diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4929 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3488,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014334862385321102,
+      "grad_norm": 10.930955258982435,
+      "learning_rate": 2.865329512893983e-07,
+      "loss": 1.1122,
+      "step": 5
+    },
+    {
+      "epoch": 0.0028669724770642203,
+      "grad_norm": 7.957126318923027,
+      "learning_rate": 5.730659025787966e-07,
+      "loss": 1.1033,
+      "step": 10
+    },
+    {
+      "epoch": 0.0043004587155963305,
+      "grad_norm": 6.252488126522345,
+      "learning_rate": 8.595988538681949e-07,
+      "loss": 1.1448,
+      "step": 15
+    },
+    {
+      "epoch": 0.005733944954128441,
+      "grad_norm": 6.589440813762702,
+      "learning_rate": 1.1461318051575932e-06,
+      "loss": 1.0284,
+      "step": 20
+    },
+    {
+      "epoch": 0.007167431192660551,
+      "grad_norm": 4.917441825534674,
+      "learning_rate": 1.4326647564469915e-06,
+      "loss": 1.0131,
+      "step": 25
+    },
+    {
+      "epoch": 0.008600917431192661,
+      "grad_norm": 4.611821748972413,
+      "learning_rate": 1.7191977077363897e-06,
+      "loss": 1.0492,
+      "step": 30
+    },
+    {
+      "epoch": 0.010034403669724771,
+      "grad_norm": 4.602326061969326,
+      "learning_rate": 2.005730659025788e-06,
+      "loss": 1.0249,
+      "step": 35
+    },
+    {
+      "epoch": 0.011467889908256881,
+      "grad_norm": 4.795975437768488,
+      "learning_rate": 2.2922636103151864e-06,
+      "loss": 0.9889,
+      "step": 40
+    },
+    {
+      "epoch": 0.012901376146788992,
+      "grad_norm": 4.881452695659322,
+      "learning_rate": 2.5787965616045845e-06,
+      "loss": 0.9654,
+      "step": 45
+    },
+    {
+      "epoch": 0.014334862385321102,
+      "grad_norm": 4.816747748013977,
+      "learning_rate": 2.865329512893983e-06,
+      "loss": 1.0479,
+      "step": 50
+    },
+    {
+      "epoch": 0.01576834862385321,
+      "grad_norm": 4.672562758642394,
+      "learning_rate": 3.151862464183381e-06,
+      "loss": 0.9692,
+      "step": 55
+    },
+    {
+      "epoch": 0.017201834862385322,
+      "grad_norm": 4.665632444621185,
+      "learning_rate": 3.4383954154727795e-06,
+      "loss": 1.0033,
+      "step": 60
+    },
+    {
+      "epoch": 0.01863532110091743,
+      "grad_norm": 4.5487339829077795,
+      "learning_rate": 3.724928366762178e-06,
+      "loss": 1.0206,
+      "step": 65
+    },
+    {
+      "epoch": 0.020068807339449542,
+      "grad_norm": 4.72120821418095,
+      "learning_rate": 4.011461318051576e-06,
+      "loss": 1.0065,
+      "step": 70
+    },
+    {
+      "epoch": 0.02150229357798165,
+      "grad_norm": 4.488698654642803,
+      "learning_rate": 4.2979942693409744e-06,
+      "loss": 0.9634,
+      "step": 75
+    },
+    {
+      "epoch": 0.022935779816513763,
+      "grad_norm": 4.502540106177661,
+      "learning_rate": 4.584527220630373e-06,
+      "loss": 0.9633,
+      "step": 80
+    },
+    {
+      "epoch": 0.02436926605504587,
+      "grad_norm": 4.538891592130377,
+      "learning_rate": 4.871060171919771e-06,
+      "loss": 0.9148,
+      "step": 85
+    },
+    {
+      "epoch": 0.025802752293577983,
+      "grad_norm": 4.89489963864345,
+      "learning_rate": 5.157593123209169e-06,
+      "loss": 1.0303,
+      "step": 90
+    },
+    {
+      "epoch": 0.02723623853211009,
+      "grad_norm": 4.546009685259081,
