 5.120816421508789,
+      "step": 12500
+    },
+    {
+      "epoch": 14.232280102476516,
+      "grad_norm": 7.739943981170654,
+      "learning_rate": 0.00013226610325586436,
+      "loss": 4.804103851318359,
+      "step": 12510
+    },
+    {
+      "epoch": 14.243666382009678,
+      "grad_norm": 6.809656620025635,
+      "learning_rate": 0.00013221192914025678,
+      "loss": 4.88025016784668,
+      "step": 12520
+    },
+    {
+      "epoch": 14.255052661542841,
+      "grad_norm": 7.113979816436768,
+      "learning_rate": 0.0001321577550246492,
+      "loss": 5.060388565063477,
+      "step": 12530
+    },
+    {
+      "epoch": 14.266438941076004,
+      "grad_norm": 7.600307941436768,
+      "learning_rate": 0.00013210358090904166,
+      "loss": 4.81654167175293,
+      "step": 12540
+    },
+    {
+      "epoch": 14.277825220609166,
+      "grad_norm": 7.201952934265137,
+      "learning_rate": 0.00013204940679343408,
+      "loss": 4.638751220703125,
+      "step": 12550
+    },
+    {
+      "epoch": 14.28921150014233,
+      "grad_norm": 7.577877521514893,
+      "learning_rate": 0.00013199523267782654,
+      "loss": 5.024491500854492,
+      "step": 12560
+    },
+    {
+      "epoch": 14.30059777967549,
+      "grad_norm": 7.727479934692383,
+      "learning_rate": 0.000131941058562219,
+      "loss": 5.108154296875,
+      "step": 12570
+    },
+    {
+      "epoch": 14.311984059208653,
+      "grad_norm": 7.286412239074707,
+      "learning_rate": 0.0001318868844466114,
+      "loss": 5.149254608154297,
+      "step": 12580
+    },
+    {
+      "epoch": 14.323370338741816,
+      "grad_norm": 7.6460723876953125,
+      "learning_rate": 0.00013183271033100386,
+      "loss": 4.752541351318359,
+      "step": 12590
+    },
+    {
+      "epoch": 14.334756618274978,
+      "grad_norm": 7.025350093841553,
+      "learning_rate": 0.0001317785362153963,
+      "loss": 4.839694213867188,
+      "step": 12600
+    },
+    {
+      "epoch": 14.346142897808141,
+      "grad_norm": 7.519998073577881,
+      "learning_rate": 0.00013172436209978874,
+      "loss": 4.982858276367187,
+      "step": 12610
+    },
+    {
+      "epoch": 14.357529177341304,
+      "grad_norm": 7.484885215759277,
+      "learning_rate": 0.00013167018798418116,
+      "loss": 4.672926330566407,
+      "step": 12620
+    },
+    {
+      "epoch": 14.368915456874467,
+      "grad_norm": 8.382508277893066,
+      "learning_rate": 0.00013161601386857362,
+      "loss": 5.045550537109375,
+      "step": 12630
+    },
+    {
+      "epoch": 14.38030173640763,
+      "grad_norm": 7.656764030456543,
+      "learning_rate": 0.00013156183975296604,
+      "loss": 5.17280158996582,
+      "step": 12640
+    },
+    {
+      "epoch": 14.391688015940792,
+      "grad_norm": 7.431979179382324,
+      "learning_rate": 0.0001315076656373585,
+      "loss": 4.8266761779785154,
+      "step": 12650
+    },
+    {
+      "epoch": 14.403074295473953,
+      "grad_norm": 7.853142738342285,
+      "learning_rate": 0.00013145349152175092,
+      "loss": 4.80140266418457,
+      "step": 12660
+    },
+    {
+      "epoch": 14.414460575007116,
+      "grad_norm": 7.3754987716674805,
+      "learning_rate": 0.00013139931740614334,
+      "loss": 4.92738037109375,
+      "step": 12670
+    },
+    {
+      "epoch": 14.425846854540278,
+      "grad_norm": 6.4924116134643555,
+      "learning_rate": 0.0001313451432905358,
+      "loss": 4.55872802734375,
+      "step": 12680
+    },
+    {
+      "epoch": 14.437233134073441,
+      "grad_norm": 8.214619636535645,
+      "learning_rate": 0.00013129096917492822,
+      "loss": 4.9627235412597654,
+      "step": 12690
+    },
+    {
+      "epoch": 14.448619413606604,
+      "grad_norm": 7.3021559715271,
+      "learning_rate": 0.00013123679505932067,
+      "loss": 5.223368072509766,
+      "step": 12700
+    },
+    {
+      "epoch": 14.460005693139767,
+      "grad_norm": 7.899365425109863,
+      "learning_rate": 0.0001311826209437131,
+      "loss": 5.1207012176513675,
+      "step": 12710
+    },
+    {
+      "epoch": 14.47139197267293,
+      "grad_norm": 7.248983860015869,
+      "learning_rate": 0.00013112844682810554,
+      "loss": 4.907715606689453,
+      "step": 12720
+    },
+    {
+      "epoch": 14.482778252206092,
+      "grad_norm": 7.069028377532959,
+      "learning_rate": 0.00013107427271249797,
+      "loss": 4.999393463134766,
+      "step": 12730
+    },
+    {
+      "epoch": 14.494164531739255,
+      "grad_norm": 7.262919902801514,
+      "learning_rate": 0.00013102009859689042,
+      "loss": 4.917198562622071,
+      "step": 12740
+    },
+    {
+      "epoch": 14.505550811272418,
+      "grad_norm": 7.535858154296875,
+      "learning_rate": 0.00013096592448128284,
+      "loss": 4.810692596435547,
+      "step": 12750
+    },
+    {
+      "epoch": 14.516937090805579,
+      "grad_norm": 7.142795562744141,
+      "learning_rate": 0.0001309117503656753,
+      "loss": 5.230551910400391,
+      "step": 12760
+    },
+    {
+      "epoch": 14.528323370338741,
+      "grad_norm": 7.323062419891357,
+      "learning_rate": 0.00013085757625006772,
+      "loss": 5.168463134765625,
+      "step": 12770
+    },
+    {
+      "epoch": 14.539709649871904,
+      "grad_norm": 8.29926872253418,
+      "learning_rate": 0.00013080340213446017,
+      "loss": 5.1340171813964846,
+      "step": 12780
+    },
+    {
+      "epoch": 14.551095929405067,
+      "grad_norm": 8.317549705505371,
+      "learning_rate": 0.0001307492280188526,
+      "loss": 4.967045593261719,
+      "step": 12790
+    },
+    {
+      "epoch": 14.56248220893823,
+      "grad_norm": 7.742959499359131,
+      "learning_rate": 0.00013069505390324505,
+      "loss": 4.707772827148437,
+      "step": 12800
+    },
+    {
+      "epoch": 14.573868488471392,
+      "grad_norm": 7.057898044586182,
+      "learning_rate": 0.00013064087978763747,
+      "loss": 4.755387496948242,
+      "step": 12810
+    },
+    {
+      "epoch": 14.585254768004555,
+      "grad_norm": 7.152231216430664,
+      "learning_rate": 0.0001305867056720299,
+      "loss": 5.241105270385742,
+      "step": 12820
+    },
+    {
+      "epoch": 14.596641047537718,
+      "grad_norm": 8.085304260253906,
+      "learning_rate": 0.00013053253155642235,
+      "loss": 4.788356781005859,
+      "step": 12830
+    },
+    {
+      "epoch": 14.608027327070879,
+      "grad_norm": 7.510035037994385,
+      "learning_rate": 0.00013047835744081477,
+      "loss": 5.091434478759766,
+      "step": 12840
+    },
+    {
+      "epoch": 14.619413606604041,
+      "grad_norm": 7.749240875244141,
+      "learning_rate": 0.00013042418332520723,
+      "loss": 4.968235778808594,
+      "step": 12850
+    },
+    {
+      "epoch": 14.630799886137204,
+      "grad_norm": 6.937013626098633,
+      "learning_rate": 0.00013037000920959965,
+      "loss": 5.242551040649414,
+      "step": 12860
+    },
+    {
+      "epoch": 14.642186165670367,
+      "grad_norm": 7.75348424911499,
+      "learning_rate": 0.0001303158350939921,
+      "loss": 4.967795944213867,
+      "step": 12870
+    },
+    {
+      "epoch": 14.65357244520353,
+      "grad_norm": 8.130202293395996,
+      "learning_rate": 0.00013026166097838453,
+      "loss": 5.108961868286133,
+      "step": 12880
+    },
+    {
+      "epoch": 14.664958724736692,
+      "grad_norm": 6.920219898223877,
+      "learning_rate": 0.00013020748686277698,
+      "loss": 4.6893165588378904,
+      "step": 12890
+    },
+    {
+      "epoch": 14.676345004269855,
+      "grad_norm": 7.73007869720459,
+      "learning_rate": 0.0001301533127471694,
+      "loss": 5.106839370727539,
+      "step": 12900
+    },
+    {
+      "epoch": 14.687731283803018,
+      "grad_norm": 7.3374924659729,
+      "learning_rate": 0.00013009913863156185,
+      "loss": 5.362062072753906,
+      "step": 12910
+    },
+    {
+      "epoch": 14.69911756333618,
+      "grad_norm": 7.493285655975342,
+      "learning_rate": 0.0001300449645159543,
+      "loss": 5.187427139282226,
+      "step": 12920
+    },
+    {
+      "epoch": 14.710503842869343,
+      "grad_norm": 7.747722625732422,
+      "learning_rate": 0.00012999079040034673,
+      "loss": 4.937312316894531,
+      "step": 12930
+    },
+    {
+      "epoch": 14.721890122402504,
+      "grad_norm": 7.39769983291626,
+      "learning_rate": 0.00012993661628473915,
+      "loss": 5.141265869140625,
+      "step": 12940
+    },
+    {
+      "epoch": 14.733276401935667,
+      "grad_norm": 6.812027931213379,
+      "learning_rate": 0.00012988244216913158,
+      "loss": 5.5300537109375,
+      "step": 12950
+    },
+    {
+      "epoch": 14.74466268146883,
+      "grad_norm": 7.199182987213135,
+      "learning_rate": 0.00012982826805352403,
+      "loss": 4.976782608032226,
+      "step": 12960
+    },
+    {
+      "epoch": 14.756048961001992,
+      "grad_norm": 7.161105155944824,
+      "learning_rate": 0.00012977409393791645,
+      "loss": 4.928173828125,
+      "step": 12970
+    },
+    {
+      "epoch": 14.767435240535155,
+      "grad_norm": 8.261829376220703,
+      "learning_rate": 0.0001297199198223089,
+      "loss": 4.858361053466797,
+      "step": 12980
+    },
+    {
+      "epoch": 14.778821520068318,
+      "grad_norm": 6.941257953643799,
+      "learning_rate": 0.00012966574570670133,
+      "loss": 4.880601119995117,
+      "step": 12990
+    },
+    {
+      "epoch": 14.79020779960148,
+      "grad_norm": 7.993049144744873,
+      "learning_rate": 0.00012961157159109378,
+      "loss": 5.036775970458985,
+      "step": 13000
+    },
+    {
+      "epoch": 14.801594079134643,
+      "grad_norm": 8.901403427124023,
+      "learning_rate": 0.0001295573974754862,
+      "loss": 5.106240844726562,
+      "step": 13010
+    },
+    {
+      "epoch": 14.812980358667804,
+      "grad_norm": 8.168084144592285,
+      "learning_rate": 0.00012950322335987866,
+      "loss": 5.213864135742187,
+      "step": 13020
+    },
+    {
+      "epoch": 14.824366638200967,
+      "grad_norm": 7.145936965942383,
+      "learning_rate": 0.0001294490492442711,
+      "loss": 5.022368621826172,
+      "step": 13030
+    },
+    {
+      "epoch": 14.83575291773413,
+      "grad_norm": 8.161090850830078,
+      "learning_rate": 0.00012939487512866353,
+      "loss": 5.109722518920899,
+      "step": 13040
+    },
+    {
+      "epoch": 14.847139197267293,
+      "grad_norm": 7.26900053024292,
+      "learning_rate": 0.00012934070101305599,
+      "loss": 5.046876907348633,
+      "step": 13050
+    },
+    {
+      "epoch": 14.858525476800455,
+      "grad_norm": 7.036923408508301,
+      "learning_rate": 0.0001292865268974484,
+      "loss": 4.823116683959961,
+      "step": 13060
+    },
+    {
+      "epoch": 14.869911756333618,
+      "grad_norm": 7.964908599853516,
+      "learning_rate": 0.00012923235278184086,
+      "loss": 5.159635925292969,
+      "step": 13070
+    },
+    {
+      "epoch": 14.88129803586678,
+      "grad_norm": 7.919378757476807,
+      "learning_rate": 0.00012917817866623329,
+      "loss": 4.9494384765625,
+      "step": 13080
+    },
+    {
+      "epoch": 14.892684315399944,
+      "grad_norm": 7.289670944213867,
+      "learning_rate": 0.0001291240045506257,
+      "loss": 5.084150695800782,
+      "step": 13090
+    },
+    {
+      "epoch": 14.904070594933106,
+      "grad_norm": 7.601792335510254,
+      "learning_rate": 0.00012906983043501813,
+      "loss": 4.942096328735351,
+      "step": 13100
+    },
+    {
+      "epoch": 14.915456874466269,
+      "grad_norm": 8.224839210510254,
+      "learning_rate": 0.0001290156563194106,
+      "loss": 4.804985046386719,
+      "step": 13110
+    },
+    {
+      "epoch": 14.92684315399943,
+      "grad_norm": 7.499617099761963,
+      "learning_rate": 0.000128961482203803,
+      "loss": 5.03905029296875,
+      "step": 13120
+    },
+    {
+      "epoch": 14.938229433532593,
+      "grad_norm": 8.021928787231445,
+      "learning_rate": 0.00012890730808819546,
+      "loss": 5.192089080810547,
+      "step": 13130
+    },
+    {
+      "epoch": 14.949615713065755,
+      "grad_norm": 7.354033470153809,
+      "learning_rate": 0.00012885313397258791,
+      "loss": 5.053718948364258,
+      "step": 13140
+    },
+    {
+      "epoch": 14.961001992598918,
+      "grad_norm": 7.821739673614502,
+      "learning_rate": 0.00012879895985698034,
+      "loss": 5.054132080078125,
+      "step": 13150
+    },
+    {
+      "epoch": 14.972388272132081,
+      "grad_norm": 7.435361862182617,
+      "learning_rate": 0.0001287447857413728,
+      "loss": 5.183145904541016,
+      "step": 13160
+    },
+    {
+      "epoch": 14.983774551665244,
+      "grad_norm": 7.934272289276123,
+      "learning_rate": 0.00012869061162576521,
+      "loss": 4.926422882080078,
+      "step": 13170
+    },
+    {
+      "epoch": 14.995160831198406,
+      "grad_norm": 8.532721519470215,
+      "learning_rate": 0.00012863643751015767,
+      "loss": 5.203876113891601,
+      "step": 13180
+    },
+    {
+      "epoch": 15.005693139766581,
+      "grad_norm": 7.0397629737854,
+      "learning_rate": 0.0001285822633945501,
+      "loss": 4.161993408203125,
+      "step": 13190
+    },
+    {
+      "epoch": 15.017079419299744,
+      "grad_norm": 8.11047649383545,
+      "learning_rate": 0.00012852808927894254,
+      "loss": 4.52406120300293,
+      "step": 13200
+    },
+    {
+      "epoch": 15.028465698832907,
+      "grad_norm": 7.387153148651123,
+      "learning_rate": 0.00012847391516333497,
+      "loss": 4.291901397705078,
+      "step": 13210
+    },
+    {
+      "epoch": 15.03985197836607,
+      "grad_norm": 7.610694885253906,
+      "learning_rate": 0.0001284197410477274,
+      "loss": 4.618151473999023,
+      "step": 13220
+    },
+    {
+      "epoch": 15.051238257899232,
+      "grad_norm": 8.041857719421387,
+      "learning_rate": 0.00012836556693211984,
+      "loss": 4.415428543090821,
+      "step": 13230
+    },
+    {
+      "epoch": 15.062624537432393,
+      "grad_norm": 8.153491020202637,
+      "learning_rate": 0.00012831139281651227,
+      "loss": 4.578215026855469,
+      "step": 13240
+    },
+    {
+      "epoch": 15.074010816965556,
+      "grad_norm": 7.969481468200684,
+      "learning_rate": 0.00012825721870090472,
+      "loss": 4.715361785888672,
+      "step": 13250
+    },
+    {
+      "epoch": 15.085397096498719,
+      "grad_norm": 6.474404811859131,
+      "learning_rate": 0.00012820304458529714,
+      "loss": 4.324704360961914,
+      "step": 13260
+    },
+    {
+      "epoch": 15.096783376031881,
+      "grad_norm": 7.767828464508057,
+      "learning_rate": 0.0001281488704696896,
+      "loss": 5.102642440795899,
+      "step": 13270
+    },
+    {
+      "epoch": 15.108169655565044,
+      "grad_norm": 6.66921854019165,
+      "learning_rate": 0.00012809469635408202,
+      "loss": 4.494071960449219,
+      "step": 13280
+    },
+    {
+      "epoch": 15.119555935098207,
+      "grad_norm": 7.946902275085449,
+      "learning_rate": 0.00012804052223847447,
+      "loss": 4.48153076171875,
+      "step": 13290
+    },
+    {
+      "epoch": 15.13094221463137,
+      "grad_norm": 7.2356486320495605,
+      "learning_rate": 0.0001279863481228669,
+      "loss": 4.334786224365234,
+      "step": 13300
+    },
+    {
+      "epoch": 15.142328494164532,
+      "grad_norm": 8.00400161743164,
+      "learning_rate": 0.00012793217400725935,
+      "loss": 4.645303344726562,
+      "step": 13310
+    },
+    {
+      "epoch": 15.153714773697695,
+      "grad_norm": 7.499370098114014,
+      "learning_rate": 0.00012787799989165177,
+      "loss": 4.6021240234375,
+      "step": 13320
+    },
+    {
+      "epoch": 15.165101053230856,
+      "grad_norm": 7.6751580238342285,
+      "learning_rate": 0.00012782382577604422,
+      "loss": 4.559888076782227,
+      "step": 13330
+    },
+    {
+      "epoch": 15.176487332764019,
+      "grad_norm": 7.785442352294922,
+      "learning_rate": 0.00012776965166043665,
+      "loss": 4.634941864013672,
+      "step": 13340
+    },
+    {
+      "epoch": 15.187873612297182,
+      "grad_norm": 7.800275802612305,
+      "learning_rate": 0.0001277154775448291,
+      "loss": 4.987696838378906,
+      "step": 13350
+    },
+    {
+      "epoch": 15.199259891830344,
+      "grad_norm": 7.645476341247559,
+      "learning_rate": 0.00012766130342922152,
+      "loss": 4.864948272705078,
+      "step": 13360
+    },
+    {
+      "epoch": 15.210646171363507,
+      "grad_norm": 7.993826866149902,
+      "learning_rate": 0.00012760712931361395,
+      "loss": 4.6918384552001955,
+      "step": 13370
+    },
+    {
+      "epoch": 15.22203245089667,
+      "grad_norm": 7.791032791137695,
+      "learning_rate": 0.0001275529551980064,
+      "loss": 4.629275131225586,
+      "step": 13380
+    },
+    {
+      "epoch": 15.233418730429833,
+      "grad_norm": 7.296651363372803,
+      "learning_rate": 0.00012749878108239882,
+      "loss": 4.688332748413086,
+      "step": 13390
+    },
+    {
+      "epoch": 15.244805009962995,
+      "grad_norm": 6.546830654144287,
+      "learning_rate": 0.00012744460696679128,
+      "loss": 4.458778381347656,
+      "step": 13400
+    },
+    {
+      "epoch": 15.256191289496158,
+      "grad_norm": 6.901986598968506,
+      "learning_rate": 0.0001273904328511837,
+      "loss": 4.439383316040039,
+      "step": 13410
+    },
+    {
+      "epoch": 15.267577569029319,
+      "grad_norm": 7.9615559577941895,
+      "learning_rate": 0.00012733625873557615,
+      "loss": 4.456288528442383,
+      "step": 13420
+    },
+    {
+      "epoch": 15.278963848562482,
+      "grad_norm": 7.0012593269348145,
+      "learning_rate": 0.00012728208461996858,
+      "loss": 4.7562213897705075,
+      "step": 13430
+    },
+    {
+      "epoch": 15.290350128095644,
+      "grad_norm": 8.112042427062988,
+      "learning_rate": 0.00012722791050436103,
+      "loss": 4.588046646118164,
+      "step": 13440
+    },
+    {
+      "epoch": 15.301736407628807,
+      "grad_norm": 8.545401573181152,
+      "learning_rate": 0.00012717373638875345,
+      "loss": 4.876891708374023,
+      "step": 13450
+    },
+    {
+      "epoch": 15.31312268716197,
+      "grad_norm": 7.401144981384277,
+      "learning_rate": 0.0001271195622731459,
+      "loss": 4.7802989959716795,
+      "step": 13460
+    },
+    {
+      "epoch": 15.324508966695133,
+      "grad_norm": 9.017389297485352,
+      "learning_rate": 0.00012706538815753836,
+      "loss": 4.951491165161133,
+      "step": 13470
+    },
+    {
+      "epoch": 15.335895246228295,
+      "grad_norm": 7.437997817993164,
+      "learning_rate": 0.00012701121404193078,
+      "loss": 4.486794662475586,
+      "step": 13480
+    },
+    {
+      "epoch": 15.347281525761458,
+      "grad_norm": 7.787178039550781,
+      "learning_rate": 0.00012695703992632323,
+      "loss": 4.632095718383789,
+      "step": 13490
+    },
+    {
+      "epoch": 15.35866780529462,
+      "grad_norm": 7.672351360321045,
+      "learning_rate": 0.00012690286581071566,
+      "loss": 4.595958709716797,
+      "step": 13500
+    },
+    {
+      "epoch": 15.370054084827782,
+      "grad_norm": 9.20431900024414,
+      "learning_rate": 0.00012684869169510808,
+      "loss": 4.731208038330078,
+      "step": 13510
+    },
+    {
+      "epoch": 15.381440364360945,
+      "grad_norm": 7.591139793395996,
+      "learning_rate": 0.0001267945175795005,
+      "loss": 4.76176643371582,
+      "step": 13520
+    },
+    {
+      "epoch": 15.392826643894107,
+      "grad_norm": 7.419577121734619,
+      "learning_rate": 0.00012674034346389296,
+      "loss": 4.38856201171875,
+      "step": 13530
+    },
+    {
+      "epoch": 15.40421292342727,
+      "grad_norm": 7.263993740081787,
+      "learning_rate": 0.00012668616934828538,
+      "loss": 4.695225524902344,
+      "step": 13540
+    },
+    {
+      "epoch": 15.415599202960433,
+      "grad_norm": 7.4275126457214355,
+      "learning_rate": 0.00012663199523267783,
+      "loss": 4.705525970458984,
+      "step": 13550
+    },
+    {
+      "epoch": 15.426985482493595,
+      "grad_norm": 8.560776710510254,
+      "learning_rate": 0.00012657782111707026,
+      "loss": 4.849109268188476,
+      "step": 13560
+    },
+    {
+      "epoch": 15.438371762026758,
+      "grad_norm": 8.176137924194336,
+      "learning_rate": 0.0001265236470014627,
+      "loss": 4.634634780883789,
+      "step": 13570
+    },
+    {
+      "epoch": 15.449758041559921,