+      "learning_rate": 5.444126074498568e-06,
+      "loss": 1.0174,
+      "step": 95
+    },
+    {
+      "epoch": 0.028669724770642203,
+      "grad_norm": 4.542941293141969,
+      "learning_rate": 5.730659025787966e-06,
+      "loss": 1.0251,
+      "step": 100
+    },
+    {
+      "epoch": 0.030103211009174312,
+      "grad_norm": 4.571406282018777,
+      "learning_rate": 6.017191977077364e-06,
+      "loss": 1.0376,
+      "step": 105
+    },
+    {
+      "epoch": 0.03153669724770642,
+      "grad_norm": 4.75986045778734,
+      "learning_rate": 6.303724928366762e-06,
+      "loss": 0.9827,
+      "step": 110
+    },
+    {
+      "epoch": 0.03297018348623853,
+      "grad_norm": 4.393838501543807,
+      "learning_rate": 6.590257879656161e-06,
+      "loss": 0.9533,
+      "step": 115
+    },
+    {
+      "epoch": 0.034403669724770644,
+      "grad_norm": 4.780066323607276,
+      "learning_rate": 6.876790830945559e-06,
+      "loss": 1.0018,
+      "step": 120
+    },
+    {
+      "epoch": 0.03583715596330275,
+      "grad_norm": 4.414441537089264,
+      "learning_rate": 7.163323782234957e-06,
+      "loss": 1.0135,
+      "step": 125
+    },
+    {
+      "epoch": 0.03727064220183486,
+      "grad_norm": 4.301618750664585,
+      "learning_rate": 7.449856733524356e-06,
+      "loss": 1.0201,
+      "step": 130
+    },
+    {
+      "epoch": 0.03870412844036697,
+      "grad_norm": 4.3543481672910245,
+      "learning_rate": 7.736389684813753e-06,
+      "loss": 1.0236,
+      "step": 135
+    },
+    {
+      "epoch": 0.040137614678899085,
+      "grad_norm": 4.31887452196869,
+      "learning_rate": 8.022922636103152e-06,
+      "loss": 1.0413,
+      "step": 140
+    },
+    {
+      "epoch": 0.04157110091743119,
+      "grad_norm": 4.19580797580846,
+      "learning_rate": 8.30945558739255e-06,
+      "loss": 0.9999,
+      "step": 145
+    },
+    {
+      "epoch": 0.0430045871559633,
+      "grad_norm": 4.2941554889344635,
+      "learning_rate": 8.595988538681949e-06,
+      "loss": 1.0824,
+      "step": 150
+    },
+    {
+      "epoch": 0.04443807339449541,
+      "grad_norm": 4.430699539319586,
+      "learning_rate": 8.882521489971347e-06,
+      "loss": 1.0545,
+      "step": 155
+    },
+    {
+      "epoch": 0.045871559633027525,
+      "grad_norm": 4.357465541547218,
+      "learning_rate": 9.169054441260746e-06,
+      "loss": 1.0662,
+      "step": 160
+    },
+    {
+      "epoch": 0.047305045871559634,
+      "grad_norm": 5.722224893724134,
+      "learning_rate": 9.455587392550144e-06,
+      "loss": 1.0197,
+      "step": 165
+    },
+    {
+      "epoch": 0.04873853211009174,
+      "grad_norm": 4.374513298911362,
+      "learning_rate": 9.742120343839543e-06,
+      "loss": 1.046,
+      "step": 170
+    },
+    {
+      "epoch": 0.05017201834862385,
+      "grad_norm": 4.593743802840811,
+      "learning_rate": 1.0028653295128941e-05,
+      "loss": 1.0739,
+      "step": 175
+    },
+    {
+      "epoch": 0.051605504587155966,
+      "grad_norm": 4.333415176606838,
+      "learning_rate": 1.0315186246418338e-05,
+      "loss": 1.0833,
+      "step": 180
+    },
+    {
+      "epoch": 0.053038990825688075,
+      "grad_norm": 4.426601251103325,
+      "learning_rate": 1.0601719197707738e-05,
+      "loss": 1.0799,
+      "step": 185
+    },
+    {
+      "epoch": 0.05447247706422018,
+      "grad_norm": 4.556659147888879,
+      "learning_rate": 1.0888252148997137e-05,