+      "grad_norm": 7.90976095199585,
+      "learning_rate": 0.00012646947288585513,
+      "loss": 4.362587356567383,
+      "step": 13580
+    },
+    {
+      "epoch": 15.461144321093084,
+      "grad_norm": 8.50546646118164,
+      "learning_rate": 0.00012641529877024758,
+      "loss": 4.5653434753417965,
+      "step": 13590
+    },
+    {
+      "epoch": 15.472530600626245,
+      "grad_norm": 7.390838146209717,
+      "learning_rate": 0.00012636112465464004,
+      "loss": 4.620360565185547,
+      "step": 13600
+    },
+    {
+      "epoch": 15.483916880159407,
+      "grad_norm": 7.908941745758057,
+      "learning_rate": 0.00012630695053903246,
+      "loss": 4.444548416137695,
+      "step": 13610
+    },
+    {
+      "epoch": 15.49530315969257,
+      "grad_norm": 7.579942226409912,
+      "learning_rate": 0.0001262527764234249,
+      "loss": 4.685836410522461,
+      "step": 13620
+    },
+    {
+      "epoch": 15.506689439225733,
+      "grad_norm": 7.692448616027832,
+      "learning_rate": 0.00012619860230781734,
+      "loss": 4.817267227172851,
+      "step": 13630
+    },
+    {
+      "epoch": 15.518075718758896,
+      "grad_norm": 7.596404552459717,
+      "learning_rate": 0.00012614442819220976,
+      "loss": 4.794780349731445,
+      "step": 13640
+    },
+    {
+      "epoch": 15.529461998292058,
+      "grad_norm": 8.101449966430664,
+      "learning_rate": 0.00012609025407660219,
+      "loss": 4.803411483764648,
+      "step": 13650
+    },
+    {
+      "epoch": 15.540848277825221,
+      "grad_norm": 7.359433650970459,
+      "learning_rate": 0.00012603607996099464,
+      "loss": 4.751076889038086,
+      "step": 13660
+    },
+    {
+      "epoch": 15.552234557358384,
+      "grad_norm": 6.875908374786377,
+      "learning_rate": 0.00012598190584538706,
+      "loss": 4.687049484252929,
+      "step": 13670
+    },
+    {
+      "epoch": 15.563620836891547,
+      "grad_norm": 7.818774700164795,
+      "learning_rate": 0.0001259277317297795,
+      "loss": 4.953191757202148,
+      "step": 13680
+    },
+    {
+      "epoch": 15.575007116424707,
+      "grad_norm": 7.3973917961120605,
+      "learning_rate": 0.00012587355761417194,
+      "loss": 4.424762725830078,
+      "step": 13690
+    },
+    {
+      "epoch": 15.58639339595787,
+      "grad_norm": 8.235298156738281,
+      "learning_rate": 0.0001258193834985644,
+      "loss": 5.169184875488281,
+      "step": 13700
+    },
+    {
+      "epoch": 15.597779675491033,
+      "grad_norm": 7.2172532081604,
+      "learning_rate": 0.00012576520938295684,
+      "loss": 4.731855392456055,
+      "step": 13710
+    },
+    {
+      "epoch": 15.609165955024196,
+      "grad_norm": 7.717671871185303,
+      "learning_rate": 0.00012571103526734927,
+      "loss": 4.4836174011230465,
+      "step": 13720
+    },
+    {
+      "epoch": 15.620552234557358,
+      "grad_norm": 8.239593505859375,
+      "learning_rate": 0.00012565686115174172,
+      "loss": 4.932541275024414,
+      "step": 13730
+    },
+    {
+      "epoch": 15.631938514090521,
+      "grad_norm": 7.339993476867676,
+      "learning_rate": 0.00012560268703613414,
+      "loss": 4.306424331665039,
+      "step": 13740
+    },
+    {
+      "epoch": 15.643324793623684,
+      "grad_norm": 7.40126371383667,
+      "learning_rate": 0.0001255485129205266,
+      "loss": 4.77293815612793,
+      "step": 13750
+    },
+    {
+      "epoch": 15.654711073156847,
+      "grad_norm": 6.878884792327881,
+      "learning_rate": 0.00012549433880491902,
+      "loss": 4.673749923706055,
+      "step": 13760
+    },
+    {
+      "epoch": 15.66609735269001,
+      "grad_norm": 8.765450477600098,
+      "learning_rate": 0.00012544016468931147,
+      "loss": 4.610214614868164,
+      "step": 13770
+    },
+    {
+      "epoch": 15.67748363222317,
+      "grad_norm": 8.095568656921387,
+      "learning_rate": 0.0001253859905737039,
+      "loss": 4.4134258270263675,
+      "step": 13780
+    },
+    {
+      "epoch": 15.688869911756333,
+      "grad_norm": 8.046052932739258,
+      "learning_rate": 0.00012533181645809632,
+      "loss": 4.598631286621094,
+      "step": 13790
+    },
+    {
+      "epoch": 15.700256191289496,
+      "grad_norm": 7.999307155609131,
+      "learning_rate": 0.00012527764234248874,
+      "loss": 4.934955596923828,
+      "step": 13800
+    },
+    {
+      "epoch": 15.711642470822659,
+      "grad_norm": 6.909728050231934,
+      "learning_rate": 0.0001252234682268812,
+      "loss": 4.704043960571289,
+      "step": 13810
+    },
+    {
+      "epoch": 15.723028750355821,
+      "grad_norm": 7.83714485168457,
+      "learning_rate": 0.00012516929411127365,
+      "loss": 4.726215362548828,
+      "step": 13820
+    },
+    {
+      "epoch": 15.734415029888984,
+      "grad_norm": 7.958460330963135,
+      "learning_rate": 0.00012511511999566607,
+      "loss": 4.580567932128906,
+      "step": 13830
+    },
+    {
+      "epoch": 15.745801309422147,
+      "grad_norm": 8.28144359588623,
+      "learning_rate": 0.00012506094588005852,
+      "loss": 4.877122116088867,
+      "step": 13840
+    },
+    {
+      "epoch": 15.75718758895531,
+      "grad_norm": 7.191257953643799,
+      "learning_rate": 0.00012500677176445095,
+      "loss": 4.728401184082031,
+      "step": 13850
+    },
+    {
+      "epoch": 15.768573868488472,
+      "grad_norm": 8.373869895935059,
+      "learning_rate": 0.0001249525976488434,
+      "loss": 4.692030715942383,
+      "step": 13860
+    },
+    {
+      "epoch": 15.779960148021633,
+      "grad_norm": 7.541042804718018,
+      "learning_rate": 0.00012489842353323582,
+      "loss": 4.673214721679687,
+      "step": 13870
+    },
+    {
+      "epoch": 15.791346427554796,
+      "grad_norm": 7.176820278167725,
+      "learning_rate": 0.00012484424941762827,
+      "loss": 4.742586135864258,
+      "step": 13880
+    },
+    {
+      "epoch": 15.802732707087959,
+      "grad_norm": 7.285589694976807,
+      "learning_rate": 0.0001247900753020207,
+      "loss": 4.660622406005859,
+      "step": 13890
+    },
+    {
+      "epoch": 15.814118986621121,
+      "grad_norm": 7.229033946990967,
+      "learning_rate": 0.0001247413185979739,
+      "loss": 4.864823532104492,
+      "step": 13900
+    },
+    {
+      "epoch": 15.825505266154284,
+      "grad_norm": 8.045304298400879,
+      "learning_rate": 0.00012468714448236635,
+      "loss": 4.963446807861328,
+      "step": 13910
+    },
+    {
+      "epoch": 15.836891545687447,
+      "grad_norm": 8.248867988586426,
+      "learning_rate": 0.00012463297036675877,
+      "loss": 4.45933837890625,
+      "step": 13920
+    },
+    {
+      "epoch": 15.84827782522061,
+      "grad_norm": 7.996612071990967,
+      "learning_rate": 0.00012457879625115122,
+      "loss": 4.713712310791015,
+      "step": 13930
+    },
+    {
+      "epoch": 15.859664104753772,
+      "grad_norm": 7.0257568359375,
+      "learning_rate": 0.00012452462213554365,
+      "loss": 4.95684814453125,
+      "step": 13940
+    },
+    {
+      "epoch": 15.871050384286935,
+      "grad_norm": 7.490429401397705,
+      "learning_rate": 0.00012447044801993607,
+      "loss": 5.097381210327148,
+      "step": 13950
+    },
+    {
+      "epoch": 15.882436663820096,
+      "grad_norm": 8.141972541809082,
+      "learning_rate": 0.00012441627390432852,
+      "loss": 4.907619476318359,
+      "step": 13960
+    },
+    {
+      "epoch": 15.893822943353259,
+      "grad_norm": 7.669741153717041,
+      "learning_rate": 0.00012436209978872095,
+      "loss": 4.52638053894043,
+      "step": 13970
+    },
+    {
+      "epoch": 15.905209222886421,
+      "grad_norm": 7.623498439788818,
+      "learning_rate": 0.0001243079256731134,
+      "loss": 4.6605175018310545,
+      "step": 13980
+    },
+    {
+      "epoch": 15.916595502419584,
+      "grad_norm": 7.301693916320801,
+      "learning_rate": 0.00012425375155750582,
+      "loss": 4.762905883789062,
+      "step": 13990
+    },
+    {
+      "epoch": 15.927981781952747,
+      "grad_norm": 7.472468376159668,
+      "learning_rate": 0.00012419957744189827,
+      "loss": 4.756846237182617,
+      "step": 14000
+    },
+    {
+      "epoch": 15.93936806148591,
+      "grad_norm": 7.766934871673584,
+      "learning_rate": 0.0001241454033262907,
+      "loss": 4.748290252685547,
+      "step": 14010
+    },
+    {
+      "epoch": 15.950754341019072,
+      "grad_norm": 6.7247796058654785,
+      "learning_rate": 0.00012409122921068315,
+      "loss": 4.70723876953125,
+      "step": 14020
+    },
+    {
+      "epoch": 15.962140620552235,
+      "grad_norm": 8.007941246032715,
+      "learning_rate": 0.00012403705509507557,
+      "loss": 4.603389739990234,
+      "step": 14030
+    },
+    {
+      "epoch": 15.973526900085398,
+      "grad_norm": 7.72274923324585,
+      "learning_rate": 0.00012398288097946803,
+      "loss": 4.74395751953125,
+      "step": 14040
+    },
+    {
+      "epoch": 15.984913179618559,
+      "grad_norm": 7.55423641204834,
+      "learning_rate": 0.00012392870686386045,
+      "loss": 4.812894439697265,
+      "step": 14050
+    },
+    {
+      "epoch": 15.996299459151722,
+      "grad_norm": 7.276206016540527,
+      "learning_rate": 0.0001238745327482529,
+      "loss": 4.808458709716797,
+      "step": 14060
+    },
+    {
+      "epoch": 16.0068317677199,
+      "grad_norm": 7.715972900390625,
+      "learning_rate": 0.00012382035863264533,
+      "loss": 4.121777725219727,
+      "step": 14070
+    },
+    {
+      "epoch": 16.01821804725306,
+      "grad_norm": 7.861795425415039,
+      "learning_rate": 0.00012376618451703775,
+      "loss": 4.087036895751953,
+      "step": 14080
+    },
+    {
+      "epoch": 16.029604326786224,
+      "grad_norm": 7.377951622009277,
+      "learning_rate": 0.0001237120104014302,
+      "loss": 4.159128570556641,
+      "step": 14090
+    },
+    {
+      "epoch": 16.040990606319387,
+      "grad_norm": 6.9755730628967285,
+      "learning_rate": 0.00012365783628582263,
+      "loss": 4.118164443969727,
+      "step": 14100
+    },
+    {
+      "epoch": 16.05237688585255,
+      "grad_norm": 7.6896467208862305,
+      "learning_rate": 0.00012360366217021508,
+      "loss": 4.3465576171875,
+      "step": 14110
+    },
+    {
+      "epoch": 16.06376316538571,
+      "grad_norm": 7.475165367126465,
+      "learning_rate": 0.0001235494880546075,
+      "loss": 4.084910202026367,
+      "step": 14120
+    },
+    {
+      "epoch": 16.07514944491887,
+      "grad_norm": 7.571929931640625,
+      "learning_rate": 0.00012349531393899995,
+      "loss": 4.203293228149414,
+      "step": 14130
+    },
+    {
+      "epoch": 16.086535724452034,
+      "grad_norm": 7.941225051879883,
+      "learning_rate": 0.00012344113982339238,
+      "loss": 4.56109733581543,
+      "step": 14140
+    },
+    {
+      "epoch": 16.097922003985197,
+      "grad_norm": 7.033837795257568,
+      "learning_rate": 0.00012338696570778483,
+      "loss": 4.267443466186523,
+      "step": 14150
+    },
+    {
+      "epoch": 16.10930828351836,
+      "grad_norm": 7.488673210144043,
+      "learning_rate": 0.00012333279159217726,
+      "loss": 4.451668167114258,
+      "step": 14160
+    },
+    {
+      "epoch": 16.120694563051522,
+      "grad_norm": 7.623510837554932,
+      "learning_rate": 0.0001232786174765697,
+      "loss": 4.3589630126953125,
+      "step": 14170
+    },
+    {
+      "epoch": 16.132080842584685,
+      "grad_norm": 7.737654685974121,
+      "learning_rate": 0.00012322444336096213,
+      "loss": 4.383876037597656,
+      "step": 14180
+    },
+    {
+      "epoch": 16.143467122117848,
+      "grad_norm": 8.437525749206543,
+      "learning_rate": 0.00012317026924535458,
+      "loss": 4.350270843505859,
+      "step": 14190
+    },
+    {
+      "epoch": 16.15485340165101,
+      "grad_norm": 7.971991539001465,
+      "learning_rate": 0.00012311609512974703,
+      "loss": 4.457265472412109,
+      "step": 14200
+    },
+    {
+      "epoch": 16.166239681184173,
+      "grad_norm": 7.322634220123291,
+      "learning_rate": 0.00012306192101413946,
+      "loss": 4.124166488647461,
+      "step": 14210
+    },
+    {
+      "epoch": 16.177625960717336,
+      "grad_norm": 6.804745674133301,
+      "learning_rate": 0.00012300774689853188,
+      "loss": 4.338391876220703,
+      "step": 14220
+    },
+    {
+      "epoch": 16.1890122402505,
+      "grad_norm": 7.241665840148926,
+      "learning_rate": 0.0001229535727829243,
+      "loss": 3.9033172607421873,
+      "step": 14230
+    },
+    {
+      "epoch": 16.20039851978366,
+      "grad_norm": 7.835842609405518,
+      "learning_rate": 0.00012289939866731676,
+      "loss": 4.350611877441406,
+      "step": 14240
+    },
+    {
+      "epoch": 16.211784799316824,
+      "grad_norm": 7.062921047210693,
+      "learning_rate": 0.00012284522455170918,
+      "loss": 4.4512939453125,
+      "step": 14250
+    },
+    {
+      "epoch": 16.223171078849987,
+      "grad_norm": 7.936232566833496,
+      "learning_rate": 0.00012279105043610164,
+      "loss": 4.246225738525391,
+      "step": 14260
+    },
+    {
+      "epoch": 16.23455735838315,
+      "grad_norm": 8.720023155212402,
+      "learning_rate": 0.00012273687632049406,
+      "loss": 4.419322967529297,
+      "step": 14270
+    },
+    {
+      "epoch": 16.245943637916312,
+      "grad_norm": 7.396350383758545,
+      "learning_rate": 0.0001226827022048865,
+      "loss": 4.686096954345703,
+      "step": 14280
+    },
+    {
+      "epoch": 16.257329917449475,
+      "grad_norm": 7.3385515213012695,
+      "learning_rate": 0.00012262852808927894,
+      "loss": 4.519886016845703,
+      "step": 14290
+    },
+    {
+      "epoch": 16.268716196982638,
+      "grad_norm": 7.164921283721924,
+      "learning_rate": 0.0001225743539736714,
+      "loss": 4.170539474487304,
+      "step": 14300
+    },
+    {
+      "epoch": 16.280102476515797,
+      "grad_norm": 6.846629619598389,
+      "learning_rate": 0.00012252017985806384,
+      "loss": 4.29273796081543,
+      "step": 14310
+    },
+    {
+      "epoch": 16.29148875604896,
+      "grad_norm": 8.011646270751953,
+      "learning_rate": 0.00012246600574245626,
+      "loss": 4.424296951293945,
+      "step": 14320
+    },
+    {
+      "epoch": 16.302875035582122,
+      "grad_norm": 7.801056861877441,
+      "learning_rate": 0.00012241183162684872,
+      "loss": 4.3867851257324215,
+      "step": 14330
+    },
+    {
+      "epoch": 16.314261315115285,
+      "grad_norm": 7.563156604766846,
+      "learning_rate": 0.00012235765751124114,
+      "loss": 4.142354583740234,
+      "step": 14340
+    },
+    {
+      "epoch": 16.325647594648448,
+      "grad_norm": 7.874721527099609,
+      "learning_rate": 0.0001223034833956336,
+      "loss": 4.672331619262695,
+      "step": 14350
+    },
+    {
+      "epoch": 16.33703387418161,
+      "grad_norm": 8.173314094543457,
+      "learning_rate": 0.00012224930928002602,
+      "loss": 4.456505584716797,
+      "step": 14360
+    },
+    {
+      "epoch": 16.348420153714773,
+      "grad_norm": 7.8878092765808105,
+      "learning_rate": 0.00012219513516441844,
+      "loss": 4.445765686035156,
+      "step": 14370
+    },
+    {
+      "epoch": 16.359806433247936,
+      "grad_norm": 8.385254859924316,
+      "learning_rate": 0.00012214096104881086,
+      "loss": 4.569265747070313,
+      "step": 14380
+    },
+    {
+      "epoch": 16.3711927127811,
+      "grad_norm": 8.698989868164062,
+      "learning_rate": 0.00012208678693320332,
+      "loss": 4.381620788574219,
+      "step": 14390
+    },
+    {
+      "epoch": 16.38257899231426,
+      "grad_norm": 6.675107479095459,
+      "learning_rate": 0.00012203261281759575,
+      "loss": 4.386390686035156,
+      "step": 14400
+    },
+    {
+      "epoch": 16.393965271847424,
+      "grad_norm": 7.738358020782471,
+      "learning_rate": 0.00012197843870198819,
+      "loss": 3.9419265747070313,
+      "step": 14410
+    },
+    {
+      "epoch": 16.405351551380587,
+      "grad_norm": 7.995990753173828,
+      "learning_rate": 0.00012192426458638064,
+      "loss": 4.322775268554688,
+      "step": 14420
+    },
+    {
+      "epoch": 16.41673783091375,
+      "grad_norm": 7.4500813484191895,
+      "learning_rate": 0.00012187009047077307,
+      "loss": 4.260313415527344,
+      "step": 14430
+    },
+    {
+      "epoch": 16.428124110446912,
+      "grad_norm": 8.59277057647705,
+      "learning_rate": 0.00012181591635516552,
+      "loss": 4.660084533691406,
+      "step": 14440
+    },
+    {
+      "epoch": 16.439510389980075,
+      "grad_norm": 7.587380409240723,
+      "learning_rate": 0.00012176174223955794,
+      "loss": 4.50414924621582,
+      "step": 14450
+    },
+    {
+      "epoch": 16.450896669513238,
+      "grad_norm": 7.656953811645508,
+      "learning_rate": 0.00012170756812395038,
+      "loss": 4.5958251953125,
+      "step": 14460
+    },
+    {
+      "epoch": 16.4622829490464,
+      "grad_norm": 7.589241981506348,
+      "learning_rate": 0.00012165339400834281,
+      "loss": 4.327259063720703,
+      "step": 14470
+    },
+    {
+      "epoch": 16.47366922857956,
+      "grad_norm": 7.885225296020508,
+      "learning_rate": 0.00012159921989273526,
+      "loss": 4.439897537231445,
+      "step": 14480
+    },
+    {
+      "epoch": 16.485055508112723,
+      "grad_norm": 7.397141456604004,
+      "learning_rate": 0.00012154504577712768,
+      "loss": 4.272943115234375,
+      "step": 14490
+    },
+    {
+      "epoch": 16.496441787645885,
+      "grad_norm": 6.983608245849609,
+      "learning_rate": 0.00012149087166152013,
+      "loss": 4.209513092041016,
+      "step": 14500
+    },
+    {
+      "epoch": 16.507828067179048,
+      "grad_norm": 8.37636947631836,
+      "learning_rate": 0.00012143669754591256,
+      "loss": 4.410722732543945,
+      "step": 14510
+    },
+    {
+      "epoch": 16.51921434671221,
+      "grad_norm": 8.623056411743164,
+      "learning_rate": 0.00012138252343030501,
+      "loss": 4.736907577514648,
+      "step": 14520
+    },
+    {
+      "epoch": 16.530600626245374,
+      "grad_norm": 7.334944725036621,
+      "learning_rate": 0.00012132834931469745,
+      "loss": 4.085528945922851,
+      "step": 14530
+    },
+    {
+      "epoch": 16.541986905778536,
+      "grad_norm": 7.641763210296631,
+      "learning_rate": 0.00012127417519908987,
+      "loss": 4.196076202392578,
+      "step": 14540
+    },
+    {
+      "epoch": 16.5533731853117,
+      "grad_norm": 7.118811130523682,
+      "learning_rate": 0.00012122000108348232,
+      "loss": 4.219401931762695,
+      "step": 14550
+    },
+    {
+      "epoch": 16.56475946484486,
+      "grad_norm": 7.666287422180176,
+      "learning_rate": 0.00012116582696787475,
+      "loss": 4.500454330444336,
+      "step": 14560
+    },
+    {
+      "epoch": 16.576145744378024,
+      "grad_norm": 8.155136108398438,
+      "learning_rate": 0.0001211116528522672,
+      "loss": 4.450511932373047,
+      "step": 14570
+    },
+    {
+      "epoch": 16.587532023911187,
+      "grad_norm": 7.149616241455078,
+      "learning_rate": 0.00012105747873665963,
+      "loss": 4.229332733154297,
+      "step": 14580
+    },
+    {
+      "epoch": 16.59891830344435,
+      "grad_norm": 8.195096969604492,
+      "learning_rate": 0.00012100330462105208,
+      "loss": 4.491680908203125,
+      "step": 14590
+    },
+    {
+      "epoch": 16.610304582977513,
+      "grad_norm": 6.949581146240234,
+      "learning_rate": 0.0001209491305054445,
+      "loss": 4.443110656738281,
+      "step": 14600
+    },
+    {
+      "epoch": 16.621690862510675,
+      "grad_norm": 7.297284126281738,
+      "learning_rate": 0.00012089495638983694,
+      "loss": 4.731317520141602,
+      "step": 14610
+    },
+    {
+      "epoch": 16.633077142043838,
+      "grad_norm": 6.88108491897583,
+      "learning_rate": 0.00012084078227422936,
+      "loss": 4.465324783325196,
+      "step": 14620
+    },
+    {
+      "epoch": 16.644463421577,
+      "grad_norm": 8.48362922668457,
+      "learning_rate": 0.00012078660815862182,
+      "loss": 4.320807266235351,
+      "step": 14630
+    },
+    {
+      "epoch": 16.655849701110164,
+      "grad_norm": 8.934884071350098,
+      "learning_rate": 0.00012073243404301427,
+      "loss": 4.569874954223633,
+      "step": 14640
+    },
+    {
+      "epoch": 16.667235980643326,
+      "grad_norm": 7.942513942718506,
+      "learning_rate": 0.00012067825992740669,
+      "loss": 4.774484634399414,
+      "step": 14650
+    },
+    {
+      "epoch": 16.67862226017649,
+      "grad_norm": 7.406705379486084,
+      "learning_rate": 0.00012062408581179914,
+      "loss": 4.290068817138672,
+      "step": 14660
+    },
+    {
+      "epoch": 16.69000853970965,
+      "grad_norm": 7.856991291046143,
+      "learning_rate": 0.00012056991169619157,
+      "loss": 4.307246398925781,
+      "step": 14670
+    },
+    {
+      "epoch": 16.70139481924281,
+      "grad_norm": 8.263503074645996,
+      "learning_rate": 0.000120515737580584,
+      "loss": 4.2724559783935545,
+      "step": 14680
+    },
+    {
+      "epoch": 16.712781098775974,
+      "grad_norm": 8.12948226928711,
+      "learning_rate": 0.00012046156346497643,
+      "loss": 4.421898651123047,
+      "step": 14690
+    },
+    {
+      "epoch": 16.724167378309136,
+      "grad_norm": 8.351273536682129,
+      "learning_rate": 0.00012040738934936888,
+      "loss": 4.521099090576172,
+      "step": 14700
+    },
+    {
+      "epoch": 16.7355536578423,
+      "grad_norm": 7.658185958862305,
+      "learning_rate": 0.0001203532152337613,
+      "loss": 4.358420562744141,
+      "step": 14710
+    },
+    {
+      "epoch": 16.746939937375462,
+      "grad_norm": 7.626881122589111,
+      "learning_rate": 0.00012029904111815376,
+      "loss": 4.407142257690429,
+      "step": 14720
+    },
+    {
+      "epoch": 16.758326216908625,
+      "grad_norm": 7.826634407043457,
+      "learning_rate": 0.00012024486700254618,
+      "loss": 4.7284095764160154,
+      "step": 14730
+    },
+    {
+      "epoch": 16.769712496441787,
+      "grad_norm": 7.835376262664795,
+      "learning_rate": 0.00012019069288693863,
+      "loss": 4.649514007568359,
+      "step": 14740
+    },
+    {
+      "epoch": 16.78109877597495,
+      "grad_norm": 7.345791816711426,
+      "learning_rate": 0.00012013651877133107,
+      "loss": 4.5487213134765625,
+      "step": 14750
+    },
+    {
+      "epoch": 16.792485055508113,
+      "grad_norm": 7.653363227844238,
+      "learning_rate": 0.0001200823446557235,
+      "loss": 4.464413833618164,
+      "step": 14760
+    },
+    {
+      "epoch": 16.803871335041276,
+      "grad_norm": 7.4951043128967285,
+      "learning_rate": 0.00012002817054011595,
+      "loss": 4.552109146118164,
+      "step": 14770
+    },
+    {
+      "epoch": 16.81525761457444,
+      "grad_norm": 6.739938259124756,
+      "learning_rate": 0.00011997399642450837,
+      "loss": 4.435204315185547,
+      "step": 14780
+    },
+    {
+      "epoch": 16.8266438941076,
+      "grad_norm": 7.946920871734619,
+      "learning_rate": 0.00011991982230890082,
+      "loss": 4.394256591796875,
+      "step": 14790
+    },
+    {
+      "epoch": 16.838030173640764,
+      "grad_norm": 7.2037224769592285,
+      "learning_rate": 0.00011986564819329325,
+      "loss": 4.656013107299804,
+      "step": 14800
+    },
+    {
+      "epoch": 16.849416453173927,
+      "grad_norm": 7.150942325592041,
+      "learning_rate": 0.00011981147407768569,
+      "loss": 4.620296478271484,
+      "step": 14810
+    },
+    {
+      "epoch": 16.86080273270709,
+      "grad_norm": 8.865056037902832,
+      "learning_rate": 0.00011975729996207812,
+      "loss": 4.58060302734375,
+      "step": 14820
+    },
+    {
+      "epoch": 16.872189012240252,
+      "grad_norm": 7.587070941925049,
+      "learning_rate": 0.00011970312584647056,
+      "loss": 4.879767990112304,
+      "step": 14830
+    },
+    {
+      "epoch": 16.88357529177341,
+      "grad_norm": 8.456056594848633,
+      "learning_rate": 0.00011964895173086299,
+      "loss": 4.5918418884277346,
+      "step": 14840
+    },
+    {
+      "epoch": 16.894961571306574,
+      "grad_norm": 7.7380900382995605,
+      "learning_rate": 0.00011959477761525544,
+      "loss": 4.7449501037597654,
+      "step": 14850
+    },
+    {
+      "epoch": 16.906347850839737,
+      "grad_norm": 7.467341423034668,
+      "learning_rate": 0.00011954060349964786,
+      "loss": 4.479496765136719,
+      "step": 14860
+    },
+    {
+      "epoch": 16.9177341303729,
+      "grad_norm": 8.6979398727417,
+      "learning_rate": 0.00011948642938404031,
+      "loss": 4.7719371795654295,
+      "step": 14870
+    },
+    {
+      "epoch": 16.929120409906062,
+      "grad_norm": 7.696811676025391,
+      "learning_rate": 0.00011943225526843275,
+      "loss": 4.6034080505371096,
+      "step": 14880
+    },
+    {
+      "epoch": 16.940506689439225,
+      "grad_norm": 6.743338584899902,
+      "learning_rate": 0.00011937808115282518,
+      "loss": 4.397494888305664,
+      "step": 14890
+    },
+    {
+      "epoch": 16.951892968972388,
+      "grad_norm": 6.509140491485596,
+      "learning_rate": 0.00011932390703721763,
+      "loss": 4.695284271240235,
+      "step": 14900
+    },
+    {
+      "epoch": 16.96327924850555,
+      "grad_norm": 7.908071994781494,
+      "learning_rate": 0.00011926973292161005,
+      "loss": 4.142197799682617,
+      "step": 14910
+    },
+    {
+      "epoch": 16.974665528038713,
+      "grad_norm": 8.120683670043945,
+      "learning_rate": 0.0001192155588060025,
+      "loss": 4.365460968017578,
+      "step": 14920
+    },
+    {
+      "epoch": 16.986051807571876,
+      "grad_norm": 7.481232643127441,
+      "learning_rate": 0.00011916138469039493,
+      "loss": 4.682232284545899,
+      "step": 14930
+    },
+    {
+      "epoch": 16.99743808710504,
+      "grad_norm": 7.541874408721924,
+      "learning_rate": 0.00011910721057478738,
+      "loss": 4.875133895874024,
+      "step": 14940
+    },
+    {
+      "epoch": 17.007970395673215,
+      "grad_norm": 6.461215019226074,
+      "learning_rate": 0.0001190530364591798,
+      "loss": 3.5713672637939453,
+      "step": 14950
+    },
+    {
+      "epoch": 17.019356675206378,
+      "grad_norm": 8.02785873413086,
+      "learning_rate": 0.00011899886234357224,
+      "loss": 3.8977893829345702,
+      "step": 14960
+    },
+    {
+      "epoch": 17.030742954739537,
+      "grad_norm": 7.833250522613525,
+      "learning_rate": 0.00011894468822796467,
+      "loss": 4.051911926269531,
+      "step": 14970
+    },
+    {
+      "epoch": 17.0421292342727,
+      "grad_norm": 7.147907257080078,
+      "learning_rate": 0.00011889051411235712,
+      "loss": 4.067853546142578,
+      "step": 14980
+    },
+    {
+      "epoch": 17.053515513805863,
+      "grad_norm": 8.514575958251953,
+      "learning_rate": 0.00011883633999674957,
+      "loss": 4.16202392578125,
+      "step": 14990
+    },
+    {
+      "epoch": 17.064901793339025,
+      "grad_norm": 6.706639766693115,
+      "learning_rate": 0.000118782165881142,
+      "loss": 3.879001998901367,
+      "step": 15000
+    },
+    {
+      "epoch": 17.076288072872188,
+      "grad_norm": 7.3274827003479,
+      "learning_rate": 0.00011872799176553445,
+      "loss": 3.8930648803710937,
+      "step": 15010
+    },
+    {
+      "epoch": 17.08767435240535,
+      "grad_norm": 8.865731239318848,
+      "learning_rate": 0.00011867381764992687,
+      "loss": 4.000261688232422,
+      "step": 15020
+    },
+    {
+      "epoch": 17.099060631938514,
+      "grad_norm": 8.527647018432617,
+      "learning_rate": 0.00011861964353431931,
+      "loss": 4.380818557739258,
+      "step": 15030
+    },
+    {
+      "epoch": 17.110446911471676,
+      "grad_norm": 6.795366287231445,
+      "learning_rate": 0.00011856546941871173,
+      "loss": 4.000965118408203,
+      "step": 15040
+    },
+    {
+      "epoch": 17.12183319100484,
+      "grad_norm": 8.024996757507324,
+      "learning_rate": 0.00011851129530310419,
+      "loss": 3.8776901245117186,
+      "step": 15050
+    },
+    {
+      "epoch": 17.133219470538002,
+      "grad_norm": 7.76874303817749,
+      "learning_rate": 0.00011845712118749661,
+      "loss": 3.907266616821289,
+      "step": 15060
+    },
+    {
+      "epoch": 17.144605750071165,
+      "grad_norm": 7.170513153076172,
+      "learning_rate": 0.00011840294707188906,
+      "loss": 4.122880172729492,
+      "step": 15070
+    },
+    {
+      "epoch": 17.155992029604327,
+      "grad_norm": 7.131034851074219,
+      "learning_rate": 0.00011834877295628149,
+      "loss": 4.134088134765625,
+      "step": 15080
+    },
+    {
+      "epoch": 17.16737830913749,
+      "grad_norm": 9.287558555603027,
+      "learning_rate": 0.00011829459884067394,
+      "loss": 4.137316131591797,
+      "step": 15090
+    },
+    {
+      "epoch": 17.178764588670653,
+      "grad_norm": 6.616999626159668,
+      "learning_rate": 0.00011824042472506638,
+      "loss": 4.016845321655273,
+      "step": 15100
+    },
+    {
+      "epoch": 17.190150868203816,
+      "grad_norm": 7.829199314117432,
+      "learning_rate": 0.0001181862506094588,
+      "loss": 4.34521369934082,
+      "step": 15110
+    },
+    {
+      "epoch": 17.20153714773698,
+      "grad_norm": 8.117554664611816,
+      "learning_rate": 0.00011813207649385125,
+      "loss": 3.9818496704101562,
+      "step": 15120
+    },
+    {
+      "epoch": 17.21292342727014,
+      "grad_norm": 8.301651954650879,
+      "learning_rate": 0.00011807790237824368,
+      "loss": 4.036884307861328,
+      "step": 15130
+    },
+    {
+      "epoch": 17.224309706803304,
+      "grad_norm": 8.472175598144531,
+      "learning_rate": 0.00011802372826263613,
+      "loss": 4.04235725402832,
+      "step": 15140
+    },
+    {
+      "epoch": 17.235695986336463,
+      "grad_norm": 8.00710391998291,
+      "learning_rate": 0.00011796955414702855,
+      "loss": 4.428823089599609,
+      "step": 15150
+    },
+    {
+      "epoch": 17.247082265869626,
+      "grad_norm": 7.678086757659912,
+      "learning_rate": 0.00011791538003142099,
+      "loss": 4.123222351074219,
+      "step": 15160
+    },
+    {
+      "epoch": 17.25846854540279,
+      "grad_norm": 7.292410850524902,
+      "learning_rate": 0.00011786120591581343,
+      "loss": 4.134667205810547,
+      "step": 15170
+    },
+    {
+      "epoch": 17.26985482493595,
+      "grad_norm": 7.336828231811523,
+      "learning_rate": 0.00011780703180020587,
+      "loss": 4.142428970336914,
+      "step": 15180
+    },
+    {
+      "epoch": 17.281241104469114,
+      "grad_norm": 6.939968109130859,
+      "learning_rate": 0.00011775285768459829,
+      "loss": 4.088996505737304,
+      "step": 15190
+    },
+    {
+      "epoch": 17.292627384002277,
+      "grad_norm": 6.900147438049316,
+      "learning_rate": 0.00011769868356899074,
+      "loss": 4.092143630981445,
+      "step": 15200
+    },
+    {
+      "epoch": 17.30401366353544,
+      "grad_norm": 8.642024993896484,
+      "learning_rate": 0.0001176445094533832,
+      "loss": 4.337299728393555,
+      "step": 15210
+    },
+    {
+      "epoch": 17.315399943068602,
+      "grad_norm": 7.520287036895752,
+      "learning_rate": 0.00011759033533777562,
+      "loss": 3.809254837036133,
+      "step": 15220
+    },
+    {
+      "epoch": 17.326786222601765,
+      "grad_norm": 8.518157958984375,
+      "learning_rate": 0.00011753616122216806,
+      "loss": 4.095516967773437,
+      "step": 15230
+    },
+    {
+      "epoch": 17.338172502134928,
+      "grad_norm": 8.106553077697754,
+      "learning_rate": 0.00011748198710656048,
+      "loss": 4.244467163085938,
+      "step": 15240
+    },
+    {
+      "epoch": 17.34955878166809,
+      "grad_norm": 7.119726181030273,
+      "learning_rate": 0.00011742781299095293,
+      "loss": 4.04039192199707,
+      "step": 15250
+    },
+    {
+      "epoch": 17.360945061201253,
+      "grad_norm": 8.118093490600586,
+      "learning_rate": 0.00011737363887534536,
+      "loss": 4.112255859375,
+      "step": 15260
+    },
+    {
+      "epoch": 17.372331340734416,
+      "grad_norm": 7.7591729164123535,
+      "learning_rate": 0.00011731946475973781,
+      "loss": 3.9870590209960937,
+      "step": 15270
+    },
+    {
+      "epoch": 17.38371762026758,
+      "grad_norm": 6.926348686218262,
+      "learning_rate": 0.00011726529064413023,
+      "loss": 4.05236701965332,
+      "step": 15280
+    },
+    {
+      "epoch": 17.39510389980074,
+      "grad_norm": 7.686641216278076,
+      "learning_rate": 0.00011721111652852268,
+      "loss": 3.9766483306884766,
+      "step": 15290
+    },
+    {
+      "epoch": 17.406490179333904,
+      "grad_norm": 7.820587635040283,
+      "learning_rate": 0.00011715694241291511,
+      "loss": 4.117971801757813,
+      "step": 15300
+    },
+    {
+      "epoch": 17.417876458867067,
+      "grad_norm": 9.149397850036621,
+      "learning_rate": 0.00011710276829730755,
+      "loss": 4.0301368713378904,
+      "step": 15310
+    },
+    {
+      "epoch": 17.42926273840023,
+      "grad_norm": 7.755941390991211,
+      "learning_rate": 0.0001170485941817,
+      "loss": 4.266557312011718,
+      "step": 15320
+    },
+    {
+      "epoch": 17.44064901793339,
+      "grad_norm": 7.495479106903076,
+      "learning_rate": 0.00011699442006609242,
+      "loss": 4.452923202514649,
+      "step": 15330
+    },
+    {
+      "epoch": 17.45203529746655,
+      "grad_norm": 8.05996322631836,
+      "learning_rate": 0.00011694024595048487,
+      "loss": 4.292583847045899,
+      "step": 15340
+    },
+    {
+      "epoch": 17.463421576999714,
+      "grad_norm": 9.973164558410645,
+      "learning_rate": 0.0001168860718348773,
+      "loss": 4.290217590332031,
+      "step": 15350
+    },
+    {
+      "epoch": 17.474807856532877,
+      "grad_norm": 8.033830642700195,
+      "learning_rate": 0.00011683189771926975,
+      "loss": 4.155846405029297,
+      "step": 15360
+    },
+    {
+      "epoch": 17.48619413606604,
+      "grad_norm": 7.986049652099609,
+      "learning_rate": 0.00011677772360366217,
+      "loss": 4.047249603271484,
+      "step": 15370
+    },
+    {
+      "epoch": 17.497580415599202,
+      "grad_norm": 7.257813930511475,
+      "learning_rate": 0.00011672354948805461,
+      "loss": 4.1319633483886715,
+      "step": 15380
+    },
+    {
+      "epoch": 17.508966695132365,
+      "grad_norm": 8.351729393005371,
+      "learning_rate": 0.00011666937537244704,
+      "loss": 4.184175872802735,
+      "step": 15390
+    },
+    {
+      "epoch": 17.520352974665528,
+      "grad_norm": 7.465641021728516,
+      "learning_rate": 0.00011661520125683949,
+      "loss": 4.227563858032227,
+      "step": 15400
+    },
+    {
+      "epoch": 17.53173925419869,
+      "grad_norm": 7.468196392059326,
+      "learning_rate": 0.00011656102714123191,
+      "loss": 4.250960922241211,
+      "step": 15410
+    },
+    {
+      "epoch": 17.543125533731853,
+      "grad_norm": 7.476719856262207,
+      "learning_rate": 0.00011650685302562437,
+      "loss": 4.10565414428711,
+      "step": 15420
+    },
+    {
+      "epoch": 17.554511813265016,
+      "grad_norm": 7.9970269203186035,
+      "learning_rate": 0.00011645267891001682,
+      "loss": 4.217912292480468,
+      "step": 15430
+    },
+    {
+      "epoch": 17.56589809279818,
+      "grad_norm": 7.878209114074707,
+      "learning_rate": 0.00011639850479440924,
+      "loss": 4.038910293579102,
+      "step": 15440
+    },
+    {
+      "epoch": 17.57728437233134,
+      "grad_norm": 7.499887466430664,
+      "learning_rate": 0.00011634433067880168,
+      "loss": 4.3691764831542965,
+      "step": 15450
+    },
+    {
+      "epoch": 17.588670651864504,
+      "grad_norm": 7.418756484985352,
+      "learning_rate": 0.0001162901565631941,
+      "loss": 4.344051361083984,
+      "step": 15460
+    },
+    {
+      "epoch": 17.600056931397667,
+      "grad_norm": 8.542085647583008,
+      "learning_rate": 0.00011623598244758656,
+      "loss": 4.26641731262207,
+      "step": 15470
+    },
+    {
+      "epoch": 17.61144321093083,
+      "grad_norm": 8.016523361206055,
+      "learning_rate": 0.00011618180833197898,
+      "loss": 4.250433349609375,
+      "step": 15480
+    },
+    {
+      "epoch": 17.622829490463992,
+      "grad_norm": 7.668478488922119,
+      "learning_rate": 0.00011612763421637143,
+      "loss": 4.095648956298828,
+      "step": 15490
+    },
+    {
+      "epoch": 17.634215769997155,
+      "grad_norm": 7.5998101234436035,
+      "learning_rate": 0.00011607346010076386,
+      "loss": 4.096474075317383,
+      "step": 15500
+    },
+    {
+      "epoch": 17.645602049530314,
+      "grad_norm": 8.259299278259277,
+      "learning_rate": 0.00011601928598515631,
+      "loss": 4.329125213623047,
+      "step": 15510
+    },
+    {
+      "epoch": 17.656988329063477,
+      "grad_norm": 7.811594009399414,
+      "learning_rate": 0.00011596511186954873,
+      "loss": 4.401547622680664,
+      "step": 15520
+    },
+    {
+      "epoch": 17.66837460859664,
+      "grad_norm": 7.514590740203857,
+      "learning_rate": 0.00011591093775394117,
+      "loss": 4.361583709716797,
+      "step": 15530
+    },
+    {
+      "epoch": 17.679760888129803,
+      "grad_norm": 7.787680625915527,
+      "learning_rate": 0.00011585676363833362,
+      "loss": 4.375646591186523,
+      "step": 15540
+    },
+    {
+      "epoch": 17.691147167662965,
+      "grad_norm": 6.984362602233887,
+      "learning_rate": 0.00011580258952272605,
+      "loss": 4.231984329223633,
+      "step": 15550
+    },
+    {
+      "epoch": 17.702533447196128,
+      "grad_norm": 8.019062042236328,
+      "learning_rate": 0.0001157484154071185,
+      "loss": 4.141991806030274,
+      "step": 15560
+    },
+    {
+      "epoch": 17.71391972672929,
+      "grad_norm": 8.658064842224121,
+      "learning_rate": 0.00011569424129151092,
+      "loss": 4.37116813659668,
+      "step": 15570
+    },
+    {
+      "epoch": 17.725306006262453,
+      "grad_norm": 8.997333526611328,
+      "learning_rate": 0.00011564006717590336,
+      "loss": 4.313310241699218,
+      "step": 15580
+    },
+    {
+      "epoch": 17.736692285795616,
+      "grad_norm": 8.168388366699219,
+      "learning_rate": 0.0001155858930602958,
+      "loss": 4.10748291015625,
+      "step": 15590
+    },
+    {
+      "epoch": 17.74807856532878,
+      "grad_norm": 8.176225662231445,
+      "learning_rate": 0.00011553171894468824,
+      "loss": 4.249481964111328,
+      "step": 15600
+    },
+    {
+      "epoch": 17.75946484486194,
+      "grad_norm": 8.127287864685059,
+      "learning_rate": 0.00011547754482908066,
+      "loss": 4.257102203369141,
+      "step": 15610
+    },
+    {
+      "epoch": 17.770851124395104,
+      "grad_norm": 7.9209794998168945,
+      "learning_rate": 0.00011542337071347311,
+      "loss": 4.099342727661133,
+      "step": 15620
+    },
+    {
+      "epoch": 17.782237403928267,
+      "grad_norm": 7.507110118865967,
+      "learning_rate": 0.00011536919659786554,
+      "loss": 4.057368469238281,
+      "step": 15630
+    },
+    {
+      "epoch": 17.79362368346143,
+      "grad_norm": 7.258909702301025,
+      "learning_rate": 0.00011531502248225799,
+      "loss": 4.278286743164062,
+      "step": 15640
+    },
+    {
+      "epoch": 17.805009962994593,
+      "grad_norm": 8.5214204788208,
+      "learning_rate": 0.00011526084836665041,
+      "loss": 4.387587356567383,
+      "step": 15650
+    },
+    {
+      "epoch": 17.816396242527755,
+      "grad_norm": 6.49737548828125,
+      "learning_rate": 0.00011520667425104285,
+      "loss": 3.87442626953125,
+      "step": 15660
+    },
+    {
+      "epoch": 17.827782522060918,
+      "grad_norm": 8.327667236328125,
+      "learning_rate": 0.0001151525001354353,
+      "loss": 4.007180404663086,
+      "step": 15670
+    },
+    {
+      "epoch": 17.83916880159408,
+      "grad_norm": 8.513700485229492,
+      "learning_rate": 0.00011509832601982773,
+      "loss": 4.101554870605469,
+      "step": 15680
+    },
+    {
+      "epoch": 17.85055508112724,
+      "grad_norm": 8.8079833984375,
+      "learning_rate": 0.00011504415190422018,
+      "loss": 4.265046310424805,
+      "step": 15690
+    },
+    {
+      "epoch": 17.861941360660403,
+      "grad_norm": 7.73423957824707,
+      "learning_rate": 0.0001149899777886126,
+      "loss": 4.407987976074219,
+      "step": 15700
+    },
+    {
+      "epoch": 17.873327640193565,
+      "grad_norm": 7.101474285125732,
+      "learning_rate": 0.00011493580367300505,
+      "loss": 4.4353492736816404,
+      "step": 15710
+    },
+    {
+      "epoch": 17.884713919726728,
+      "grad_norm": 7.298971652984619,
+      "learning_rate": 0.00011488162955739748,
+      "loss": 4.033404159545898,
+      "step": 15720
+    },
+    {
+      "epoch": 17.89610019925989,
+      "grad_norm": 7.350695610046387,