+      "loss": 1.0405,
+      "step": 190
+    },
+    {
+      "epoch": 0.05590596330275229,
+      "grad_norm": 4.3383547238785605,
+      "learning_rate": 1.1174785100286533e-05,
+      "loss": 1.0189,
+      "step": 195
+    },
+    {
+      "epoch": 0.05733944954128441,
+      "grad_norm": 4.125983340413087,
+      "learning_rate": 1.1461318051575932e-05,
+      "loss": 1.0592,
+      "step": 200
+    },
+    {
+      "epoch": 0.058772935779816515,
+      "grad_norm": 4.479422603616749,
+      "learning_rate": 1.1747851002865332e-05,
+      "loss": 1.0626,
+      "step": 205
+    },
+    {
+      "epoch": 0.060206422018348624,
+      "grad_norm": 4.506035260293105,
+      "learning_rate": 1.2034383954154729e-05,
+      "loss": 1.0506,
+      "step": 210
+    },
+    {
+      "epoch": 0.06163990825688073,
+      "grad_norm": 5.22525942718489,
+      "learning_rate": 1.2320916905444127e-05,
+      "loss": 1.0855,
+      "step": 215
+    },
+    {
+      "epoch": 0.06307339449541284,
+      "grad_norm": 4.779500493076456,
+      "learning_rate": 1.2607449856733524e-05,
+      "loss": 1.0872,
+      "step": 220
+    },
+    {
+      "epoch": 0.06450688073394495,
+      "grad_norm": 4.1419371246845555,
+      "learning_rate": 1.2893982808022924e-05,
+      "loss": 1.0158,
+      "step": 225
+    },
+    {
+      "epoch": 0.06594036697247706,
+      "grad_norm": 4.55193247523381,
+      "learning_rate": 1.3180515759312323e-05,
+      "loss": 1.1341,
+      "step": 230
+    },
+    {
+      "epoch": 0.06737385321100918,
+      "grad_norm": 4.821922591302919,
+      "learning_rate": 1.346704871060172e-05,
+      "loss": 1.0317,
+      "step": 235
+    },
+    {
+      "epoch": 0.06880733944954129,
+      "grad_norm": 5.7218678278363075,
+      "learning_rate": 1.3753581661891118e-05,
+      "loss": 1.0929,
+      "step": 240
+    },
+    {
+      "epoch": 0.0702408256880734,
+      "grad_norm": 4.5897462960681885,
+      "learning_rate": 1.4040114613180518e-05,
+      "loss": 1.0707,
+      "step": 245
+    },
+    {
+      "epoch": 0.0716743119266055,
+      "grad_norm": 4.9044704729529265,
+      "learning_rate": 1.4326647564469915e-05,
+      "loss": 1.0589,
+      "step": 250
+    },
+    {
+      "epoch": 0.07310779816513761,
+      "grad_norm": 4.317942267526155,
+      "learning_rate": 1.4613180515759313e-05,
+      "loss": 1.0944,
+      "step": 255
+    },
+    {
+      "epoch": 0.07454128440366972,
+      "grad_norm": 4.019443011772793,
+      "learning_rate": 1.4899713467048712e-05,
+      "loss": 1.1014,
+      "step": 260
+    },
+    {
+      "epoch": 0.07597477064220183,
+      "grad_norm": 5.05645456273739,
+      "learning_rate": 1.518624641833811e-05,
+      "loss": 1.0855,
+      "step": 265
+    },
+    {
+      "epoch": 0.07740825688073394,
+      "grad_norm": 4.631991611940915,
+      "learning_rate": 1.5472779369627507e-05,
+      "loss": 1.1797,
+      "step": 270
+    },
+    {
+      "epoch": 0.07884174311926606,
+      "grad_norm": 4.863391448983006,
+      "learning_rate": 1.5759312320916907e-05,
+      "loss": 1.1189,
+      "step": 275
+    },
+    {
+      "epoch": 0.08027522935779817,
+      "grad_norm": 5.67171228826149,
+      "learning_rate": 1.6045845272206304e-05,
+      "loss": 1.0687,
+      "step": 280
+    },
+    {
+      "epoch": 0.08170871559633028,
+      "grad_norm": 4.863402157719064,
+      "learning_rate": 1.6332378223495704e-05,
+      "loss": 1.1033,
+      "step": 285
+    },