+      "learning_rate": 0.00011482745544178992,
+      "loss": 4.468685150146484,
+      "step": 15730
+    },
+    {
+      "epoch": 17.907486478793054,
+      "grad_norm": 7.839898109436035,
+      "learning_rate": 0.00011477328132618234,
+      "loss": 4.322669601440429,
+      "step": 15740
+    },
+    {
+      "epoch": 17.918872758326216,
+      "grad_norm": 7.159393787384033,
+      "learning_rate": 0.00011471910721057479,
+      "loss": 4.0323749542236325,
+      "step": 15750
+    },
+    {
+      "epoch": 17.93025903785938,
+      "grad_norm": 7.601138591766357,
+      "learning_rate": 0.00011466493309496722,
+      "loss": 4.170376586914062,
+      "step": 15760
+    },
+    {
+      "epoch": 17.941645317392542,
+      "grad_norm": 7.914937496185303,
+      "learning_rate": 0.00011461075897935967,
+      "loss": 4.421757507324219,
+      "step": 15770
+    },
+    {
+      "epoch": 17.953031596925705,
+      "grad_norm": 7.36163854598999,
+      "learning_rate": 0.00011455658486375212,
+      "loss": 4.248828506469726,
+      "step": 15780
+    },
+    {
+      "epoch": 17.964417876458867,
+      "grad_norm": 7.326484680175781,
+      "learning_rate": 0.00011450241074814454,
+      "loss": 4.062700271606445,
+      "step": 15790
+    },
+    {
+      "epoch": 17.97580415599203,
+      "grad_norm": 7.293338775634766,
+      "learning_rate": 0.00011444823663253698,
+      "loss": 4.451213455200195,
+      "step": 15800
+    },
+    {
+      "epoch": 17.987190435525193,
+      "grad_norm": 7.929836273193359,
+      "learning_rate": 0.00011439406251692941,
+      "loss": 4.008718109130859,
+      "step": 15810
+    },
+    {
+      "epoch": 17.998576715058356,
+      "grad_norm": 7.278785705566406,
+      "learning_rate": 0.00011433988840132186,
+      "loss": 4.249869537353516,
+      "step": 15820
+    },
+    {
+      "epoch": 18.00910902362653,
+      "grad_norm": 7.38397216796875,
+      "learning_rate": 0.00011428571428571428,
+      "loss": 3.487931823730469,
+      "step": 15830
+    },
+    {
+      "epoch": 18.02049530315969,
+      "grad_norm": 7.779951572418213,
+      "learning_rate": 0.00011423154017010674,
+      "loss": 3.6420997619628905,
+      "step": 15840
+    },
+    {
+      "epoch": 18.031881582692854,
+      "grad_norm": 7.421573162078857,
+      "learning_rate": 0.00011417736605449916,
+      "loss": 3.697202682495117,
+      "step": 15850
+    },
+    {
+      "epoch": 18.043267862226017,
+      "grad_norm": 7.718578815460205,
+      "learning_rate": 0.00011412319193889161,
+      "loss": 3.6515987396240233,
+      "step": 15860
+    },
+    {
+      "epoch": 18.05465414175918,
+      "grad_norm": 8.21097183227539,
+      "learning_rate": 0.0001140744352348448,
+      "loss": 4.043758392333984,
+      "step": 15870
+    },
+    {
+      "epoch": 18.066040421292342,
+      "grad_norm": 8.3043794631958,
+      "learning_rate": 0.00011402026111923723,
+      "loss": 3.8113327026367188,
+      "step": 15880
+    },
+    {
+      "epoch": 18.077426700825505,
+      "grad_norm": 7.186902046203613,
+      "learning_rate": 0.00011396608700362967,
+      "loss": 3.746054458618164,
+      "step": 15890
+    },
+    {
+      "epoch": 18.088812980358668,
+      "grad_norm": 7.123048305511475,
+      "learning_rate": 0.0001139119128880221,
+      "loss": 3.691455078125,
+      "step": 15900
+    },
+    {
+      "epoch": 18.10019925989183,
+      "grad_norm": 7.646660804748535,
+      "learning_rate": 0.00011385773877241455,
+      "loss": 3.715811538696289,
+      "step": 15910
+    },
+    {
+      "epoch": 18.111585539424993,
+      "grad_norm": 7.243655204772949,
+      "learning_rate": 0.000113803564656807,
+      "loss": 3.995361328125,
+      "step": 15920
+    },
+    {
+      "epoch": 18.122971818958156,
+      "grad_norm": 7.36957311630249,
+      "learning_rate": 0.00011374939054119942,
+      "loss": 3.660219192504883,
+      "step": 15930
+    },
+    {
+      "epoch": 18.13435809849132,
+      "grad_norm": 8.030665397644043,
+      "learning_rate": 0.00011369521642559187,
+      "loss": 3.9243896484375,
+      "step": 15940
+    },
+    {
+      "epoch": 18.14574437802448,
+      "grad_norm": 7.206502914428711,
+      "learning_rate": 0.0001136410423099843,
+      "loss": 3.791769027709961,
+      "step": 15950
+    },
+    {
+      "epoch": 18.157130657557644,
+      "grad_norm": 8.390379905700684,
+      "learning_rate": 0.00011358686819437674,
+      "loss": 3.6325794219970704,
+      "step": 15960
+    },
+    {
+      "epoch": 18.168516937090807,
+      "grad_norm": 7.630048751831055,
+      "learning_rate": 0.00011353269407876916,
+      "loss": 3.683578872680664,
+      "step": 15970
+    },
+    {
+      "epoch": 18.17990321662397,
+      "grad_norm": 7.336859703063965,
+      "learning_rate": 0.00011347851996316161,
+      "loss": 3.736345672607422,
+      "step": 15980
+    },
+    {
+      "epoch": 18.19128949615713,
+      "grad_norm": 6.535525798797607,
+      "learning_rate": 0.00011342434584755404,
+      "loss": 3.869864654541016,
+      "step": 15990
+    },
+    {
+      "epoch": 18.20267577569029,
+      "grad_norm": 7.941544532775879,
+      "learning_rate": 0.00011337017173194649,
+      "loss": 3.8768684387207033,
+      "step": 16000
+    },
+    {
+      "epoch": 18.214062055223454,
+      "grad_norm": 8.93309497833252,
+      "learning_rate": 0.00011331599761633891,
+      "loss": 4.1240489959716795,
+      "step": 16010
+    },
+    {
+      "epoch": 18.225448334756617,
+      "grad_norm": 8.11670970916748,
+      "learning_rate": 0.00011326182350073135,
+      "loss": 3.7689064025878904,
+      "step": 16020
+    },
+    {
+      "epoch": 18.23683461428978,
+      "grad_norm": 8.06661319732666,
+      "learning_rate": 0.0001132076493851238,
+      "loss": 3.8555408477783204,
+      "step": 16030
+    },
+    {
+      "epoch": 18.248220893822943,
+      "grad_norm": 7.596058368682861,
+      "learning_rate": 0.00011315347526951623,
+      "loss": 3.899448013305664,
+      "step": 16040
+    },
+    {
+      "epoch": 18.259607173356105,
+      "grad_norm": 7.202267646789551,
+      "learning_rate": 0.00011309930115390868,
+      "loss": 3.954229736328125,
+      "step": 16050
+    },
+    {
+      "epoch": 18.270993452889268,
+      "grad_norm": 6.913619518280029,
+      "learning_rate": 0.0001130451270383011,
+      "loss": 4.055387496948242,
+      "step": 16060
+    },
+    {
+      "epoch": 18.28237973242243,
+      "grad_norm": 7.013467311859131,
+      "learning_rate": 0.00011299095292269355,
+      "loss": 3.91961784362793,
+      "step": 16070
+    },
+    {
+      "epoch": 18.293766011955594,
+      "grad_norm": 8.460749626159668,
+      "learning_rate": 0.00011293677880708598,
+      "loss": 3.9316688537597657,
+      "step": 16080
+    },
+    {
+      "epoch": 18.305152291488756,
+      "grad_norm": 6.905110836029053,
+      "learning_rate": 0.00011288260469147842,
+      "loss": 4.052811813354492,
+      "step": 16090
+    },
+    {
+      "epoch": 18.31653857102192,
+      "grad_norm": 6.9299116134643555,
+      "learning_rate": 0.00011282843057587084,
+      "loss": 3.9361194610595702,
+      "step": 16100
+    },
+    {
+      "epoch": 18.327924850555082,
+      "grad_norm": 6.814231872558594,
+      "learning_rate": 0.00011277425646026329,
+      "loss": 3.73162841796875,
+      "step": 16110
+    },
+    {
+      "epoch": 18.339311130088245,
+      "grad_norm": 7.320927619934082,
+      "learning_rate": 0.00011272008234465572,
+      "loss": 3.8358139038085937,
+      "step": 16120
+    },
+    {
+      "epoch": 18.350697409621407,
+      "grad_norm": 7.525396823883057,
+      "learning_rate": 0.00011266590822904817,
+      "loss": 3.948038864135742,
+      "step": 16130
+    },
+    {
+      "epoch": 18.36208368915457,
+      "grad_norm": 8.188949584960938,
+      "learning_rate": 0.00011261173411344062,
+      "loss": 4.153988647460937,
+      "step": 16140
+    },
+    {
+      "epoch": 18.373469968687733,
+      "grad_norm": 7.659078121185303,
+      "learning_rate": 0.00011255755999783304,
+      "loss": 3.8572532653808596,
+      "step": 16150
+    },
+    {
+      "epoch": 18.384856248220895,
+      "grad_norm": 7.888333320617676,
+      "learning_rate": 0.00011250338588222548,
+      "loss": 4.0505119323730465,
+      "step": 16160
+    },
+    {
+      "epoch": 18.396242527754055,
+      "grad_norm": 9.681838989257812,
+      "learning_rate": 0.0001124492117666179,
+      "loss": 3.6657783508300783,
+      "step": 16170
+    },
+    {
+      "epoch": 18.407628807287217,
+      "grad_norm": 7.816950798034668,
+      "learning_rate": 0.00011239503765101036,
+      "loss": 4.182400894165039,
+      "step": 16180
+    },
+    {
+      "epoch": 18.41901508682038,
+      "grad_norm": 8.955124855041504,
+      "learning_rate": 0.00011234086353540278,
+      "loss": 4.1284526824951175,
+      "step": 16190
+    },
+    {
+      "epoch": 18.430401366353543,
+      "grad_norm": 7.874047756195068,
+      "learning_rate": 0.00011228668941979523,
+      "loss": 4.089582824707032,
+      "step": 16200
+    },
+    {
+      "epoch": 18.441787645886706,
+      "grad_norm": 7.964369297027588,
+      "learning_rate": 0.00011223251530418766,
+      "loss": 4.226484298706055,
+      "step": 16210
+    },
+    {
+      "epoch": 18.45317392541987,
+      "grad_norm": 7.5684614181518555,
+      "learning_rate": 0.00011217834118858011,
+      "loss": 3.7357833862304686,
+      "step": 16220
+    },
+    {
+      "epoch": 18.46456020495303,
+      "grad_norm": 8.19275188446045,
+      "learning_rate": 0.00011212416707297253,
+      "loss": 3.658647918701172,
+      "step": 16230
+    },
+    {
+      "epoch": 18.475946484486194,
+      "grad_norm": 7.265321731567383,
+      "learning_rate": 0.00011206999295736497,
+      "loss": 4.110098266601563,
+      "step": 16240
+    },
+    {
+      "epoch": 18.487332764019357,
+      "grad_norm": 8.811004638671875,
+      "learning_rate": 0.0001120158188417574,
+      "loss": 4.014685821533203,
+      "step": 16250
+    },
+    {
+      "epoch": 18.49871904355252,
+      "grad_norm": 8.132495880126953,
+      "learning_rate": 0.00011196164472614985,
+      "loss": 3.972939682006836,
+      "step": 16260
+    },
+    {
+      "epoch": 18.510105323085682,
+      "grad_norm": 7.8982672691345215,
+      "learning_rate": 0.0001119074706105423,
+      "loss": 3.900859069824219,
+      "step": 16270
+    },
+    {
+      "epoch": 18.521491602618845,
+      "grad_norm": 8.57484245300293,
+      "learning_rate": 0.00011185329649493472,
+      "loss": 3.8809505462646485,
+      "step": 16280
+    },
+    {
+      "epoch": 18.532877882152007,
+      "grad_norm": 8.151063919067383,
+      "learning_rate": 0.00011179912237932718,
+      "loss": 3.9721370697021485,
+      "step": 16290
+    },
+    {
+      "epoch": 18.54426416168517,
+      "grad_norm": 7.32830810546875,
+      "learning_rate": 0.0001117449482637196,
+      "loss": 4.091455078125,
+      "step": 16300
+    },
+    {
+      "epoch": 18.555650441218333,
+      "grad_norm": 7.877411842346191,
+      "learning_rate": 0.00011169077414811204,
+      "loss": 3.9833248138427733,
+      "step": 16310
+    },
+    {
+      "epoch": 18.567036720751496,
+      "grad_norm": 7.7397379875183105,
+      "learning_rate": 0.00011163660003250446,
+      "loss": 3.792256164550781,
+      "step": 16320
+    },
+    {
+      "epoch": 18.57842300028466,
+      "grad_norm": 7.100308418273926,
+      "learning_rate": 0.00011158242591689692,
+      "loss": 3.683748245239258,
+      "step": 16330
+    },
+    {
+      "epoch": 18.58980927981782,
+      "grad_norm": 8.094539642333984,
+      "learning_rate": 0.00011152825180128934,
+      "loss": 4.14416732788086,
+      "step": 16340
+    },
+    {
+      "epoch": 18.60119555935098,
+      "grad_norm": 7.870161056518555,
+      "learning_rate": 0.00011147407768568179,
+      "loss": 3.89129753112793,
+      "step": 16350
+    },
+    {
+      "epoch": 18.612581838884143,
+      "grad_norm": 7.73281192779541,
+      "learning_rate": 0.00011141990357007422,
+      "loss": 4.153730773925782,
+      "step": 16360
+    },
+    {
+      "epoch": 18.623968118417306,
+      "grad_norm": 7.23974084854126,
+      "learning_rate": 0.00011136572945446667,
+      "loss": 3.8958515167236327,
+      "step": 16370
+    },
+    {
+      "epoch": 18.63535439795047,
+      "grad_norm": 11.619794845581055,
+      "learning_rate": 0.0001113115553388591,
+      "loss": 4.148723220825195,
+      "step": 16380
+    },
+    {
+      "epoch": 18.64674067748363,
+      "grad_norm": 7.968571186065674,
+      "learning_rate": 0.00011125738122325153,
+      "loss": 4.051975250244141,
+      "step": 16390
+    },
+    {
+      "epoch": 18.658126957016794,
+      "grad_norm": 8.169166564941406,
+      "learning_rate": 0.00011120320710764398,
+      "loss": 3.804905319213867,
+      "step": 16400
+    },
+    {
+      "epoch": 18.669513236549957,
+      "grad_norm": 7.773144721984863,
+      "learning_rate": 0.0001111490329920364,
+      "loss": 3.9082794189453125,
+      "step": 16410
+    },
+    {
+      "epoch": 18.68089951608312,
+      "grad_norm": 8.190699577331543,
+      "learning_rate": 0.00011109485887642886,
+      "loss": 4.108190536499023,
+      "step": 16420
+    },
+    {
+      "epoch": 18.692285795616282,
+      "grad_norm": 7.462385177612305,
+      "learning_rate": 0.00011104068476082128,
+      "loss": 4.018275451660156,
+      "step": 16430
+    },
+    {
+      "epoch": 18.703672075149445,
+      "grad_norm": 8.5626859664917,
+      "learning_rate": 0.00011098651064521372,
+      "loss": 3.7356529235839844,
+      "step": 16440
+    },
+    {
+      "epoch": 18.715058354682608,
+      "grad_norm": 8.161300659179688,
+      "learning_rate": 0.00011093233652960616,
+      "loss": 4.018674850463867,
+      "step": 16450
+    },
+    {
+      "epoch": 18.72644463421577,
+      "grad_norm": 7.462896347045898,
+      "learning_rate": 0.0001108781624139986,
+      "loss": 3.7080970764160157,
+      "step": 16460
+    },
+    {
+      "epoch": 18.737830913748933,
+      "grad_norm": 7.781260967254639,
+      "learning_rate": 0.00011082398829839102,
+      "loss": 4.161582183837891,
+      "step": 16470
+    },
+    {
+      "epoch": 18.749217193282096,
+      "grad_norm": 7.212603569030762,
+      "learning_rate": 0.00011076981418278347,
+      "loss": 3.770726776123047,
+      "step": 16480
+    },
+    {
+      "epoch": 18.76060347281526,
+      "grad_norm": 8.329229354858398,
+      "learning_rate": 0.00011071564006717592,
+      "loss": 4.128889083862305,
+      "step": 16490
+    },
+    {
+      "epoch": 18.77198975234842,
+      "grad_norm": 7.012000560760498,
+      "learning_rate": 0.00011066146595156835,
+      "loss": 4.022500991821289,
+      "step": 16500
+    },
+    {
+      "epoch": 18.783376031881584,
+      "grad_norm": 7.675992965698242,
+      "learning_rate": 0.00011060729183596079,
+      "loss": 3.725032424926758,
+      "step": 16510
+    },
+    {
+      "epoch": 18.794762311414747,
+      "grad_norm": 7.538510322570801,
+      "learning_rate": 0.00011055311772035321,
+      "loss": 3.657960891723633,
+      "step": 16520
+    },
+    {
+      "epoch": 18.806148590947906,
+      "grad_norm": 6.402365207672119,
+      "learning_rate": 0.00011049894360474566,
+      "loss": 3.9226734161376955,
+      "step": 16530
+    },
+    {
+      "epoch": 18.81753487048107,
+      "grad_norm": 6.849089622497559,
+      "learning_rate": 0.00011044476948913809,
+      "loss": 3.683644104003906,
+      "step": 16540
+    },
+    {
+      "epoch": 18.82892115001423,
+      "grad_norm": 7.539620399475098,
+      "learning_rate": 0.00011039059537353054,
+      "loss": 3.788827896118164,
+      "step": 16550
+    },
+    {
+      "epoch": 18.840307429547394,
+      "grad_norm": 8.003284454345703,
+      "learning_rate": 0.00011033642125792296,
+      "loss": 4.023321533203125,
+      "step": 16560
+    },
+    {
+      "epoch": 18.851693709080557,
+      "grad_norm": 8.320795059204102,
+      "learning_rate": 0.00011028224714231541,
+      "loss": 3.955525588989258,
+      "step": 16570
+    },
+    {
+      "epoch": 18.86307998861372,
+      "grad_norm": 8.324459075927734,
+      "learning_rate": 0.00011022807302670784,
+      "loss": 3.9589923858642577,
+      "step": 16580
+    },
+    {
+      "epoch": 18.874466268146882,
+      "grad_norm": 8.919875144958496,
+      "learning_rate": 0.00011017389891110028,
+      "loss": 4.231829833984375,
+      "step": 16590
+    },
+    {
+      "epoch": 18.885852547680045,
+      "grad_norm": 7.729628562927246,
+      "learning_rate": 0.00011011972479549273,
+      "loss": 3.9778141021728515,
+      "step": 16600
+    },
+    {
+      "epoch": 18.897238827213208,
+      "grad_norm": 7.793543815612793,
+      "learning_rate": 0.00011006555067988515,
+      "loss": 4.103470230102539,
+      "step": 16610
+    },
+    {
+      "epoch": 18.90862510674637,
+      "grad_norm": 7.603933334350586,
+      "learning_rate": 0.0001100113765642776,
+      "loss": 3.918026351928711,
+      "step": 16620
+    },
+    {
+      "epoch": 18.920011386279533,
+      "grad_norm": 9.164023399353027,
+      "learning_rate": 0.00010995720244867003,
+      "loss": 4.239971542358399,
+      "step": 16630
+    },
+    {
+      "epoch": 18.931397665812696,
+      "grad_norm": 8.179018020629883,
+      "learning_rate": 0.00010990302833306248,
+      "loss": 4.0833740234375,
+      "step": 16640
+    },
+    {
+      "epoch": 18.94278394534586,
+      "grad_norm": 8.107283592224121,
+      "learning_rate": 0.0001098488542174549,
+      "loss": 3.886615753173828,
+      "step": 16650
+    },
+    {
+      "epoch": 18.95417022487902,
+      "grad_norm": 8.483085632324219,
+      "learning_rate": 0.00010979468010184734,
+      "loss": 4.466776275634766,
+      "step": 16660
+    },
+    {
+      "epoch": 18.965556504412184,
+      "grad_norm": 7.803193092346191,
+      "learning_rate": 0.00010974050598623977,
+      "loss": 3.9009479522705077,
+      "step": 16670
+    },
+    {
+      "epoch": 18.976942783945347,
+      "grad_norm": 8.053387641906738,
+      "learning_rate": 0.00010968633187063222,
+      "loss": 4.092340469360352,
+      "step": 16680
+    },
+    {
+      "epoch": 18.98832906347851,
+      "grad_norm": 7.657060623168945,
+      "learning_rate": 0.00010963215775502464,
+      "loss": 4.109097290039062,
+      "step": 16690
+    },
+    {
+      "epoch": 18.999715343011673,
+      "grad_norm": 7.931739807128906,
+      "learning_rate": 0.0001095779836394171,
+      "loss": 3.970943069458008,
+      "step": 16700
+    },
+    {
+      "epoch": 19.010247651579846,
+      "grad_norm": 8.304778099060059,
+      "learning_rate": 0.00010952380952380953,
+      "loss": 3.4373973846435546,
+      "step": 16710
+    },
+    {
+      "epoch": 19.02163393111301,
+      "grad_norm": 8.502395629882812,
+      "learning_rate": 0.00010946963540820197,
+      "loss": 3.6647174835205076,
+      "step": 16720
+    },
+    {
+      "epoch": 19.03302021064617,
+      "grad_norm": 8.930685043334961,
+      "learning_rate": 0.00010941546129259441,
+      "loss": 3.4595081329345705,
+      "step": 16730
+    },
+    {
+      "epoch": 19.044406490179334,
+      "grad_norm": 7.8097920417785645,
+      "learning_rate": 0.00010936128717698683,
+      "loss": 3.548003005981445,
+      "step": 16740
+    },
+    {
+      "epoch": 19.055792769712497,
+      "grad_norm": 6.608278751373291,
+      "learning_rate": 0.00010930711306137929,
+      "loss": 3.3133895874023436,
+      "step": 16750
+    },
+    {
+      "epoch": 19.06717904924566,
+      "grad_norm": 7.428323745727539,
+      "learning_rate": 0.00010925293894577171,
+      "loss": 3.576006317138672,
+      "step": 16760
+    },
+    {
+      "epoch": 19.078565328778822,
+      "grad_norm": 7.900843620300293,
+      "learning_rate": 0.00010919876483016416,
+      "loss": 3.676096725463867,
+      "step": 16770
+    },
+    {
+      "epoch": 19.089951608311985,
+      "grad_norm": 7.658603191375732,
+      "learning_rate": 0.00010914459071455659,
+      "loss": 3.7328716278076173,
+      "step": 16780
+    },
+    {
+      "epoch": 19.101337887845148,
+      "grad_norm": 7.360223293304443,
+      "learning_rate": 0.00010909041659894902,
+      "loss": 3.652063751220703,
+      "step": 16790
+    },
+    {
+      "epoch": 19.11272416737831,
+      "grad_norm": 7.122159957885742,
+      "learning_rate": 0.00010903624248334146,
+      "loss": 3.690329360961914,
+      "step": 16800
+    },
+    {
+      "epoch": 19.124110446911473,
+      "grad_norm": 6.8670501708984375,
+      "learning_rate": 0.0001089820683677339,
+      "loss": 3.667049026489258,
+      "step": 16810
+    },
+    {
+      "epoch": 19.135496726444636,
+      "grad_norm": 7.987428665161133,
+      "learning_rate": 0.00010892789425212635,
+      "loss": 3.607272720336914,
+      "step": 16820
+    },
+    {
+      "epoch": 19.146883005977795,
+      "grad_norm": 7.282992839813232,
+      "learning_rate": 0.00010887372013651878,
+      "loss": 3.4856422424316404,
+      "step": 16830
+    },
+    {
+      "epoch": 19.158269285510958,
+      "grad_norm": 8.548340797424316,
+      "learning_rate": 0.00010881954602091123,
+      "loss": 3.8047679901123046,
+      "step": 16840
+    },
+    {
+      "epoch": 19.16965556504412,
+      "grad_norm": 7.989595413208008,
+      "learning_rate": 0.00010876537190530365,
+      "loss": 3.664881134033203,
+      "step": 16850
+    },
+    {
+      "epoch": 19.181041844577283,
+      "grad_norm": 7.799252033233643,
+      "learning_rate": 0.00010871119778969609,
+      "loss": 4.083931350708008,
+      "step": 16860
+    },
+    {
+      "epoch": 19.192428124110446,
+      "grad_norm": 7.790616512298584,
+      "learning_rate": 0.00010865702367408851,
+      "loss": 3.279020309448242,
+      "step": 16870
+    },
+    {
+      "epoch": 19.20381440364361,
+      "grad_norm": 7.22285795211792,
+      "learning_rate": 0.00010860284955848097,
+      "loss": 3.5854366302490233,
+      "step": 16880
+    },
+    {
+      "epoch": 19.21520068317677,
+      "grad_norm": 8.332810401916504,
+      "learning_rate": 0.00010854867544287339,
+      "loss": 3.9076087951660154,
+      "step": 16890
+    },
+    {
+      "epoch": 19.226586962709934,
+      "grad_norm": 6.827261924743652,
+      "learning_rate": 0.00010849450132726584,
+      "loss": 3.7439952850341798,
+      "step": 16900
+    },
+    {
+      "epoch": 19.237973242243097,
+      "grad_norm": 7.577035427093506,
+      "learning_rate": 0.00010844032721165827,
+      "loss": 3.5914188385009767,
+      "step": 16910
+    },
+    {
+      "epoch": 19.24935952177626,
+      "grad_norm": 8.372159957885742,
+      "learning_rate": 0.00010838615309605072,
+      "loss": 3.650664520263672,
+      "step": 16920
+    },
+    {
+      "epoch": 19.260745801309422,
+      "grad_norm": 7.719248294830322,
+      "learning_rate": 0.00010833197898044314,
+      "loss": 3.8901805877685547,
+      "step": 16930
+    },
+    {
+      "epoch": 19.272132080842585,
+      "grad_norm": 6.905912399291992,
+      "learning_rate": 0.00010827780486483558,
+      "loss": 3.5376930236816406,
+      "step": 16940
+    },
+    {
+      "epoch": 19.283518360375748,
+      "grad_norm": 6.742889404296875,
+      "learning_rate": 0.00010822363074922803,
+      "loss": 3.475457763671875,
+      "step": 16950
+    },
+    {
+      "epoch": 19.29490463990891,
+      "grad_norm": 7.5838446617126465,
+      "learning_rate": 0.00010816945663362046,
+      "loss": 3.4266204833984375,
+      "step": 16960
+    },
+    {
+      "epoch": 19.306290919442073,
+      "grad_norm": 7.7189836502075195,
+      "learning_rate": 0.00010811528251801291,
+      "loss": 3.5107288360595703,
+      "step": 16970
+    },
+    {
+      "epoch": 19.317677198975236,
+      "grad_norm": 7.462057590484619,
+      "learning_rate": 0.00010806110840240533,
+      "loss": 3.7074642181396484,
+      "step": 16980
+    },
+    {
+      "epoch": 19.3290634785084,
+      "grad_norm": 8.049522399902344,
+      "learning_rate": 0.00010800693428679778,
+      "loss": 3.521027374267578,
+      "step": 16990
+    },
+    {
+      "epoch": 19.34044975804156,
+      "grad_norm": 7.91304349899292,
+      "learning_rate": 0.00010795276017119021,
+      "loss": 3.9731681823730467,
+      "step": 17000
+    },
+    {
+      "epoch": 19.35183603757472,
+      "grad_norm": 8.217313766479492,
+      "learning_rate": 0.00010789858605558265,
+      "loss": 3.8252967834472655,
+      "step": 17010
+    },
+    {
+      "epoch": 19.363222317107883,
+      "grad_norm": 8.448884963989258,
+      "learning_rate": 0.00010784441193997507,
+      "loss": 3.677239990234375,
+      "step": 17020
+    },
+    {
+      "epoch": 19.374608596641046,
+      "grad_norm": 7.586091041564941,
+      "learning_rate": 0.00010779023782436752,
+      "loss": 3.6121559143066406,
+      "step": 17030
+    },
+    {
+      "epoch": 19.38599487617421,
+      "grad_norm": 7.399051189422607,
+      "learning_rate": 0.00010773606370875995,
+      "loss": 3.6527149200439455,
+      "step": 17040
+    },
+    {
+      "epoch": 19.39738115570737,
+      "grad_norm": 8.452932357788086,
+      "learning_rate": 0.0001076818895931524,
+      "loss": 3.5067722320556642,
+      "step": 17050
+    },
+    {
+      "epoch": 19.408767435240534,
+      "grad_norm": 7.3610382080078125,
+      "learning_rate": 0.00010762771547754485,
+      "loss": 3.557135009765625,
+      "step": 17060
+    },
+    {
+      "epoch": 19.420153714773697,
+      "grad_norm": 7.684902191162109,
+      "learning_rate": 0.00010757354136193727,
+      "loss": 3.7063770294189453,
+      "step": 17070
+    },
+    {
+      "epoch": 19.43153999430686,
+      "grad_norm": 7.2472825050354,
+      "learning_rate": 0.00010751936724632971,
+      "loss": 3.5020095825195314,
+      "step": 17080
+    },
+    {
+      "epoch": 19.442926273840023,
+      "grad_norm": 8.710956573486328,
+      "learning_rate": 0.00010746519313072214,
+      "loss": 3.7144115447998045,
+      "step": 17090
+    },
+    {
+      "epoch": 19.454312553373185,
+      "grad_norm": 7.865204811096191,
+      "learning_rate": 0.00010741101901511459,
+      "loss": 3.860599136352539,
+      "step": 17100
+    },
+    {
+      "epoch": 19.465698832906348,
+      "grad_norm": 7.315329551696777,
+      "learning_rate": 0.00010735684489950701,
+      "loss": 3.8806697845458986,
+      "step": 17110
+    },
+    {
+      "epoch": 19.47708511243951,
+      "grad_norm": 8.162956237792969,
+      "learning_rate": 0.00010730267078389946,
+      "loss": 3.6206798553466797,
+      "step": 17120
+    },
+    {
+      "epoch": 19.488471391972674,
+      "grad_norm": 7.473461627960205,
+      "learning_rate": 0.00010724849666829189,
+      "loss": 3.7630165100097654,
+      "step": 17130
+    },
+    {
+      "epoch": 19.499857671505836,
+      "grad_norm": 7.086675643920898,
+      "learning_rate": 0.00010719432255268434,
+      "loss": 3.6255001068115233,
+      "step": 17140
+    },
+    {
+      "epoch": 19.511243951039,
+      "grad_norm": 7.986152172088623,
+      "learning_rate": 0.00010714014843707677,
+      "loss": 3.654983139038086,
+      "step": 17150
+    },
+    {
+      "epoch": 19.52263023057216,
+      "grad_norm": 6.950078010559082,
+      "learning_rate": 0.0001070859743214692,
+      "loss": 3.5265995025634767,
+      "step": 17160
+    },
+    {
+      "epoch": 19.534016510105324,
+      "grad_norm": 8.496960639953613,
+      "learning_rate": 0.00010703180020586165,
+      "loss": 3.8141246795654298,
+      "step": 17170
+    },
+    {
+      "epoch": 19.545402789638487,
+      "grad_norm": 7.757366180419922,
+      "learning_rate": 0.00010697762609025408,
+      "loss": 3.748949432373047,
+      "step": 17180
+    },
+    {
+      "epoch": 19.55678906917165,
+      "grad_norm": 8.865738868713379,
+      "learning_rate": 0.00010692345197464653,
+      "loss": 3.841624069213867,
+      "step": 17190
+    },
+    {
+      "epoch": 19.56817534870481,
+      "grad_norm": 8.705485343933105,
+      "learning_rate": 0.00010686927785903896,
+      "loss": 3.5993213653564453,
+      "step": 17200
+    },
+    {
+      "epoch": 19.579561628237972,
+      "grad_norm": 8.318339347839355,
+      "learning_rate": 0.0001068151037434314,
+      "loss": 3.897611618041992,
+      "step": 17210
+    },
+    {
+      "epoch": 19.590947907771135,
+      "grad_norm": 8.347416877746582,
+      "learning_rate": 0.00010676092962782382,
+      "loss": 3.714302825927734,
+      "step": 17220
+    },
+    {
+      "epoch": 19.602334187304297,
+      "grad_norm": 6.775099754333496,
+      "learning_rate": 0.00010670675551221627,
+      "loss": 3.8760570526123046,
+      "step": 17230
+    },
+    {
+      "epoch": 19.61372046683746,
+      "grad_norm": 8.375359535217285,
+      "learning_rate": 0.0001066525813966087,
+      "loss": 3.7001274108886717,
+      "step": 17240
+    },
+    {
+      "epoch": 19.625106746370623,
+      "grad_norm": 8.114470481872559,
+      "learning_rate": 0.00010659840728100115,
+      "loss": 3.857496643066406,
+      "step": 17250
+    },
+    {
+      "epoch": 19.636493025903786,
+      "grad_norm": 8.257019996643066,
+      "learning_rate": 0.00010654423316539357,
+      "loss": 3.6363399505615233,
+      "step": 17260
+    },
+    {
+      "epoch": 19.64787930543695,
+      "grad_norm": 7.418999671936035,
+      "learning_rate": 0.00010649005904978602,
+      "loss": 3.6371334075927733,
+      "step": 17270
+    },
+    {
+      "epoch": 19.65926558497011,
+      "grad_norm": 8.225639343261719,
+      "learning_rate": 0.00010643588493417846,
+      "loss": 3.6147201538085936,
+      "step": 17280
+    },
+    {
+      "epoch": 19.670651864503274,
+      "grad_norm": 7.634995937347412,
+      "learning_rate": 0.00010638171081857088,
+      "loss": 3.783678436279297,
+      "step": 17290
+    },
+    {
+      "epoch": 19.682038144036436,
+      "grad_norm": 7.631542682647705,
+      "learning_rate": 0.00010632753670296334,
+      "loss": 3.7224945068359374,
+      "step": 17300
+    },
+    {
+      "epoch": 19.6934244235696,
+      "grad_norm": 8.868590354919434,
+      "learning_rate": 0.00010627336258735576,
+      "loss": 3.8564884185791017,
+      "step": 17310
+    },
+    {
+      "epoch": 19.704810703102762,
+      "grad_norm": 7.66297721862793,
+      "learning_rate": 0.00010621918847174821,
+      "loss": 3.8662670135498045,
+      "step": 17320
+    },
+    {
+      "epoch": 19.716196982635925,
+      "grad_norm": 7.4999589920043945,
+      "learning_rate": 0.00010616501435614064,
+      "loss": 3.648085021972656,
+      "step": 17330
+    },
+    {
+      "epoch": 19.727583262169087,
+      "grad_norm": 7.6754984855651855,
+      "learning_rate": 0.00010611084024053309,
+      "loss": 3.6652565002441406,
+      "step": 17340
+    },
+    {
+      "epoch": 19.73896954170225,
+      "grad_norm": 7.409824371337891,
+      "learning_rate": 0.00010605666612492551,
+      "loss": 3.8480754852294923,
+      "step": 17350
+    },
+    {
+      "epoch": 19.750355821235413,
+      "grad_norm": 7.804997444152832,
+      "learning_rate": 0.00010600249200931795,
+      "loss": 3.830272674560547,
+      "step": 17360
+    },
+    {
+      "epoch": 19.761742100768572,
+      "grad_norm": 9.533950805664062,
+      "learning_rate": 0.00010594831789371037,
+      "loss": 4.024043273925781,
+      "step": 17370
+    },
+    {
+      "epoch": 19.773128380301735,
+      "grad_norm": 7.675684452056885,
+      "learning_rate": 0.00010589414377810283,
+      "loss": 3.693775939941406,
+      "step": 17380
+    },
+    {
+      "epoch": 19.784514659834898,
+      "grad_norm": 8.96830940246582,
+      "learning_rate": 0.00010583996966249528,
+      "loss": 3.858016586303711,
+      "step": 17390
+    },
+    {
+      "epoch": 19.79590093936806,
+      "grad_norm": 6.618140697479248,
+      "learning_rate": 0.0001057857955468877,
+      "loss": 3.8667743682861326,
+      "step": 17400
+    },
+    {
+      "epoch": 19.807287218901223,
+      "grad_norm": 8.536417961120605,
+      "learning_rate": 0.00010573162143128015,
+      "loss": 3.633957290649414,
+      "step": 17410
+    },
+    {
+      "epoch": 19.818673498434386,
+      "grad_norm": 8.237014770507812,
+      "learning_rate": 0.00010567744731567258,
+      "loss": 3.944471740722656,
+      "step": 17420
+    },
+    {
+      "epoch": 19.83005977796755,
+      "grad_norm": 7.2452545166015625,
+      "learning_rate": 0.00010562327320006502,
+      "loss": 3.8378509521484374,
+      "step": 17430
+    },
+    {
+      "epoch": 19.84144605750071,
+      "grad_norm": 6.854305744171143,
+      "learning_rate": 0.00010556909908445744,
+      "loss": 3.6727760314941404,
+      "step": 17440
+    },
+    {
+      "epoch": 19.852832337033874,
+      "grad_norm": 7.254054546356201,
+      "learning_rate": 0.00010551492496884989,
+      "loss": 3.783167266845703,
+      "step": 17450
+    },
+    {
+      "epoch": 19.864218616567037,
+      "grad_norm": 7.683404445648193,
+      "learning_rate": 0.00010546075085324232,
+      "loss": 3.615367889404297,
+      "step": 17460
+    },
+    {
+      "epoch": 19.8756048961002,
+      "grad_norm": 7.9448065757751465,
+      "learning_rate": 0.00010540657673763477,
+      "loss": 3.5793502807617186,
+      "step": 17470
+    },
+    {
+      "epoch": 19.886991175633362,
+      "grad_norm": 7.222136974334717,
+      "learning_rate": 0.00010535240262202719,
+      "loss": 3.8391284942626953,
+      "step": 17480
+    },
+    {
+      "epoch": 19.898377455166525,
+      "grad_norm": 8.161956787109375,
+      "learning_rate": 0.00010529822850641964,
+      "loss": 3.8968780517578123,
+      "step": 17490
+    },
+    {
+      "epoch": 19.909763734699688,
+      "grad_norm": 7.755995750427246,
+      "learning_rate": 0.00010524405439081208,
+      "loss": 3.5014904022216795,
+      "step": 17500
+    },
+    {
+      "epoch": 19.92115001423285,
+      "grad_norm": 7.664623737335205,
+      "learning_rate": 0.00010518988027520451,
+      "loss": 3.861225128173828,
+      "step": 17510
+    },
+    {
+      "epoch": 19.932536293766013,
+      "grad_norm": 8.745182037353516,
+      "learning_rate": 0.00010513570615959696,
+      "loss": 3.8261661529541016,
+      "step": 17520
+    },
+    {
+      "epoch": 19.943922573299176,
+      "grad_norm": 7.4443206787109375,
+      "learning_rate": 0.00010508153204398938,
+      "loss": 3.686481475830078,
+      "step": 17530
+    },
+    {
+      "epoch": 19.95530885283234,
+      "grad_norm": 8.966758728027344,
+      "learning_rate": 0.00010502735792838183,
+      "loss": 3.7222663879394533,
+      "step": 17540
+    },
+    {
+      "epoch": 19.9666951323655,
+      "grad_norm": 9.107354164123535,
+      "learning_rate": 0.00010497318381277426,
+      "loss": 3.9835369110107424,
+      "step": 17550
+    },
+    {
+      "epoch": 19.97808141189866,
+      "grad_norm": 7.190637111663818,
+      "learning_rate": 0.0001049190096971667,
+      "loss": 3.5804458618164063,
+      "step": 17560
+    },
+    {
+      "epoch": 19.989467691431823,
+      "grad_norm": 8.058660507202148,
+      "learning_rate": 0.00010486483558155914,
+      "loss": 3.7318145751953127,
+      "step": 17570
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 5.256486892700195,
+      "learning_rate": 0.00010481066146595157,
+      "loss": 3.4294113159179687,
+      "step": 17580
+    },
+    {
+      "epoch": 20.011386279533163,
+      "grad_norm": 6.450402736663818,
+      "learning_rate": 0.000104756487350344,
+      "loss": 3.1654720306396484,
+      "step": 17590
+    },
+    {
+      "epoch": 20.022772559066325,
+      "grad_norm": 8.31795597076416,
+      "learning_rate": 0.00010470231323473645,
+      "loss": 3.4220176696777345,
+      "step": 17600
+    },
+    {
+      "epoch": 20.034158838599488,
+      "grad_norm": 7.338985443115234,
+      "learning_rate": 0.0001046481391191289,
+      "loss": 3.2845233917236327,
+      "step": 17610
+    },
+    {
+      "epoch": 20.04554511813265,
+      "grad_norm": 7.033785820007324,
+      "learning_rate": 0.00010459396500352133,
+      "loss": 3.3925506591796877,
+      "step": 17620
+    },
+    {
+      "epoch": 20.056931397665814,
+      "grad_norm": 8.403705596923828,
+      "learning_rate": 0.00010453979088791376,
+      "loss": 3.3477813720703127,
+      "step": 17630
+    },
+    {
+      "epoch": 20.068317677198976,
+      "grad_norm": 8.051742553710938,
+      "learning_rate": 0.00010448561677230619,
+      "loss": 3.4361392974853517,
+      "step": 17640
+    },
+    {
+      "epoch": 20.07970395673214,
+      "grad_norm": 8.155322074890137,
+      "learning_rate": 0.00010443144265669864,
+      "loss": 3.4342941284179687,
+      "step": 17650
+    },
+    {
+      "epoch": 20.091090236265302,
+      "grad_norm": 7.207828998565674,
+      "learning_rate": 0.00010437726854109106,
+      "loss": 3.397259521484375,
+      "step": 17660
+    },
+    {
+      "epoch": 20.102476515798465,
+      "grad_norm": 8.539883613586426,
+      "learning_rate": 0.00010432309442548352,
+      "loss": 3.2583744049072267,
+      "step": 17670
+    },
+    {
+      "epoch": 20.113862795331624,
+      "grad_norm": 7.494978904724121,
+      "learning_rate": 0.00010426892030987594,
+      "loss": 3.2995159149169924,
+      "step": 17680
+    },
+    {
+      "epoch": 20.125249074864787,
+      "grad_norm": 8.173662185668945,
+      "learning_rate": 0.00010421474619426839,
+      "loss": 3.5137447357177733,
+      "step": 17690
+    },
+    {
+      "epoch": 20.13663535439795,
+      "grad_norm": 7.584948539733887,
+      "learning_rate": 0.00010416057207866082,
+      "loss": 3.537697601318359,
+      "step": 17700
+    },
+    {
+      "epoch": 20.148021633931112,
+      "grad_norm": 7.191871643066406,
+      "learning_rate": 0.00010410639796305325,
+      "loss": 3.5673038482666017,
+      "step": 17710
+    },
+    {
+      "epoch": 20.159407913464275,
+      "grad_norm": 7.3911356925964355,
+      "learning_rate": 0.00010405222384744568,
+      "loss": 3.557468032836914,
+      "step": 17720
+    },
+    {
+      "epoch": 20.170794192997437,
+      "grad_norm": 6.331779479980469,
+      "learning_rate": 0.00010399804973183813,
+      "loss": 3.5360912322998046,
+      "step": 17730
+    },
+    {
+      "epoch": 20.1821804725306,
+      "grad_norm": 7.614666938781738,
+      "learning_rate": 0.00010394387561623058,
+      "loss": 3.3655654907226564,
+      "step": 17740
+    },
+    {
+      "epoch": 20.193566752063763,
+      "grad_norm": 7.935511112213135,
+      "learning_rate": 0.000103889701500623,
+      "loss": 3.340676498413086,
+      "step": 17750
+    },
+    {
+      "epoch": 20.204953031596926,
+      "grad_norm": 8.085060119628906,
+      "learning_rate": 0.00010383552738501546,
+      "loss": 3.363076400756836,
+      "step": 17760
+    },
+    {
+      "epoch": 20.21633931113009,
+      "grad_norm": 7.689126491546631,
+      "learning_rate": 0.00010378135326940788,
+      "loss": 3.4896373748779297,
+      "step": 17770
+    },
+    {
+      "epoch": 20.22772559066325,
+      "grad_norm": 8.10032844543457,
+      "learning_rate": 0.00010372717915380032,
+      "loss": 3.325590133666992,
+      "step": 17780
+    },
+    {
+      "epoch": 20.239111870196414,
+      "grad_norm": 8.31187629699707,
+      "learning_rate": 0.00010367300503819274,
+      "loss": 3.2680404663085936,
+      "step": 17790
+    },
+    {
+      "epoch": 20.250498149729577,
+      "grad_norm": 7.616562366485596,
+      "learning_rate": 0.0001036188309225852,
+      "loss": 3.3835872650146483,
+      "step": 17800
+    },
+    {
+      "epoch": 20.26188442926274,
+      "grad_norm": 7.7585344314575195,
+      "learning_rate": 0.00010356465680697762,
+      "loss": 3.3168540954589845,
+      "step": 17810
+    },
+    {
+      "epoch": 20.273270708795902,
+      "grad_norm": 7.622613430023193,
+      "learning_rate": 0.00010351048269137007,
+      "loss": 3.4925868988037108,
+      "step": 17820
+    },
+    {
+      "epoch": 20.284656988329065,
+      "grad_norm": 7.538808345794678,
+      "learning_rate": 0.0001034563085757625,
+      "loss": 3.5138877868652343,
+      "step": 17830
+    },
+    {
+      "epoch": 20.296043267862228,
+      "grad_norm": 7.818309783935547,
+      "learning_rate": 0.00010340213446015495,
+      "loss": 3.594626617431641,
+      "step": 17840
+    },
+    {
+      "epoch": 20.30742954739539,
+      "grad_norm": 7.420721530914307,
+      "learning_rate": 0.00010334796034454739,
+      "loss": 3.3362926483154296,
+      "step": 17850
+    },
+    {
+      "epoch": 20.31881582692855,
+      "grad_norm": 8.078766822814941,
+      "learning_rate": 0.00010329378622893981,
+      "loss": 3.376148986816406,
+      "step": 17860
+    },
+    {
+      "epoch": 20.330202106461712,
+      "grad_norm": 8.338269233703613,
+      "learning_rate": 0.00010323961211333226,
+      "loss": 3.398274230957031,
+      "step": 17870
+    },
+    {