+    {
+      "epoch": 0.08314220183486239,
+      "grad_norm": 4.660992837053409,
+      "learning_rate": 1.66189111747851e-05,
+      "loss": 1.1543,
+      "step": 290
+    },
+    {
+      "epoch": 0.0845756880733945,
+      "grad_norm": 4.552069083735871,
+      "learning_rate": 1.69054441260745e-05,
+      "loss": 1.1464,
+      "step": 295
+    },
+    {
+      "epoch": 0.0860091743119266,
+      "grad_norm": 4.61726981167071,
+      "learning_rate": 1.7191977077363898e-05,
+      "loss": 1.1332,
+      "step": 300
+    },
+    {
+      "epoch": 0.08744266055045871,
+      "grad_norm": 4.9734394128477755,
+      "learning_rate": 1.7478510028653298e-05,
+      "loss": 1.0608,
+      "step": 305
+    },
+    {
+      "epoch": 0.08887614678899082,
+      "grad_norm": 4.080774915213187,
+      "learning_rate": 1.7765042979942695e-05,
+      "loss": 1.1737,
+      "step": 310
+    },
+    {
+      "epoch": 0.09030963302752294,
+      "grad_norm": 4.9010757003007805,
+      "learning_rate": 1.805157593123209e-05,
+      "loss": 1.152,
+      "step": 315
+    },
+    {
+      "epoch": 0.09174311926605505,
+      "grad_norm": 4.348268075612745,
+      "learning_rate": 1.833810888252149e-05,
+      "loss": 1.1493,
+      "step": 320
+    },
+    {
+      "epoch": 0.09317660550458716,
+      "grad_norm": 5.525129625035676,
+      "learning_rate": 1.8624641833810892e-05,
+      "loss": 1.109,
+      "step": 325
+    },
+    {
+      "epoch": 0.09461009174311927,
+      "grad_norm": 6.128793939259254,
+      "learning_rate": 1.891117478510029e-05,
+      "loss": 1.1883,
+      "step": 330
+    },
+    {
+      "epoch": 0.09604357798165138,
+      "grad_norm": 4.553408862504849,
+      "learning_rate": 1.9197707736389685e-05,
+      "loss": 1.1294,
+      "step": 335
+    },
+    {
+      "epoch": 0.09747706422018348,
+      "grad_norm": 4.512129266800345,
+      "learning_rate": 1.9484240687679085e-05,
+      "loss": 1.2076,
+      "step": 340
+    },
+    {
+      "epoch": 0.0989105504587156,
+      "grad_norm": 4.294698339999475,
+      "learning_rate": 1.9770773638968482e-05,
+      "loss": 1.1455,
+      "step": 345
+    },
+    {
+      "epoch": 0.1003440366972477,
+      "grad_norm": 4.649384755228939,
+      "learning_rate": 1.999999499173752e-05,
+      "loss": 1.142,
+      "step": 350
+    },
+    {
+      "epoch": 0.10177752293577981,
+      "grad_norm": 4.639825467109486,
+      "learning_rate": 1.999981970307739e-05,
+      "loss": 1.19,
+      "step": 355
+    },
+    {
+      "epoch": 0.10321100917431193,
+      "grad_norm": 4.447236882707046,
+      "learning_rate": 1.999939400630968e-05,
+      "loss": 1.1753,
+      "step": 360
+    },
+    {
+      "epoch": 0.10464449541284404,
+      "grad_norm": 4.092068716884874,
+      "learning_rate": 1.999871791209438e-05,
+      "loss": 1.1893,
+      "step": 365
+    },
+    {
+      "epoch": 0.10607798165137615,
+      "grad_norm": 4.398948638445328,
+      "learning_rate": 1.9997791437361734e-05,
+      "loss": 1.1555,
+      "step": 370
+    },
+    {
+      "epoch": 0.10751146788990826,
+      "grad_norm": 4.701129620714216,
+      "learning_rate": 1.9996614605311848e-05,
+      "loss": 1.2011,
+      "step": 375
+    },
+    {
+      "epoch": 0.10894495412844037,
+      "grad_norm": 4.821236633722439,
+      "learning_rate": 1.999518744541407e-05,
+      "loss": 1.1776,
+      "step": 380
+    },
+    {
+      "epoch": 0.11037844036697247,