+      "epoch": 20.341588385994875,
+      "grad_norm": 8.21082592010498,
+      "learning_rate": 0.00010318543799772469,
+      "loss": 3.543021011352539,
+      "step": 17880
+    },
+    {
+      "epoch": 20.352974665528038,
+      "grad_norm": 7.475930213928223,
+      "learning_rate": 0.00010313126388211714,
+      "loss": 3.3988021850585937,
+      "step": 17890
+    },
+    {
+      "epoch": 20.3643609450612,
+      "grad_norm": 8.236494064331055,
+      "learning_rate": 0.00010307708976650956,
+      "loss": 3.582375717163086,
+      "step": 17900
+    },
+    {
+      "epoch": 20.375747224594363,
+      "grad_norm": 6.666841506958008,
+      "learning_rate": 0.000103022915650902,
+      "loss": 3.476043701171875,
+      "step": 17910
+    },
+    {
+      "epoch": 20.387133504127526,
+      "grad_norm": 8.686334609985352,
+      "learning_rate": 0.00010296874153529444,
+      "loss": 3.4657127380371096,
+      "step": 17920
+    },
+    {
+      "epoch": 20.39851978366069,
+      "grad_norm": 7.781078338623047,
+      "learning_rate": 0.00010291456741968688,
+      "loss": 3.4596488952636717,
+      "step": 17930
+    },
+    {
+      "epoch": 20.40990606319385,
+      "grad_norm": 6.893237590789795,
+      "learning_rate": 0.0001028603933040793,
+      "loss": 3.6022926330566407,
+      "step": 17940
+    },
+    {
+      "epoch": 20.421292342727014,
+      "grad_norm": 8.505749702453613,
+      "learning_rate": 0.00010280621918847175,
+      "loss": 3.511888122558594,
+      "step": 17950
+    },
+    {
+      "epoch": 20.432678622260177,
+      "grad_norm": 7.29093074798584,
+      "learning_rate": 0.0001027520450728642,
+      "loss": 3.803249740600586,
+      "step": 17960
+    },
+    {
+      "epoch": 20.44406490179334,
+      "grad_norm": 7.3153462409973145,
+      "learning_rate": 0.00010269787095725663,
+      "loss": 3.3840301513671873,
+      "step": 17970
+    },
+    {
+      "epoch": 20.455451181326502,
+      "grad_norm": 7.5833306312561035,
+      "learning_rate": 0.00010264369684164907,
+      "loss": 3.48956413269043,
+      "step": 17980
+    },
+    {
+      "epoch": 20.466837460859665,
+      "grad_norm": 8.305935859680176,
+      "learning_rate": 0.00010258952272604149,
+      "loss": 3.399882507324219,
+      "step": 17990
+    },
+    {
+      "epoch": 20.478223740392828,
+      "grad_norm": 8.22060775756836,
+      "learning_rate": 0.00010253534861043394,
+      "loss": 3.2761825561523437,
+      "step": 18000
+    },
+    {
+      "epoch": 20.48961001992599,
+      "grad_norm": 8.05759334564209,
+      "learning_rate": 0.00010248117449482637,
+      "loss": 3.5770782470703124,
+      "step": 18010
+    },
+    {
+      "epoch": 20.500996299459153,
+      "grad_norm": 8.089456558227539,
+      "learning_rate": 0.00010242700037921882,
+      "loss": 3.424999237060547,
+      "step": 18020
+    },
+    {
+      "epoch": 20.512382578992316,
+      "grad_norm": 7.183095932006836,
+      "learning_rate": 0.00010237282626361124,
+      "loss": 3.1551158905029295,
+      "step": 18030
+    },
+    {
+      "epoch": 20.523768858525475,
+      "grad_norm": 7.149391174316406,
+      "learning_rate": 0.0001023186521480037,
+      "loss": 3.384687805175781,
+      "step": 18040
+    },
+    {
+      "epoch": 20.535155138058638,
+      "grad_norm": 6.47637939453125,
+      "learning_rate": 0.00010226447803239612,
+      "loss": 3.601530075073242,
+      "step": 18050
+    },
+    {
+      "epoch": 20.5465414175918,
+      "grad_norm": 8.601676940917969,
+      "learning_rate": 0.00010221030391678856,
+      "loss": 3.6373306274414063,
+      "step": 18060
+    },
+    {
+      "epoch": 20.557927697124963,
+      "grad_norm": 8.294757843017578,
+      "learning_rate": 0.00010215612980118101,
+      "loss": 3.5030460357666016,
+      "step": 18070
+    },
+    {
+      "epoch": 20.569313976658126,
+      "grad_norm": 7.412970066070557,
+      "learning_rate": 0.00010210195568557343,
+      "loss": 3.206818389892578,
+      "step": 18080
+    },
+    {
+      "epoch": 20.58070025619129,
+      "grad_norm": 7.663300514221191,
+      "learning_rate": 0.00010204778156996589,
+      "loss": 3.5179752349853515,
+      "step": 18090
+    },
+    {
+      "epoch": 20.59208653572445,
+      "grad_norm": 7.107182025909424,
+      "learning_rate": 0.00010199360745435831,
+      "loss": 3.446649932861328,
+      "step": 18100
+    },
+    {
+      "epoch": 20.603472815257614,
+      "grad_norm": 7.1297688484191895,
+      "learning_rate": 0.00010193943333875076,
+      "loss": 3.4354694366455076,
+      "step": 18110
+    },
+    {
+      "epoch": 20.614859094790777,
+      "grad_norm": 7.961850643157959,
+      "learning_rate": 0.00010188525922314319,
+      "loss": 3.3658203125,
+      "step": 18120
+    },
+    {
+      "epoch": 20.62624537432394,
+      "grad_norm": 7.603863716125488,
+      "learning_rate": 0.00010183108510753562,
+      "loss": 3.439925765991211,
+      "step": 18130
+    },
+    {
+      "epoch": 20.637631653857103,
+      "grad_norm": 7.988809108734131,
+      "learning_rate": 0.00010177691099192805,
+      "loss": 3.531351089477539,
+      "step": 18140
+    },
+    {
+      "epoch": 20.649017933390265,
+      "grad_norm": 8.041953086853027,
+      "learning_rate": 0.0001017227368763205,
+      "loss": 3.673780822753906,
+      "step": 18150
+    },
+    {
+      "epoch": 20.660404212923428,
+      "grad_norm": 8.37690258026123,
+      "learning_rate": 0.00010166856276071292,
+      "loss": 3.794148254394531,
+      "step": 18160
+    },
+    {
+      "epoch": 20.67179049245659,
+      "grad_norm": 8.250043869018555,
+      "learning_rate": 0.00010161438864510538,
+      "loss": 3.5557697296142576,
+      "step": 18170
+    },
+    {
+      "epoch": 20.683176771989753,
+      "grad_norm": 8.75919246673584,
+      "learning_rate": 0.00010156021452949783,
+      "loss": 3.7775032043457033,
+      "step": 18180
+    },
+    {
+      "epoch": 20.694563051522916,
+      "grad_norm": 8.309146881103516,
+      "learning_rate": 0.00010150604041389025,
+      "loss": 3.4888229370117188,
+      "step": 18190
+    },
+    {
+      "epoch": 20.70594933105608,
+      "grad_norm": 8.30295467376709,
+      "learning_rate": 0.00010145186629828269,
+      "loss": 3.7718467712402344,
+      "step": 18200
+    },
+    {
+      "epoch": 20.71733561058924,
+      "grad_norm": 7.974371910095215,
+      "learning_rate": 0.00010139769218267511,
+      "loss": 3.401205062866211,
+      "step": 18210
+    },
+    {
+      "epoch": 20.7287218901224,
+      "grad_norm": 8.394783973693848,
+      "learning_rate": 0.00010134351806706757,
+      "loss": 3.713974380493164,
+      "step": 18220
+    },
+    {
+      "epoch": 20.740108169655564,
+      "grad_norm": 8.151832580566406,
+      "learning_rate": 0.00010128934395145999,
+      "loss": 3.8104171752929688,
+      "step": 18230
+    },
+    {
+      "epoch": 20.751494449188726,
+      "grad_norm": 7.931938171386719,
+      "learning_rate": 0.00010123516983585244,
+      "loss": 3.417683410644531,
+      "step": 18240
+    },
+    {
+      "epoch": 20.76288072872189,
+      "grad_norm": 7.705905914306641,
+      "learning_rate": 0.00010118099572024487,
+      "loss": 3.6185550689697266,
+      "step": 18250
+    },
+    {
+      "epoch": 20.774267008255052,
+      "grad_norm": 9.284891128540039,
+      "learning_rate": 0.00010112682160463732,
+      "loss": 3.7930877685546873,
+      "step": 18260
+    },
+    {
+      "epoch": 20.785653287788215,
+      "grad_norm": 7.982856750488281,
+      "learning_rate": 0.00010107264748902974,
+      "loss": 3.4403953552246094,
+      "step": 18270
+    },
+    {
+      "epoch": 20.797039567321377,
+      "grad_norm": 8.39303207397461,
+      "learning_rate": 0.00010101847337342218,
+      "loss": 3.756923294067383,
+      "step": 18280
+    },
+    {
+      "epoch": 20.80842584685454,
+      "grad_norm": 7.579355239868164,
+      "learning_rate": 0.00010096429925781463,
+      "loss": 3.770177459716797,
+      "step": 18290
+    },
+    {
+      "epoch": 20.819812126387703,
+      "grad_norm": 7.803186416625977,
+      "learning_rate": 0.00010091012514220706,
+      "loss": 3.6111522674560548,
+      "step": 18300
+    },
+    {
+      "epoch": 20.831198405920865,
+      "grad_norm": 7.868539810180664,
+      "learning_rate": 0.00010085595102659951,
+      "loss": 3.5871612548828127,
+      "step": 18310
+    },
+    {
+      "epoch": 20.842584685454028,
+      "grad_norm": 8.147734642028809,
+      "learning_rate": 0.00010080177691099193,
+      "loss": 3.5940319061279298,
+      "step": 18320
+    },
+    {
+      "epoch": 20.85397096498719,
+      "grad_norm": 7.550434589385986,
+      "learning_rate": 0.00010074760279538437,
+      "loss": 3.7070236206054688,
+      "step": 18330
+    },
+    {
+      "epoch": 20.865357244520354,
+      "grad_norm": 7.593775272369385,
+      "learning_rate": 0.00010069342867977681,
+      "loss": 3.5908023834228517,
+      "step": 18340
+    },
+    {
+      "epoch": 20.876743524053516,
+      "grad_norm": 9.302807807922363,
+      "learning_rate": 0.00010063925456416925,
+      "loss": 3.5747802734375,
+      "step": 18350
+    },
+    {
+      "epoch": 20.88812980358668,
+      "grad_norm": 8.078301429748535,
+      "learning_rate": 0.00010058508044856167,
+      "loss": 3.404593658447266,
+      "step": 18360
+    },
+    {
+      "epoch": 20.899516083119842,
+      "grad_norm": 7.820472240447998,
+      "learning_rate": 0.00010053090633295412,
+      "loss": 3.6110614776611327,
+      "step": 18370
+    },
+    {
+      "epoch": 20.910902362653005,
+      "grad_norm": 7.314359664916992,
+      "learning_rate": 0.00010047673221734655,
+      "loss": 3.4265087127685545,
+      "step": 18380
+    },
+    {
+      "epoch": 20.922288642186167,
+      "grad_norm": 7.681244373321533,
+      "learning_rate": 0.000100422558101739,
+      "loss": 3.6056175231933594,
+      "step": 18390
+    },
+    {
+      "epoch": 20.933674921719327,
+      "grad_norm": 8.012112617492676,
+      "learning_rate": 0.00010036838398613142,
+      "loss": 3.682481384277344,
+      "step": 18400
+    },
+    {
+      "epoch": 20.94506120125249,
+      "grad_norm": 7.457361221313477,
+      "learning_rate": 0.00010031420987052386,
+      "loss": 3.843639373779297,
+      "step": 18410
+    },
+    {
+      "epoch": 20.956447480785652,
+      "grad_norm": 8.518708229064941,
+      "learning_rate": 0.00010026003575491631,
+      "loss": 3.5751590728759766,
+      "step": 18420
+    },
+    {
+      "epoch": 20.967833760318815,
+      "grad_norm": 7.371086597442627,
+      "learning_rate": 0.00010020586163930874,
+      "loss": 3.6571311950683594,
+      "step": 18430
+    },
+    {
+      "epoch": 20.979220039851977,
+      "grad_norm": 7.497313976287842,
+      "learning_rate": 0.00010015168752370119,
+      "loss": 3.5427356719970704,
+      "step": 18440
+    },
+    {
+      "epoch": 20.99060631938514,
+      "grad_norm": 8.249246597290039,
+      "learning_rate": 0.00010009751340809361,
+      "loss": 3.492108154296875,
+      "step": 18450
+    },
+    {
+      "epoch": 21.001138627953317,
+      "grad_norm": 7.12626314163208,
+      "learning_rate": 0.00010004333929248607,
+      "loss": 3.4620594024658202,
+      "step": 18460
+    },
+    {
+      "epoch": 21.01252490748648,
+      "grad_norm": 7.223207473754883,
+      "learning_rate": 9.99891651768785e-05,
+      "loss": 2.993937301635742,
+      "step": 18470
+    },
+    {
+      "epoch": 21.023911187019642,
+      "grad_norm": 8.60847282409668,
+      "learning_rate": 9.993499106127093e-05,
+      "loss": 3.044596481323242,
+      "step": 18480
+    },
+    {
+      "epoch": 21.035297466552805,
+      "grad_norm": 7.056385040283203,
+      "learning_rate": 9.988081694566337e-05,
+      "loss": 3.329801559448242,
+      "step": 18490
+    },
+    {
+      "epoch": 21.046683746085968,
+      "grad_norm": 8.232537269592285,
+      "learning_rate": 9.98266428300558e-05,
+      "loss": 3.359906005859375,
+      "step": 18500
+    },
+    {
+      "epoch": 21.05807002561913,
+      "grad_norm": 6.871865272521973,
+      "learning_rate": 9.977246871444824e-05,
+      "loss": 3.1155139923095705,
+      "step": 18510
+    },
+    {
+      "epoch": 21.06945630515229,
+      "grad_norm": 8.1388578414917,
+      "learning_rate": 9.971829459884068e-05,
+      "loss": 3.1933937072753906,
+      "step": 18520
+    },
+    {
+      "epoch": 21.080842584685453,
+      "grad_norm": 7.495057106018066,
+      "learning_rate": 9.966412048323312e-05,
+      "loss": 3.4371055603027343,
+      "step": 18530
+    },
+    {
+      "epoch": 21.092228864218615,
+      "grad_norm": 6.833533763885498,
+      "learning_rate": 9.960994636762556e-05,
+      "loss": 3.0802495956420897,
+      "step": 18540
+    },
+    {
+      "epoch": 21.103615143751778,
+      "grad_norm": 6.975308895111084,
+      "learning_rate": 9.955577225201798e-05,
+      "loss": 3.318703842163086,
+      "step": 18550
+    },
+    {
+      "epoch": 21.11500142328494,
+      "grad_norm": 8.011495590209961,
+      "learning_rate": 9.950159813641042e-05,
+      "loss": 3.2149578094482423,
+      "step": 18560
+    },
+    {
+      "epoch": 21.126387702818104,
+      "grad_norm": 7.448869705200195,
+      "learning_rate": 9.944742402080287e-05,
+      "loss": 3.2515499114990236,
+      "step": 18570
+    },
+    {
+      "epoch": 21.137773982351266,
+      "grad_norm": 7.280409812927246,
+      "learning_rate": 9.939324990519531e-05,
+      "loss": 3.5192329406738283,
+      "step": 18580
+    },
+    {
+      "epoch": 21.14916026188443,
+      "grad_norm": 7.909364700317383,
+      "learning_rate": 9.933907578958775e-05,
+      "loss": 3.397074890136719,
+      "step": 18590
+    },
+    {
+      "epoch": 21.16054654141759,
+      "grad_norm": 7.611769676208496,
+      "learning_rate": 9.928490167398018e-05,
+      "loss": 3.0764091491699217,
+      "step": 18600
+    },
+    {
+      "epoch": 21.171932820950754,
+      "grad_norm": 7.476166248321533,
+      "learning_rate": 9.923072755837262e-05,
+      "loss": 3.295338821411133,
+      "step": 18610
+    },
+    {
+      "epoch": 21.183319100483917,
+      "grad_norm": 6.955434799194336,
+      "learning_rate": 9.917655344276505e-05,
+      "loss": 3.357422637939453,
+      "step": 18620
+    },
+    {
+      "epoch": 21.19470538001708,
+      "grad_norm": 8.147130966186523,
+      "learning_rate": 9.912237932715748e-05,
+      "loss": 3.1608652114868163,
+      "step": 18630
+    },
+    {
+      "epoch": 21.206091659550243,
+      "grad_norm": 8.146050453186035,
+      "learning_rate": 9.906820521154992e-05,
+      "loss": 3.1724903106689455,
+      "step": 18640
+    },
+    {
+      "epoch": 21.217477939083405,
+      "grad_norm": 7.68185567855835,
+      "learning_rate": 9.901403109594236e-05,
+      "loss": 3.063700485229492,
+      "step": 18650
+    },
+    {
+      "epoch": 21.228864218616568,
+      "grad_norm": 7.399899005889893,
+      "learning_rate": 9.89598569803348e-05,
+      "loss": 3.2527538299560548,
+      "step": 18660
+    },
+    {
+      "epoch": 21.24025049814973,
+      "grad_norm": 8.725862503051758,
+      "learning_rate": 9.890568286472724e-05,
+      "loss": 3.3397960662841797,
+      "step": 18670
+    },
+    {
+      "epoch": 21.251636777682894,
+      "grad_norm": 8.439692497253418,
+      "learning_rate": 9.885150874911967e-05,
+      "loss": 3.127163314819336,
+      "step": 18680
+    },
+    {
+      "epoch": 21.263023057216056,
+      "grad_norm": 7.275774955749512,
+      "learning_rate": 9.879733463351211e-05,
+      "loss": 3.3399112701416014,
+      "step": 18690
+    },
+    {
+      "epoch": 21.274409336749216,
+      "grad_norm": 8.231805801391602,
+      "learning_rate": 9.874316051790455e-05,
+      "loss": 3.3167404174804687,
+      "step": 18700
+    },
+    {
+      "epoch": 21.28579561628238,
+      "grad_norm": 8.807353019714355,
+      "learning_rate": 9.868898640229699e-05,
+      "loss": 3.2518096923828126,
+      "step": 18710
+    },
+    {
+      "epoch": 21.29718189581554,
+      "grad_norm": 7.745279788970947,
+      "learning_rate": 9.863481228668943e-05,
+      "loss": 2.9932077407836912,
+      "step": 18720
+    },
+    {
+      "epoch": 21.308568175348704,
+      "grad_norm": 8.119501113891602,
+      "learning_rate": 9.858063817108186e-05,
+      "loss": 3.360261917114258,
+      "step": 18730
+    },
+    {
+      "epoch": 21.319954454881866,
+      "grad_norm": 7.655581951141357,
+      "learning_rate": 9.85264640554743e-05,
+      "loss": 3.311513900756836,
+      "step": 18740
+    },
+    {
+      "epoch": 21.33134073441503,
+      "grad_norm": 7.7340240478515625,
+      "learning_rate": 9.847228993986674e-05,
+      "loss": 3.241596221923828,
+      "step": 18750
+    },
+    {
+      "epoch": 21.342727013948192,
+      "grad_norm": 7.133695602416992,
+      "learning_rate": 9.841811582425917e-05,
+      "loss": 3.315245819091797,
+      "step": 18760
+    },
+    {
+      "epoch": 21.354113293481355,
+      "grad_norm": 7.775753021240234,
+      "learning_rate": 9.83639417086516e-05,
+      "loss": 3.2109920501708986,
+      "step": 18770
+    },
+    {
+      "epoch": 21.365499573014517,
+      "grad_norm": 7.912374019622803,
+      "learning_rate": 9.830976759304404e-05,
+      "loss": 3.256867218017578,
+      "step": 18780
+    },
+    {
+      "epoch": 21.37688585254768,
+      "grad_norm": 7.180131435394287,
+      "learning_rate": 9.825559347743648e-05,
+      "loss": 3.2446563720703123,
+      "step": 18790
+    },
+    {
+      "epoch": 21.388272132080843,
+      "grad_norm": 8.159407615661621,
+      "learning_rate": 9.820141936182893e-05,
+      "loss": 3.262491226196289,
+      "step": 18800
+    },
+    {
+      "epoch": 21.399658411614006,
+      "grad_norm": 7.895713806152344,
+      "learning_rate": 9.814724524622137e-05,
+      "loss": 3.2164894104003907,
+      "step": 18810
+    },
+    {
+      "epoch": 21.41104469114717,
+      "grad_norm": 8.5209321975708,
+      "learning_rate": 9.809307113061381e-05,
+      "loss": 3.3610679626464846,
+      "step": 18820
+    },
+    {
+      "epoch": 21.42243097068033,
+      "grad_norm": 6.682577610015869,
+      "learning_rate": 9.803889701500623e-05,
+      "loss": 3.2081188201904296,
+      "step": 18830
+    },
+    {
+      "epoch": 21.433817250213494,
+      "grad_norm": 7.8088812828063965,
+      "learning_rate": 9.798472289939867e-05,
+      "loss": 3.5212165832519533,
+      "step": 18840
+    },
+    {
+      "epoch": 21.445203529746657,
+      "grad_norm": 7.557262897491455,
+      "learning_rate": 9.793054878379111e-05,
+      "loss": 3.161390495300293,
+      "step": 18850
+    },
+    {
+      "epoch": 21.45658980927982,
+      "grad_norm": 6.671349048614502,
+      "learning_rate": 9.787637466818355e-05,
+      "loss": 3.4181884765625,
+      "step": 18860
+    },
+    {
+      "epoch": 21.467976088812982,
+      "grad_norm": 8.085090637207031,
+      "learning_rate": 9.782220055257598e-05,
+      "loss": 3.2536998748779298,
+      "step": 18870
+    },
+    {
+      "epoch": 21.47936236834614,
+      "grad_norm": 7.863136291503906,
+      "learning_rate": 9.776802643696842e-05,
+      "loss": 3.254865264892578,
+      "step": 18880
+    },
+    {
+      "epoch": 21.490748647879304,
+      "grad_norm": 8.205495834350586,
+      "learning_rate": 9.771385232136086e-05,
+      "loss": 3.396427536010742,
+      "step": 18890
+    },
+    {
+      "epoch": 21.502134927412467,
+      "grad_norm": 7.5775861740112305,
+      "learning_rate": 9.76596782057533e-05,
+      "loss": 3.1754524230957033,
+      "step": 18900
+    },
+    {
+      "epoch": 21.51352120694563,
+      "grad_norm": 7.762941837310791,
+      "learning_rate": 9.760550409014574e-05,
+      "loss": 3.3684154510498048,
+      "step": 18910
+    },
+    {
+      "epoch": 21.524907486478792,
+      "grad_norm": 7.523013114929199,
+      "learning_rate": 9.755132997453817e-05,
+      "loss": 3.414249801635742,
+      "step": 18920
+    },
+    {
+      "epoch": 21.536293766011955,
+      "grad_norm": 8.015236854553223,
+      "learning_rate": 9.749715585893061e-05,
+      "loss": 3.2509552001953126,
+      "step": 18930
+    },
+    {
+      "epoch": 21.547680045545118,
+      "grad_norm": 9.543829917907715,
+      "learning_rate": 9.744298174332305e-05,
+      "loss": 3.6638416290283202,
+      "step": 18940
+    },
+    {
+      "epoch": 21.55906632507828,
+      "grad_norm": 8.284276008605957,
+      "learning_rate": 9.738880762771549e-05,