+      "grad_norm": 4.335100849608752,
+      "learning_rate": 1.9993509993406297e-05,
+      "loss": 1.1902,
+      "step": 385
+    },
+    {
+      "epoch": 0.11181192660550458,
+      "grad_norm": 4.068825093466267,
+      "learning_rate": 1.9991582291294042e-05,
+      "loss": 1.1836,
+      "step": 390
+    },
+    {
+      "epoch": 0.11324541284403669,
+      "grad_norm": 3.9520872546047143,
+      "learning_rate": 1.9989404387349393e-05,
+      "loss": 1.1439,
+      "step": 395
+    },
+    {
+      "epoch": 0.11467889908256881,
+      "grad_norm": 4.582077126321336,
+      "learning_rate": 1.998697633610982e-05,
+      "loss": 1.2094,
+      "step": 400
+    },
+    {
+      "epoch": 0.11611238532110092,
+      "grad_norm": 4.38687782745242,
+      "learning_rate": 1.998429819837679e-05,
+      "loss": 1.1866,
+      "step": 405
+    },
+    {
+      "epoch": 0.11754587155963303,
+      "grad_norm": 4.6771014427247275,
+      "learning_rate": 1.998137004121425e-05,
+      "loss": 1.1587,
+      "step": 410
+    },
+    {
+      "epoch": 0.11897935779816514,
+      "grad_norm": 4.218200613872475,
+      "learning_rate": 1.9978191937946955e-05,
+      "loss": 1.1411,
+      "step": 415
+    },
+    {
+      "epoch": 0.12041284403669725,
+      "grad_norm": 4.509609936000089,
+      "learning_rate": 1.9974763968158614e-05,
+      "loss": 1.1772,
+      "step": 420
+    },
+    {
+      "epoch": 0.12184633027522936,
+      "grad_norm": 5.925174301278395,
+      "learning_rate": 1.9971086217689928e-05,
+      "loss": 1.2411,
+      "step": 425
+    },
+    {
+      "epoch": 0.12327981651376146,
+      "grad_norm": 4.022377669658578,
+      "learning_rate": 1.9967158778636405e-05,
+      "loss": 1.1887,
+      "step": 430
+    },
+    {
+      "epoch": 0.12471330275229357,
+      "grad_norm": 4.539584147379229,
+      "learning_rate": 1.996298174934608e-05,
+      "loss": 1.1717,
+      "step": 435
+    },
+    {
+      "epoch": 0.12614678899082568,
+      "grad_norm": 9.545718854729863,
+      "learning_rate": 1.9958555234417035e-05,
+      "loss": 1.1879,
+      "step": 440
+    },
+    {
+      "epoch": 0.1275802752293578,
+      "grad_norm": 22.053598752620175,
+      "learning_rate": 1.995387934469479e-05,
+      "loss": 1.172,
+      "step": 445
+    },
+    {
+      "epoch": 0.1290137614678899,
+      "grad_norm": 12.911881630750296,
+      "learning_rate": 1.994895419726953e-05,
+      "loss": 1.2034,
+      "step": 450
+    },
+    {
+      "epoch": 0.13044724770642202,
+      "grad_norm": 4.856805114224933,
+      "learning_rate": 1.9943779915473165e-05,
+      "loss": 1.1795,
+      "step": 455
+    },
+    {
+      "epoch": 0.13188073394495411,
+      "grad_norm": 4.286105662410557,
+      "learning_rate": 1.9938356628876235e-05,
+      "loss": 1.1866,
+      "step": 460
+    },
+    {
+      "epoch": 0.13331422018348624,
+      "grad_norm": 4.2624287742900995,
+      "learning_rate": 1.9932684473284687e-05,
+      "loss": 1.1894,
+      "step": 465
+    },
+    {
+      "epoch": 0.13474770642201836,
+      "grad_norm": 4.716264114725749,
+      "learning_rate": 1.9926763590736457e-05,
+      "loss": 1.2835,
+      "step": 470
+    },
+    {
+      "epoch": 0.13618119266055045,
+      "grad_norm": 4.2462330499210115,
+      "learning_rate": 1.992059412949791e-05,
+      "loss": 1.2412,
+      "step": 475
+    },
+    {
+      "epoch": 0.13761467889908258,
+      "grad_norm": 4.391420347453553,
+      "learning_rate": 1.9914176244060158e-05,
+      "loss": 1.1516,
+      "step": 480
+    },
+    {
+      "epoch": 0.13904816513761467,
+      "grad_norm": 75.66535297260735,
+      "learning_rate": 1.9907510095135142e-05,
+      "loss": 1.2722,
+      "step": 485
+    },
+    {
+      "epoch": 0.1404816513761468,
+      "grad_norm": 5.310902667535963,
+      "learning_rate": 1.9900595849651645e-05,
+      "loss": 1.2421,
+      "step": 490
+    },
+    {
+      "epoch": 0.1419151376146789,
+      "grad_norm": 5.025394485398511,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 1.2633,
+      "step": 495
+    },
+    {
+      "epoch": 0.143348623853211,
+      "grad_norm": 5.520994924491881,
+      "learning_rate": 1.988602376778327e-05,
+      "loss": 1.189,
+      "step": 500
+    },
+    {
+      "epoch": 0.14478211009174313,
+      "grad_norm": 4.515316969142533,
+      "learning_rate": 1.9878366296301713e-05,
+      "loss": 1.1607,
+      "step": 505
+    },
+    {
+      "epoch": 0.14621559633027523,
+      "grad_norm": 4.532857835025907,
+      "learning_rate": 1.9870461458059188e-05,
+      "loss": 1.2315,
+      "step": 510
+    },
+    {
+      "epoch": 0.14764908256880735,
+      "grad_norm": 4.001206919662932,
+      "learning_rate": 1.9862309451002827e-05,
+      "loss": 1.1659,
+      "step": 515
+    },
+    {
+      "epoch": 0.14908256880733944,
+      "grad_norm": 4.137717431705149,
+      "learning_rate": 1.9853910479269165e-05,
+      "loss": 1.1732,
+      "step": 520
+    },
+    {
+      "epoch": 0.15051605504587157,
+      "grad_norm": 4.367905503345653,
+      "learning_rate": 1.9845264753179064e-05,
+      "loss": 1.2219,
+      "step": 525
+    },
+    {
+      "epoch": 0.15194954128440366,
+      "grad_norm": 3.942691834220351,
+      "learning_rate": 1.9836372489232416e-05,
+      "loss": 1.1208,
+      "step": 530
+    },
+    {
+      "epoch": 0.15338302752293578,
+      "grad_norm": 4.028647286087348,
+      "learning_rate": 1.982723391010273e-05,
+      "loss": 1.1499,
+      "step": 535
+    },
+    {
+      "epoch": 0.15481651376146788,
+      "grad_norm": 4.262774154684462,
+      "learning_rate": 1.9817849244631575e-05,
+      "loss": 1.2052,
+      "step": 540
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 4.0350996506196175,
+      "learning_rate": 1.9808218727822808e-05,
+      "loss": 1.1899,
+      "step": 545
+    },
+    {
+      "epoch": 0.15768348623853212,
+      "grad_norm": 3.806300412200292,
+      "learning_rate": 1.979834260083673e-05,
+      "loss": 1.2099,
+      "step": 550
+    },
+    {
+      "epoch": 0.15911697247706422,
+      "grad_norm": 4.447162495747635,
+      "learning_rate": 1.9788221110984026e-05,
+      "loss": 1.2123,
+      "step": 555
+    },
+    {
+      "epoch": 0.16055045871559634,
+      "grad_norm": 4.094827599966712,
+      "learning_rate": 1.977785451171958e-05,
+      "loss": 1.1411,
+      "step": 560
+    },
+    {
+      "epoch": 0.16198394495412843,
+      "grad_norm": 4.6429132125696535,
+      "learning_rate": 1.9767243062636122e-05,
+      "loss": 1.1914,
+      "step": 565
+    },
+    {
+      "epoch": 0.16341743119266056,
+      "grad_norm": 3.9726685572488227,
+      "learning_rate": 1.975638702945773e-05,
+      "loss": 1.1758,
+      "step": 570
+    },
+    {
+      "epoch": 0.16485091743119265,
+      "grad_norm": 3.9582277412235367,
+      "learning_rate": 1.974528668403318e-05,
+      "loss": 1.1839,
+      "step": 575
+    },