+      "loss": 3.4558319091796874,
+      "step": 18950
+    },
+    {
+      "epoch": 21.570452604611443,
+      "grad_norm": 8.424079895019531,
+      "learning_rate": 9.733463351210793e-05,
+      "loss": 3.319378662109375,
+      "step": 18960
+    },
+    {
+      "epoch": 21.581838884144606,
+      "grad_norm": 7.417494773864746,
+      "learning_rate": 9.728045939650035e-05,
+      "loss": 3.0912622451782226,
+      "step": 18970
+    },
+    {
+      "epoch": 21.59322516367777,
+      "grad_norm": 8.443395614624023,
+      "learning_rate": 9.722628528089279e-05,
+      "loss": 3.599252700805664,
+      "step": 18980
+    },
+    {
+      "epoch": 21.60461144321093,
+      "grad_norm": 7.053211688995361,
+      "learning_rate": 9.717211116528523e-05,
+      "loss": 3.4184902191162108,
+      "step": 18990
+    },
+    {
+      "epoch": 21.615997722744094,
+      "grad_norm": 7.322805881500244,
+      "learning_rate": 9.711793704967766e-05,
+      "loss": 3.3825252532958983,
+      "step": 19000
+    },
+    {
+      "epoch": 21.627384002277257,
+      "grad_norm": 7.776912212371826,
+      "learning_rate": 9.70637629340701e-05,
+      "loss": 3.413040542602539,
+      "step": 19010
+    },
+    {
+      "epoch": 21.63877028181042,
+      "grad_norm": 7.955855846405029,
+      "learning_rate": 9.700958881846254e-05,
+      "loss": 3.3618423461914064,
+      "step": 19020
+    },
+    {
+      "epoch": 21.650156561343582,
+      "grad_norm": 7.364816188812256,
+      "learning_rate": 9.695541470285499e-05,
+      "loss": 3.470820999145508,
+      "step": 19030
+    },
+    {
+      "epoch": 21.661542840876745,
+      "grad_norm": 7.257011413574219,
+      "learning_rate": 9.690124058724742e-05,
+      "loss": 3.3685569763183594,
+      "step": 19040
+    },
+    {
+      "epoch": 21.672929120409908,
+      "grad_norm": 7.3103485107421875,
+      "learning_rate": 9.684706647163985e-05,
+      "loss": 3.5275413513183596,
+      "step": 19050
+    },
+    {
+      "epoch": 21.684315399943067,
+      "grad_norm": 8.958032608032227,
+      "learning_rate": 9.679289235603229e-05,
+      "loss": 3.6240314483642577,
+      "step": 19060
+    },
+    {
+      "epoch": 21.69570167947623,
+      "grad_norm": 7.291479110717773,
+      "learning_rate": 9.673871824042473e-05,
+      "loss": 3.4977294921875,
+      "step": 19070
+    },
+    {
+      "epoch": 21.707087959009392,
+      "grad_norm": 7.926621437072754,
+      "learning_rate": 9.668454412481717e-05,
+      "loss": 3.3956787109375,
+      "step": 19080
+    },
+    {
+      "epoch": 21.718474238542555,
+      "grad_norm": 7.5746588706970215,
+      "learning_rate": 9.66303700092096e-05,
+      "loss": 3.4186546325683596,
+      "step": 19090
+    },
+    {
+      "epoch": 21.729860518075718,
+      "grad_norm": 8.237885475158691,
+      "learning_rate": 9.657619589360204e-05,
+      "loss": 3.315296173095703,
+      "step": 19100
+    },
+    {
+      "epoch": 21.74124679760888,
+      "grad_norm": 7.103343963623047,
+      "learning_rate": 9.652202177799447e-05,
+      "loss": 3.2063358306884764,
+      "step": 19110
+    },
+    {
+      "epoch": 21.752633077142043,
+      "grad_norm": 7.7268500328063965,
+      "learning_rate": 9.646784766238691e-05,
+      "loss": 3.576313781738281,
+      "step": 19120
+    },
+    {
+      "epoch": 21.764019356675206,
+      "grad_norm": 8.264037132263184,
+      "learning_rate": 9.641367354677935e-05,
+      "loss": 3.452584075927734,
+      "step": 19130
+    },
+    {
+      "epoch": 21.77540563620837,
+      "grad_norm": 7.947970390319824,
+      "learning_rate": 9.63594994311718e-05,
+      "loss": 3.2328662872314453,
+      "step": 19140
+    },
+    {
+      "epoch": 21.78679191574153,
+      "grad_norm": 8.009827613830566,
+      "learning_rate": 9.630532531556423e-05,
+      "loss": 3.3568878173828125,
+      "step": 19150
+    },
+    {
+      "epoch": 21.798178195274694,
+      "grad_norm": 7.520967483520508,
+      "learning_rate": 9.625115119995667e-05,
+      "loss": 3.2189865112304688,
+      "step": 19160
+    },
+    {
+      "epoch": 21.809564474807857,
+      "grad_norm": 6.894745349884033,
+      "learning_rate": 9.619697708434911e-05,
+      "loss": 3.3676876068115233,
+      "step": 19170
+    },
+    {
+      "epoch": 21.82095075434102,
+      "grad_norm": 7.391615867614746,
+      "learning_rate": 9.614280296874154e-05,
+      "loss": 3.1733936309814452,
+      "step": 19180
+    },
+    {
+      "epoch": 21.832337033874182,
+      "grad_norm": 8.549541473388672,
+      "learning_rate": 9.608862885313397e-05,
+      "loss": 3.4626365661621095,
+      "step": 19190
+    },
+    {
+      "epoch": 21.843723313407345,
+      "grad_norm": 7.7757182121276855,
+      "learning_rate": 9.603445473752641e-05,
+      "loss": 3.3878768920898437,
+      "step": 19200
+    },
+    {
+      "epoch": 21.855109592940508,
+      "grad_norm": 9.261902809143066,
+      "learning_rate": 9.598028062191885e-05,
+      "loss": 3.0820999145507812,
+      "step": 19210
+    },
+    {
+      "epoch": 21.86649587247367,
+      "grad_norm": 7.974034309387207,
+      "learning_rate": 9.592610650631129e-05,
+      "loss": 3.532638931274414,
+      "step": 19220
+    },
+    {
+      "epoch": 21.877882152006833,
+      "grad_norm": 7.275153160095215,
+      "learning_rate": 9.587193239070373e-05,
+      "loss": 3.409342956542969,
+      "step": 19230
+    },
+    {
+      "epoch": 21.889268431539996,
+      "grad_norm": 7.821238040924072,
+      "learning_rate": 9.581775827509616e-05,
+      "loss": 3.2864406585693358,
+      "step": 19240
+    },
+    {
+      "epoch": 21.900654711073155,
+      "grad_norm": 6.836928844451904,
+      "learning_rate": 9.57635841594886e-05,
+      "loss": 3.399283218383789,
+      "step": 19250
+    },
+    {
+      "epoch": 21.912040990606318,
+      "grad_norm": 8.900579452514648,
+      "learning_rate": 9.570941004388104e-05,
+      "loss": 3.277374267578125,
+      "step": 19260
+    },
+    {
+      "epoch": 21.92342727013948,
+      "grad_norm": 7.689421653747559,
+      "learning_rate": 9.565523592827348e-05,
+      "loss": 3.502179718017578,
+      "step": 19270
+    },
+    {
+      "epoch": 21.934813549672644,
+      "grad_norm": 7.474170684814453,
+      "learning_rate": 9.560106181266592e-05,
+      "loss": 3.3328819274902344,
+      "step": 19280
+    },
+    {
+      "epoch": 21.946199829205806,
+      "grad_norm": 7.40029764175415,
+      "learning_rate": 9.554688769705835e-05,
+      "loss": 3.339038848876953,
+      "step": 19290
+    },
+    {
+      "epoch": 21.95758610873897,
+      "grad_norm": 8.339094161987305,
+      "learning_rate": 9.549271358145079e-05,
+      "loss": 3.642498779296875,
+      "step": 19300
+    },
+    {
+      "epoch": 21.96897238827213,
+      "grad_norm": 8.725251197814941,
+      "learning_rate": 9.543853946584323e-05,
+      "loss": 3.3834274291992186,
+      "step": 19310
+    },
+    {
+      "epoch": 21.980358667805294,
+      "grad_norm": 7.735589504241943,
+      "learning_rate": 9.538436535023565e-05,
+      "loss": 3.362496185302734,
+      "step": 19320
+    },
+    {
+      "epoch": 21.991744947338457,
+      "grad_norm": 8.0684175491333,
+      "learning_rate": 9.533019123462809e-05,
+      "loss": 3.6285259246826174,
+      "step": 19330
+    },
+    {
+      "epoch": 22.002277255906634,
+      "grad_norm": 7.213688850402832,
+      "learning_rate": 9.527601711902053e-05,
+      "loss": 3.0752532958984373,
+      "step": 19340
+    },
+    {
+      "epoch": 22.013663535439797,
+      "grad_norm": 7.230661392211914,
+      "learning_rate": 9.522184300341297e-05,
+      "loss": 3.045562171936035,
+      "step": 19350
+    },
+    {
+      "epoch": 22.025049814972956,
+      "grad_norm": 7.4309587478637695,
+      "learning_rate": 9.51676688878054e-05,
+      "loss": 3.1042001724243162,
+      "step": 19360
+    },
+    {
+      "epoch": 22.03643609450612,
+      "grad_norm": 7.176165580749512,
+      "learning_rate": 9.511349477219786e-05,
+      "loss": 2.9163785934448243,
+      "step": 19370
+    },
+    {
+      "epoch": 22.04782237403928,
+      "grad_norm": 7.511384963989258,
+      "learning_rate": 9.50593206565903e-05,
+      "loss": 2.967784118652344,
+      "step": 19380
+    },
+    {
+      "epoch": 22.059208653572444,
+      "grad_norm": 7.659293174743652,
+      "learning_rate": 9.500514654098272e-05,
+      "loss": 3.1063777923583986,
+      "step": 19390
+    },
+    {
+      "epoch": 22.070594933105607,
+      "grad_norm": 7.430224895477295,
+      "learning_rate": 9.495097242537516e-05,
+      "loss": 3.028996467590332,
+      "step": 19400
+    },
+    {
+      "epoch": 22.08198121263877,
+      "grad_norm": 7.423468589782715,
+      "learning_rate": 9.48967983097676e-05,
+      "loss": 3.157891273498535,
+      "step": 19410
+    },
+    {
+      "epoch": 22.093367492171932,
+      "grad_norm": 7.423767566680908,
+      "learning_rate": 9.484262419416003e-05,
+      "loss": 2.9452770233154295,
+      "step": 19420
+    },
+    {
+      "epoch": 22.104753771705095,
+      "grad_norm": 7.975661754608154,
+      "learning_rate": 9.478845007855247e-05,
+      "loss": 3.0032615661621094,
+      "step": 19430
+    },
+    {
+      "epoch": 22.116140051238258,
+      "grad_norm": 7.372896194458008,
+      "learning_rate": 9.473427596294491e-05,
+      "loss": 3.0240869522094727,
+      "step": 19440
+    },
+    {
+      "epoch": 22.12752633077142,
+      "grad_norm": 8.073140144348145,
+      "learning_rate": 9.468010184733735e-05,
+      "loss": 3.047717475891113,
+      "step": 19450
+    },
+    {
+      "epoch": 22.138912610304583,
+      "grad_norm": 7.785470008850098,
+      "learning_rate": 9.462592773172979e-05,
+      "loss": 3.0755859375,
+      "step": 19460
+    },
+    {
+      "epoch": 22.150298889837746,
+      "grad_norm": 8.51516342163086,
+      "learning_rate": 9.457175361612221e-05,
+      "loss": 3.0527263641357423,
+      "step": 19470
+    },
+    {
+      "epoch": 22.16168516937091,
+      "grad_norm": 7.476900577545166,
+      "learning_rate": 9.451757950051466e-05,
+      "loss": 3.286362075805664,
+      "step": 19480
+    },
+    {
+      "epoch": 22.17307144890407,
+      "grad_norm": 9.160499572753906,
+      "learning_rate": 9.44634053849071e-05,
+      "loss": 3.2243637084960937,
+      "step": 19490
+    },
+    {
+      "epoch": 22.184457728437234,
+      "grad_norm": 7.854064464569092,
+      "learning_rate": 9.440923126929954e-05,
+      "loss": 2.984610748291016,
+      "step": 19500
+    },
+    {
+      "epoch": 22.195844007970397,
+      "grad_norm": 7.314716815948486,
+      "learning_rate": 9.435505715369198e-05,
+      "loss": 3.109171676635742,
+      "step": 19510
+    },
+    {
+      "epoch": 22.20723028750356,
+      "grad_norm": 6.959289073944092,
+      "learning_rate": 9.430088303808441e-05,
+      "loss": 3.1544565200805663,
+      "step": 19520
+    },
+    {
+      "epoch": 22.218616567036722,
+      "grad_norm": 7.609602928161621,
+      "learning_rate": 9.424670892247684e-05,
+      "loss": 3.071408271789551,
+      "step": 19530
+    },
+    {
+      "epoch": 22.23000284656988,
+      "grad_norm": 7.927331924438477,
+      "learning_rate": 9.419253480686928e-05,
+      "loss": 3.055502510070801,
+      "step": 19540
+    },
+    {
+      "epoch": 22.241389126103044,
+      "grad_norm": 8.021471977233887,
+      "learning_rate": 9.413836069126171e-05,
+      "loss": 2.963591194152832,
+      "step": 19550
+    },
+    {
+      "epoch": 22.252775405636207,
+      "grad_norm": 7.8159260749816895,
+      "learning_rate": 9.408418657565415e-05,
+      "loss": 3.1434553146362303,
+      "step": 19560
+    },
+    {
+      "epoch": 22.26416168516937,
+      "grad_norm": 7.591141223907471,
+      "learning_rate": 9.403001246004659e-05,
+      "loss": 3.1236942291259764,
+      "step": 19570
+    },
+    {
+      "epoch": 22.275547964702533,
+      "grad_norm": 7.809198379516602,
+      "learning_rate": 9.397583834443903e-05,
+      "loss": 2.9383840560913086,
+      "step": 19580
+    },
+    {
+      "epoch": 22.286934244235695,
+      "grad_norm": 7.057039737701416,
+      "learning_rate": 9.392166422883148e-05,
+      "loss": 2.704611396789551,
+      "step": 19590
+    },
+    {
+      "epoch": 22.298320523768858,
+      "grad_norm": 8.17033863067627,
+      "learning_rate": 9.38674901132239e-05,
+      "loss": 3.0170907974243164,
+      "step": 19600
+    },
+    {
+      "epoch": 22.30970680330202,
+      "grad_norm": 8.437378883361816,
+      "learning_rate": 9.381331599761634e-05,
+      "loss": 3.192144012451172,
+      "step": 19610
+    },
+    {
+      "epoch": 22.321093082835183,
+      "grad_norm": 8.427281379699707,
+      "learning_rate": 9.375914188200878e-05,
+      "loss": 3.265237808227539,
+      "step": 19620
+    },
+    {
+      "epoch": 22.332479362368346,
+      "grad_norm": 7.595215797424316,
+      "learning_rate": 9.370496776640122e-05,
+      "loss": 3.1780752182006835,
+      "step": 19630
+    },
+    {
+      "epoch": 22.34386564190151,
+      "grad_norm": 7.446441173553467,
+      "learning_rate": 9.365079365079366e-05,
+      "loss": 3.126137542724609,
+      "step": 19640
+    },
+    {
+      "epoch": 22.35525192143467,
+      "grad_norm": 8.297301292419434,
+      "learning_rate": 9.35966195351861e-05,
+      "loss": 3.281117630004883,
+      "step": 19650
+    },
+    {
+      "epoch": 22.366638200967834,
+      "grad_norm": 7.048451900482178,
+      "learning_rate": 9.354244541957853e-05,
+      "loss": 3.1563640594482423,
+      "step": 19660
+    },
+    {
+      "epoch": 22.378024480500997,
+      "grad_norm": 7.006303787231445,
+      "learning_rate": 9.348827130397097e-05,
+      "loss": 2.991481971740723,
+      "step": 19670
+    },
+    {
+      "epoch": 22.38941076003416,
+      "grad_norm": 6.830575942993164,
+      "learning_rate": 9.34340971883634e-05,
+      "loss": 3.372071075439453,
+      "step": 19680
+    },
+    {
+      "epoch": 22.400797039567323,
+      "grad_norm": 8.474934577941895,
+      "learning_rate": 9.337992307275583e-05,
+      "loss": 3.1545629501342773,
+      "step": 19690
+    },
+    {
+      "epoch": 22.412183319100485,
+      "grad_norm": 8.20909595489502,
+      "learning_rate": 9.332574895714827e-05,
+      "loss": 3.1375335693359374,
+      "step": 19700
+    },
+    {
+      "epoch": 22.423569598633648,
+      "grad_norm": 7.151069164276123,
+      "learning_rate": 9.327157484154072e-05,
+      "loss": 3.147255706787109,
+      "step": 19710
+    },
+    {
+      "epoch": 22.434955878166807,
+      "grad_norm": 8.133870124816895,
+      "learning_rate": 9.321740072593316e-05,
+      "loss": 3.096915435791016,
+      "step": 19720
+    },
+    {
+      "epoch": 22.44634215769997,
+      "grad_norm": 7.726839065551758,
+      "learning_rate": 9.31632266103256e-05,
+      "loss": 3.2699676513671876,
+      "step": 19730
+    },
+    {
+      "epoch": 22.457728437233133,
+      "grad_norm": 7.36300802230835,
+      "learning_rate": 9.310905249471802e-05,
+      "loss": 3.1540500640869142,
+      "step": 19740
+    },
+    {
+      "epoch": 22.469114716766295,
+      "grad_norm": 8.004473686218262,
+      "learning_rate": 9.305487837911046e-05,
+      "loss": 3.1127681732177734,
+      "step": 19750
+    },
+    {
+      "epoch": 22.480500996299458,
+      "grad_norm": 7.39958381652832,
+      "learning_rate": 9.30007042635029e-05,
+      "loss": 2.9369550704956056,
+      "step": 19760
+    },
+    {
+      "epoch": 22.49188727583262,
+      "grad_norm": 7.182417869567871,
+      "learning_rate": 9.294653014789534e-05,
+      "loss": 3.0352306365966797,
+      "step": 19770
+    },
+    {
+      "epoch": 22.503273555365784,
+      "grad_norm": 7.487427711486816,
+      "learning_rate": 9.289235603228778e-05,
+      "loss": 3.1564273834228516,
+      "step": 19780
+    },
+    {
+      "epoch": 22.514659834898946,
+      "grad_norm": 8.101351737976074,
+      "learning_rate": 9.283818191668021e-05,
+      "loss": 3.135332489013672,
+      "step": 19790
+    },
+    {
+      "epoch": 22.52604611443211,
+      "grad_norm": 7.65117883682251,
+      "learning_rate": 9.278400780107265e-05,
+      "loss": 3.1288909912109375,
+      "step": 19800
+    },
+    {
+      "epoch": 22.537432393965272,
+      "grad_norm": 10.03753662109375,
+      "learning_rate": 9.272983368546509e-05,
+      "loss": 2.8136066436767577,
+      "step": 19810
+    },
+    {
+      "epoch": 22.548818673498435,
+      "grad_norm": 7.884417533874512,
+      "learning_rate": 9.267565956985753e-05,
+      "loss": 3.3287643432617187,
+      "step": 19820
+    },
+    {
+      "epoch": 22.560204953031597,
+      "grad_norm": 7.343411922454834,
+      "learning_rate": 9.262148545424997e-05,
+      "loss": 3.0845478057861326,
+      "step": 19830
+    },
+    {
+      "epoch": 22.57159123256476,
+      "grad_norm": 7.07990026473999,
+      "learning_rate": 9.25673113386424e-05,
+      "loss": 3.0407432556152343,
+      "step": 19840
+    },
+    {
+      "epoch": 22.582977512097923,
+      "grad_norm": 7.047083377838135,
+      "learning_rate": 9.251313722303484e-05,
+      "loss": 2.975852394104004,
+      "step": 19850
+    },
+    {
+      "epoch": 22.594363791631086,
+      "grad_norm": 7.662022590637207,
+      "learning_rate": 9.245896310742728e-05,
+      "loss": 3.1719154357910155,
+      "step": 19860
+    },
+    {
+      "epoch": 22.60575007116425,
+      "grad_norm": 7.884767532348633,
+      "learning_rate": 9.240478899181972e-05,
+      "loss": 3.1837724685668944,
+      "step": 19870
+    },
+    {
+      "epoch": 22.61713635069741,
+      "grad_norm": 7.272804260253906,
+      "learning_rate": 9.235061487621214e-05,
+      "loss": 2.9902538299560546,
+      "step": 19880
+    },
+    {
+      "epoch": 22.628522630230574,
+      "grad_norm": 8.392932891845703,
+      "learning_rate": 9.229644076060458e-05,
+      "loss": 3.203661346435547,
+      "step": 19890
+    },
+    {
+      "epoch": 22.639908909763733,
+      "grad_norm": 7.606863975524902,
+      "learning_rate": 9.224226664499702e-05,
+      "loss": 3.245991516113281,
+      "step": 19900
+    },
+    {
+      "epoch": 22.651295189296896,
+      "grad_norm": 8.281023979187012,
+      "learning_rate": 9.218809252938946e-05,
+      "loss": 3.256304168701172,
+      "step": 19910
+    },
+    {
+      "epoch": 22.66268146883006,
+      "grad_norm": 9.497676849365234,
+      "learning_rate": 9.21339184137819e-05,
+      "loss": 3.2775390625,
+      "step": 19920
+    },
+    {
+      "epoch": 22.67406774836322,
+      "grad_norm": 8.350468635559082,
+      "learning_rate": 9.207974429817435e-05,
+      "loss": 3.1510446548461912,
+      "step": 19930
+    },
+    {
+      "epoch": 22.685454027896384,
+      "grad_norm": 8.590188980102539,
+      "learning_rate": 9.202557018256678e-05,
+      "loss": 3.364548110961914,
+      "step": 19940
+    },
+    {
+      "epoch": 22.696840307429547,
+      "grad_norm": 8.386192321777344,
+      "learning_rate": 9.197139606695921e-05,
+      "loss": 3.446294403076172,
+      "step": 19950
+    },
+    {
+      "epoch": 22.70822658696271,
+      "grad_norm": 8.23085880279541,
+      "learning_rate": 9.191722195135165e-05,
+      "loss": 3.0785924911499025,
+      "step": 19960
+    },
+    {
+      "epoch": 22.719612866495872,
+      "grad_norm": 7.4072771072387695,
+      "learning_rate": 9.186304783574408e-05,
+      "loss": 3.299879455566406,
+      "step": 19970
+    },
+    {
+      "epoch": 22.730999146029035,
+      "grad_norm": 6.338870048522949,
+      "learning_rate": 9.180887372013652e-05,
+      "loss": 3.1906932830810546,
+      "step": 19980
+    },
+    {
+      "epoch": 22.742385425562198,
+      "grad_norm": 7.546908378601074,
+      "learning_rate": 9.175469960452896e-05,
+      "loss": 3.0080894470214843,
+      "step": 19990
+    },
+    {
+      "epoch": 22.75377170509536,
+      "grad_norm": 8.439484596252441,
+      "learning_rate": 9.17005254889214e-05,
+      "loss": 3.1106647491455077,
+      "step": 20000
+    },
+    {
+      "epoch": 22.765157984628523,
+      "grad_norm": 8.297350883483887,
+      "learning_rate": 9.164635137331384e-05,
+      "loss": 2.9686201095581053,
+      "step": 20010
+    },
+    {
+      "epoch": 22.776544264161686,
+      "grad_norm": 7.963090896606445,
+      "learning_rate": 9.159217725770628e-05,
+      "loss": 3.098881721496582,
+      "step": 20020
+    },
+    {
+      "epoch": 22.78793054369485,
+      "grad_norm": 8.618762016296387,
+      "learning_rate": 9.15380031420987e-05,
+      "loss": 3.041916084289551,
+      "step": 20030
+    },
+    {
+      "epoch": 22.79931682322801,
+      "grad_norm": 7.110545635223389,
+      "learning_rate": 9.148382902649115e-05,
+      "loss": 3.047221565246582,
+      "step": 20040
+    },
+    {
+      "epoch": 22.810703102761174,
+      "grad_norm": 8.43532657623291,
+      "learning_rate": 9.142965491088359e-05,
+      "loss": 3.025491142272949,
+      "step": 20050
+    },
+    {
+      "epoch": 22.822089382294337,
+      "grad_norm": 7.121822834014893,
+      "learning_rate": 9.137548079527603e-05,
+      "loss": 3.299955368041992,
+      "step": 20060
+    },
+    {
+      "epoch": 22.8334756618275,
+      "grad_norm": 7.654300212860107,
+      "learning_rate": 9.132130667966847e-05,
+      "loss": 3.023772430419922,
+      "step": 20070
+    },
+    {
+      "epoch": 22.844861941360662,
+      "grad_norm": 7.399035930633545,
+      "learning_rate": 9.12671325640609e-05,
+      "loss": 3.1401979446411135,
+      "step": 20080
+    },
+    {
+      "epoch": 22.85624822089382,
+      "grad_norm": 8.279058456420898,
+      "learning_rate": 9.121295844845333e-05,
+      "loss": 3.2731346130371093,
+      "step": 20090
+    },
+    {
+      "epoch": 22.867634500426984,
+      "grad_norm": 8.097005844116211,
+      "learning_rate": 9.115878433284577e-05,
+      "loss": 3.2051586151123046,
+      "step": 20100
+    },
+    {
+      "epoch": 22.879020779960147,
+      "grad_norm": 7.273082256317139,
+      "learning_rate": 9.11046102172382e-05,
+      "loss": 3.2332443237304687,
+      "step": 20110
+    },
+    {
+      "epoch": 22.89040705949331,
+      "grad_norm": 8.696020126342773,
+      "learning_rate": 9.105043610163064e-05,
+      "loss": 3.323190689086914,
+      "step": 20120
+    },
+    {
+      "epoch": 22.901793339026472,
+      "grad_norm": 7.149501323699951,
+      "learning_rate": 9.099626198602308e-05,
+      "loss": 3.2843936920166015,
+      "step": 20130
+    },
+    {
+      "epoch": 22.913179618559635,
+      "grad_norm": 8.37787914276123,
+      "learning_rate": 9.094208787041552e-05,
+      "loss": 3.404499053955078,
+      "step": 20140
+    },
+    {
+      "epoch": 22.924565898092798,
+      "grad_norm": 8.074170112609863,
+      "learning_rate": 9.088791375480796e-05,
+      "loss": 3.290465545654297,
+      "step": 20150
+    },
+    {
+      "epoch": 22.93595217762596,
+      "grad_norm": 7.932847023010254,
+      "learning_rate": 9.08337396392004e-05,
+      "loss": 3.2561920166015623,
+      "step": 20160
+    },
+    {
+      "epoch": 22.947338457159123,
+      "grad_norm": 11.540017127990723,
+      "learning_rate": 9.077956552359283e-05,
+      "loss": 3.1390569686889647,
+      "step": 20170
+    },
+    {
+      "epoch": 22.958724736692286,
+      "grad_norm": 7.136811256408691,
+      "learning_rate": 9.072539140798527e-05,
+      "loss": 3.2981540679931642,
+      "step": 20180
+    },
+    {
+      "epoch": 22.97011101622545,
+      "grad_norm": 7.825453758239746,
+      "learning_rate": 9.067121729237771e-05,
+      "loss": 3.0614255905151366,
+      "step": 20190
+    },
+    {
+      "epoch": 22.98149729575861,
+      "grad_norm": 7.4651007652282715,
+      "learning_rate": 9.061704317677015e-05,
+      "loss": 3.2027751922607424,
+      "step": 20200
+    },
+    {
+      "epoch": 22.992883575291774,
+      "grad_norm": 7.570321559906006,
+      "learning_rate": 9.056286906116258e-05,
+      "loss": 3.1534257888793946,
+      "step": 20210
+    },
+    {
+      "epoch": 23.003415883859947,
+      "grad_norm": 6.957921028137207,
+      "learning_rate": 9.050869494555502e-05,
+      "loss": 2.89981746673584,
+      "step": 20220
+    },
+    {
+      "epoch": 23.01480216339311,
+      "grad_norm": 7.340804576873779,
+      "learning_rate": 9.045452082994746e-05,
+      "loss": 2.898666000366211,
+      "step": 20230
+    },
+    {
+      "epoch": 23.026188442926273,
+      "grad_norm": 7.689297676086426,
+      "learning_rate": 9.040034671433988e-05,
+      "loss": 3.076976776123047,
+      "step": 20240
+    },
+    {
+      "epoch": 23.037574722459436,
+      "grad_norm": 7.607706069946289,
+      "learning_rate": 9.034617259873232e-05,
+      "loss": 3.11181697845459,
+      "step": 20250
+    },
+    {
+      "epoch": 23.0489610019926,
+      "grad_norm": 7.3909502029418945,
+      "learning_rate": 9.029199848312476e-05,
+      "loss": 3.0006891250610352,
+      "step": 20260
+    },
+    {
+      "epoch": 23.06034728152576,
+      "grad_norm": 6.984787464141846,
+      "learning_rate": 9.023782436751721e-05,
+      "loss": 2.905653381347656,
+      "step": 20270
+    },
+    {
+      "epoch": 23.071733561058924,
+      "grad_norm": 8.02444839477539,
+      "learning_rate": 9.018365025190965e-05,
+      "loss": 2.8838172912597657,
+      "step": 20280
+    },
+    {
+      "epoch": 23.083119840592087,
+      "grad_norm": 7.629673480987549,
+      "learning_rate": 9.012947613630209e-05,
+      "loss": 3.2186203002929688,
+      "step": 20290
+    },
+    {
+      "epoch": 23.09450612012525,
+      "grad_norm": 7.7079877853393555,
+      "learning_rate": 9.007530202069451e-05,
+      "loss": 2.9591360092163086,
+      "step": 20300
+    },
+    {
+      "epoch": 23.105892399658412,
+      "grad_norm": 6.760339736938477,
+      "learning_rate": 9.002112790508695e-05,
+      "loss": 2.7239349365234373,
+      "step": 20310
+    },
+    {
+      "epoch": 23.117278679191575,
+      "grad_norm": 9.964285850524902,
+      "learning_rate": 8.996695378947939e-05,
+      "loss": 2.7104719161987303,
+      "step": 20320
+    },
+    {
+      "epoch": 23.128664958724737,
+      "grad_norm": 7.464658737182617,
+      "learning_rate": 8.991277967387183e-05,
+      "loss": 2.8070682525634765,
+      "step": 20330
+    },
+    {
+      "epoch": 23.1400512382579,
+      "grad_norm": 7.5616960525512695,
+      "learning_rate": 8.985860555826426e-05,
+      "loss": 2.708431434631348,
+      "step": 20340
+    },
+    {
+      "epoch": 23.151437517791063,
+      "grad_norm": 7.403007984161377,
+      "learning_rate": 8.98044314426567e-05,
+      "loss": 2.8072072982788088,
+      "step": 20350
+    },
+    {
+      "epoch": 23.162823797324226,
+      "grad_norm": 8.316858291625977,
+      "learning_rate": 8.975025732704914e-05,
+      "loss": 3.012776565551758,
+      "step": 20360
+    },
+    {
+      "epoch": 23.17421007685739,
+      "grad_norm": 7.397007942199707,
+      "learning_rate": 8.969608321144158e-05,
+      "loss": 3.178554344177246,
+      "step": 20370
+    },
+    {
+      "epoch": 23.185596356390548,
+      "grad_norm": 7.289053916931152,
+      "learning_rate": 8.964190909583402e-05,
+      "loss": 2.6939666748046873,
+      "step": 20380
+    },
+    {
+      "epoch": 23.19698263592371,
+      "grad_norm": 7.612484455108643,
+      "learning_rate": 8.958773498022645e-05,
+      "loss": 3.1522052764892576,
+      "step": 20390
+    },
+    {
+      "epoch": 23.208368915456873,
+      "grad_norm": 7.743655681610107,
+      "learning_rate": 8.953356086461889e-05,
+      "loss": 2.716688346862793,
+      "step": 20400
+    },
+    {
+      "epoch": 23.219755194990036,
+      "grad_norm": 10.625493049621582,
+      "learning_rate": 8.947938674901133e-05,
+      "loss": 3.075759506225586,
+      "step": 20410
+    },
+    {
+      "epoch": 23.2311414745232,
+      "grad_norm": 7.2747883796691895,
+      "learning_rate": 8.942521263340377e-05,
+      "loss": 2.8906108856201174,
+      "step": 20420
+    },
+    {
+      "epoch": 23.24252775405636,
+      "grad_norm": 8.012664794921875,
+      "learning_rate": 8.937103851779621e-05,
+      "loss": 2.891446113586426,
+      "step": 20430
+    },
+    {
+      "epoch": 23.253914033589524,
+      "grad_norm": 6.6297430992126465,
+      "learning_rate": 8.931686440218863e-05,
+      "loss": 2.9877086639404298,
+      "step": 20440
+    },
+    {
+      "epoch": 23.265300313122687,
+      "grad_norm": 7.2411274909973145,
+      "learning_rate": 8.926269028658107e-05,
+      "loss": 2.8750194549560546,
+      "step": 20450
+    },
+    {
+      "epoch": 23.27668659265585,
+      "grad_norm": 7.526220321655273,
+      "learning_rate": 8.920851617097351e-05,
+      "loss": 2.7831699371337892,
+      "step": 20460
+    },
+    {
+      "epoch": 23.288072872189012,
+      "grad_norm": 7.155262470245361,
+      "learning_rate": 8.915434205536595e-05,
+      "loss": 2.864755630493164,
+      "step": 20470
+    },
+    {
+      "epoch": 23.299459151722175,
+      "grad_norm": 8.7533540725708,
+      "learning_rate": 8.910016793975838e-05,
+      "loss": 3.0703664779663087,
+      "step": 20480
+    },
+    {
+      "epoch": 23.310845431255338,
+      "grad_norm": 8.419981956481934,
+      "learning_rate": 8.904599382415082e-05,
+      "loss": 3.047661781311035,
+      "step": 20490
+    },
+    {
+      "epoch": 23.3222317107885,
+      "grad_norm": 7.7640204429626465,
+      "learning_rate": 8.899181970854327e-05,
+      "loss": 2.887657356262207,
+      "step": 20500
+    },
+    {
+      "epoch": 23.333617990321663,
+      "grad_norm": 7.903047561645508,
+      "learning_rate": 8.89376455929357e-05,
+      "loss": 2.9257579803466798,
+      "step": 20510
+    },
+    {
+      "epoch": 23.345004269854826,
+      "grad_norm": 7.807440757751465,
+      "learning_rate": 8.888347147732814e-05,
+      "loss": 2.984356689453125,
+      "step": 20520
+    },
+    {
+      "epoch": 23.35639054938799,
+      "grad_norm": 7.732547760009766,
+      "learning_rate": 8.882929736172057e-05,
+      "loss": 3.2712615966796874,
+      "step": 20530
+    },
+    {
+      "epoch": 23.36777682892115,
+      "grad_norm": 9.053909301757812,
+      "learning_rate": 8.877512324611301e-05,
+      "loss": 2.9967681884765627,
+      "step": 20540
+    },
+    {
+      "epoch": 23.379163108454314,
+      "grad_norm": 7.440546035766602,
+      "learning_rate": 8.872094913050545e-05,
+      "loss": 3.0605968475341796,
+      "step": 20550
+    },
+    {
+      "epoch": 23.390549387987477,
+      "grad_norm": 6.801850318908691,
+      "learning_rate": 8.866677501489789e-05,
+      "loss": 3.1235517501831054,
+      "step": 20560
+    },
+    {
+      "epoch": 23.401935667520636,
+      "grad_norm": 8.020711898803711,
+      "learning_rate": 8.861260089929033e-05,
+      "loss": 2.8574342727661133,
+      "step": 20570
+    },
+    {
+      "epoch": 23.4133219470538,
+      "grad_norm": 7.850024223327637,
+      "learning_rate": 8.855842678368276e-05,
+      "loss": 3.050294303894043,
+      "step": 20580
+    },
+    {
+      "epoch": 23.42470822658696,
+      "grad_norm": 7.764718532562256,
+      "learning_rate": 8.850425266807519e-05,
+      "loss": 3.0289308547973635,
+      "step": 20590
+    },
+    {
+      "epoch": 23.436094506120124,
+      "grad_norm": 6.614105701446533,
+      "learning_rate": 8.845007855246763e-05,
+      "loss": 2.7201135635375975,
+      "step": 20600
+    },
+    {
+      "epoch": 23.447480785653287,
+      "grad_norm": 7.509239673614502,
+      "learning_rate": 8.839590443686008e-05,
+      "loss": 3.0174678802490233,
+      "step": 20610
+    },
+    {
+      "epoch": 23.45886706518645,
+      "grad_norm": 7.190005302429199,
+      "learning_rate": 8.834173032125252e-05,
+      "loss": 2.841200828552246,
+      "step": 20620
+    },
+    {
+      "epoch": 23.470253344719612,
+      "grad_norm": 8.263400077819824,
+      "learning_rate": 8.828755620564495e-05,
+      "loss": 2.7141368865966795,
+      "step": 20630
+    },
+    {
+      "epoch": 23.481639624252775,
+      "grad_norm": 6.695593357086182,
+      "learning_rate": 8.823338209003739e-05,
+      "loss": 2.7828269958496095,
+      "step": 20640
+    },
+    {
+      "epoch": 23.493025903785938,
+      "grad_norm": 8.470293998718262,
+      "learning_rate": 8.817920797442982e-05,
+      "loss": 3.0112863540649415,
+      "step": 20650
+    },
+    {
+      "epoch": 23.5044121833191,
+      "grad_norm": 7.583410263061523,
+      "learning_rate": 8.812503385882225e-05,
+      "loss": 2.9472209930419924,
+      "step": 20660
+    },
+    {
+      "epoch": 23.515798462852263,
+      "grad_norm": 7.364217758178711,
+      "learning_rate": 8.807085974321469e-05,
+      "loss": 2.9927459716796876,
+      "step": 20670
+    },
+    {
+      "epoch": 23.527184742385426,
+      "grad_norm": 8.428694725036621,
+      "learning_rate": 8.801668562760713e-05,
+      "loss": 3.037489128112793,
+      "step": 20680
+    },
+    {
+      "epoch": 23.53857102191859,
+      "grad_norm": 8.37234115600586,
+      "learning_rate": 8.796251151199957e-05,
+      "loss": 2.8356271743774415,
+      "step": 20690
+    },
+    {
+      "epoch": 23.54995730145175,
+      "grad_norm": 7.239820957183838,
+      "learning_rate": 8.7908337396392e-05,
+      "loss": 2.9129606246948243,
+      "step": 20700
+    },
+    {
+      "epoch": 23.561343580984914,
+      "grad_norm": 7.503726959228516,
+      "learning_rate": 8.785416328078444e-05,
+      "loss": 2.9780012130737306,
+      "step": 20710
+    },
+    {
+      "epoch": 23.572729860518077,
+      "grad_norm": 6.911001205444336,
+      "learning_rate": 8.779998916517688e-05,
+      "loss": 2.829633331298828,
+      "step": 20720
+    },
+    {
+      "epoch": 23.58411614005124,
+      "grad_norm": 8.57630729675293,
+      "learning_rate": 8.774581504956932e-05,
+      "loss": 3.0809423446655275,
+      "step": 20730
+    },
+    {
+      "epoch": 23.5955024195844,
+      "grad_norm": 7.118391990661621,
+      "learning_rate": 8.769164093396176e-05,
+      "loss": 2.9418434143066405,
+      "step": 20740
+    },
+    {
+      "epoch": 23.60688869911756,
+      "grad_norm": 8.470577239990234,
+      "learning_rate": 8.76374668183542e-05,
+      "loss": 3.073319435119629,
+      "step": 20750
+    },
+    {
+      "epoch": 23.618274978650724,
+      "grad_norm": 7.737377643585205,
+      "learning_rate": 8.758329270274663e-05,
+      "loss": 2.973114013671875,
+      "step": 20760
+    },
+    {
+      "epoch": 23.629661258183887,
+      "grad_norm": 7.212495803833008,
+      "learning_rate": 8.752911858713907e-05,
+      "loss": 3.256830596923828,
+      "step": 20770
+    },
+    {
+      "epoch": 23.64104753771705,
+      "grad_norm": 8.123977661132812,
+      "learning_rate": 8.747494447153151e-05,
+      "loss": 2.8294878005981445,
+      "step": 20780
+    },
+    {
+      "epoch": 23.652433817250213,
+      "grad_norm": 8.355793952941895,
+      "learning_rate": 8.742077035592395e-05,
+      "loss": 2.869054412841797,
+      "step": 20790
+    },
+    {
+      "epoch": 23.663820096783375,
+      "grad_norm": 8.310219764709473,
+      "learning_rate": 8.736659624031637e-05,
+      "loss": 3.0746349334716796,
+      "step": 20800
+    },
+    {
+      "epoch": 23.675206376316538,
+      "grad_norm": 7.146482944488525,
+      "learning_rate": 8.731242212470881e-05,
+      "loss": 2.7864452362060548,
+      "step": 20810
+    },
+    {
+      "epoch": 23.6865926558497,
+      "grad_norm": 7.662703990936279,
+      "learning_rate": 8.725824800910125e-05,
+      "loss": 2.792169952392578,
+      "step": 20820
+    },
+    {
+      "epoch": 23.697978935382864,
+      "grad_norm": 7.96389627456665,
+      "learning_rate": 8.720407389349369e-05,
+      "loss": 3.1457258224487306,
+      "step": 20830
+    },
+    {
+      "epoch": 23.709365214916026,
+      "grad_norm": 6.582839012145996,
+      "learning_rate": 8.714989977788614e-05,
+      "loss": 2.910396385192871,
+      "step": 20840
+    },
+    {
+      "epoch": 23.72075149444919,
+      "grad_norm": 7.634960651397705,
+      "learning_rate": 8.709572566227858e-05,
+      "loss": 3.0249223709106445,
+      "step": 20850
+    },
+    {
+      "epoch": 23.732137773982352,
+      "grad_norm": 8.20224380493164,
+      "learning_rate": 8.7041551546671e-05,
+      "loss": 2.8246204376220705,
+      "step": 20860
+    },
+    {
+      "epoch": 23.743524053515515,
+      "grad_norm": 7.890283584594727,
+      "learning_rate": 8.698737743106344e-05,
+      "loss": 3.098486137390137,
+      "step": 20870
+    },
+    {
+      "epoch": 23.754910333048677,
+      "grad_norm": 7.693148612976074,
+      "learning_rate": 8.693320331545588e-05,
+      "loss": 2.9983917236328126,
+      "step": 20880
+    },
+    {
+      "epoch": 23.76629661258184,
+      "grad_norm": 7.275303363800049,
+      "learning_rate": 8.687902919984832e-05,
+      "loss": 3.028023529052734,
+      "step": 20890
+    },
+    {
+      "epoch": 23.777682892115003,
+      "grad_norm": 9.735662460327148,
+      "learning_rate": 8.682485508424075e-05,
+      "loss": 2.907429313659668,
+      "step": 20900
+    },
+    {
+      "epoch": 23.789069171648165,
+      "grad_norm": 7.711111545562744,
+      "learning_rate": 8.677068096863319e-05,
+      "loss": 2.9936874389648436,
+      "step": 20910
+    },
+    {
+      "epoch": 23.800455451181328,
+      "grad_norm": 7.945051193237305,
+      "learning_rate": 8.671650685302563e-05,
+      "loss": 3.211922836303711,
+      "step": 20920
+    },
+    {
+      "epoch": 23.811841730714487,
+      "grad_norm": 7.160228729248047,
+      "learning_rate": 8.666233273741807e-05,
+      "loss": 2.911302375793457,
+      "step": 20930
+    },
+    {
+      "epoch": 23.82322801024765,
+      "grad_norm": 7.800594329833984,
+      "learning_rate": 8.660815862181049e-05,
+      "loss": 3.017009735107422,
+      "step": 20940
+    },
+    {
+      "epoch": 23.834614289780813,
+      "grad_norm": 7.576048851013184,
+      "learning_rate": 8.655398450620294e-05,
+      "loss": 3.0092130661010743,
+      "step": 20950
+    },
+    {
+      "epoch": 23.846000569313976,
+      "grad_norm": 7.396543979644775,
+      "learning_rate": 8.649981039059538e-05,
+      "loss": 3.004311180114746,
+      "step": 20960
+    },
+    {
+      "epoch": 23.85738684884714,
+      "grad_norm": 7.172977447509766,
+      "learning_rate": 8.644563627498782e-05,
+      "loss": 2.650033187866211,
+      "step": 20970
+    },
+    {
+      "epoch": 23.8687731283803,
+      "grad_norm": 7.44117546081543,
+      "learning_rate": 8.639146215938026e-05,
+      "loss": 2.9172443389892577,
+      "step": 20980
+    },
+    {
+      "epoch": 23.880159407913464,
+      "grad_norm": 8.184832572937012,
+      "learning_rate": 8.63372880437727e-05,
+      "loss": 3.129201316833496,
+      "step": 20990
+    },
+    {
+      "epoch": 23.891545687446627,
+      "grad_norm": 7.225976467132568,
+      "learning_rate": 8.628311392816512e-05,
+      "loss": 2.9618593215942384,
+      "step": 21000
+    },
+    {
+      "epoch": 23.90293196697979,
+      "grad_norm": 7.948321342468262,
+      "learning_rate": 8.622893981255756e-05,
+      "loss": 3.0456144332885744,
+      "step": 21010
+    },
+    {
+      "epoch": 23.914318246512952,
+      "grad_norm": 8.000722885131836,
+      "learning_rate": 8.617476569695e-05,
+      "loss": 3.007844924926758,
+      "step": 21020
+    },
+    {
+      "epoch": 23.925704526046115,
+      "grad_norm": 7.9927215576171875,
+      "learning_rate": 8.612059158134243e-05,
+      "loss": 2.997426223754883,
+      "step": 21030
+    },
+    {
+      "epoch": 23.937090805579277,
+      "grad_norm": 9.083395957946777,
+      "learning_rate": 8.606641746573487e-05,
+      "loss": 3.4572025299072267,
+      "step": 21040
+    },
+    {
+      "epoch": 23.94847708511244,
+      "grad_norm": 7.858376502990723,
+      "learning_rate": 8.601224335012731e-05,
+      "loss": 2.95123291015625,
+      "step": 21050
+    },
+    {
+      "epoch": 23.959863364645603,
+      "grad_norm": 6.826962471008301,
+      "learning_rate": 8.595806923451976e-05,
+      "loss": 3.145720672607422,
+      "step": 21060
+    },
+    {
+      "epoch": 23.971249644178766,
+      "grad_norm": 8.091140747070312,
+      "learning_rate": 8.590389511891219e-05,
+      "loss": 2.7883310317993164,
+      "step": 21070
+    },
+    {
+      "epoch": 23.98263592371193,
+      "grad_norm": 7.429564952850342,
+      "learning_rate": 8.584972100330462e-05,
+      "loss": 3.2133480072021485,
+      "step": 21080
+    },
+    {
+      "epoch": 23.99402220324509,
+      "grad_norm": 6.603694438934326,
+      "learning_rate": 8.579554688769706e-05,
+      "loss": 2.7781740188598634,
+      "step": 21090
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 36918,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 42,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.6981989206528e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}