+    {
+      "epoch": 0.16628440366972477,
+      "grad_norm": 4.611199474059946,
+      "learning_rate": 1.973394230432913e-05,
+      "loss": 1.2155,
+      "step": 580
+    },
+    {
+      "epoch": 0.16771788990825687,
+      "grad_norm": 3.964842824608755,
+      "learning_rate": 1.972235417442317e-05,
+      "loss": 1.168,
+      "step": 585
+    },
+    {
+      "epoch": 0.169151376146789,
+      "grad_norm": 4.083964710145448,
+      "learning_rate": 1.9710522584496695e-05,
+      "loss": 1.2135,
+      "step": 590
+    },
+    {
+      "epoch": 0.1705848623853211,
+      "grad_norm": 3.8905730852414155,
+      "learning_rate": 1.9698447830827655e-05,
+      "loss": 1.1893,
+      "step": 595
+    },
+    {
+      "epoch": 0.1720183486238532,
+      "grad_norm": 4.399950123881084,
+      "learning_rate": 1.9686130215783124e-05,
+      "loss": 1.2244,
+      "step": 600
+    },
+    {
+      "epoch": 0.17345183486238533,
+      "grad_norm": 4.3066835429404495,
+      "learning_rate": 1.967357004781173e-05,
+      "loss": 1.1532,
+      "step": 605
+    },
+    {
+      "epoch": 0.17488532110091742,
+      "grad_norm": 5.071390390632784,
+      "learning_rate": 1.9660767641435926e-05,
+      "loss": 1.2137,
+      "step": 610
+    },
+    {
+      "epoch": 0.17631880733944955,
+      "grad_norm": 3.738266843064544,
+      "learning_rate": 1.964772331724414e-05,
+      "loss": 1.165,
+      "step": 615
+    },
+    {
+      "epoch": 0.17775229357798164,
+      "grad_norm": 3.885352429025116,
+      "learning_rate": 1.9634437401882707e-05,
+      "loss": 1.1899,
+      "step": 620
+    },
+    {
+      "epoch": 0.17918577981651376,
+      "grad_norm": 3.653935761439161,
+      "learning_rate": 1.962091022804772e-05,
+      "loss": 1.1506,
+      "step": 625
+    },
+    {
+      "epoch": 0.18061926605504589,
+      "grad_norm": 4.559227033021131,
+      "learning_rate": 1.960714213447668e-05,
+      "loss": 1.1645,
+      "step": 630
+    },
+    {
+      "epoch": 0.18205275229357798,
+      "grad_norm": 4.618925411728112,
+      "learning_rate": 1.959313346594004e-05,
+      "loss": 1.1802,
+      "step": 635
+    },
+    {
+      "epoch": 0.1834862385321101,
+      "grad_norm": 4.308837236054774,
+      "learning_rate": 1.9578884573232538e-05,
+      "loss": 1.2023,
+      "step": 640
+    },
+    {
+      "epoch": 0.1849197247706422,
+      "grad_norm": 4.304427359268019,
+      "learning_rate": 1.9564395813164428e-05,
+      "loss": 1.2244,
+      "step": 645
+    },
+    {
+      "epoch": 0.18635321100917432,
+      "grad_norm": 3.486413458583046,
+      "learning_rate": 1.9549667548552557e-05,
+      "loss": 1.2111,
+      "step": 650
+    },
+    {
+      "epoch": 0.1877866972477064,
+      "grad_norm": 3.7640427311235403,
+      "learning_rate": 1.9534700148211255e-05,
+      "loss": 1.1956,
+      "step": 655
+    },
+    {
+      "epoch": 0.18922018348623854,
+      "grad_norm": 4.222766727318349,
+      "learning_rate": 1.9519493986943125e-05,
+      "loss": 1.1996,
+      "step": 660
+    },
+    {
+      "epoch": 0.19065366972477063,
+      "grad_norm": 3.7868143111703256,
+      "learning_rate": 1.9504049445529632e-05,
+      "loss": 1.2093,
+      "step": 665
+    },
+    {
+      "epoch": 0.19208715596330275,
+      "grad_norm": 3.579431474524795,
+      "learning_rate": 1.94883669107216e-05,
+      "loss": 1.1785,
+      "step": 670
+    },
+    {
+      "epoch": 0.19352064220183487,
+      "grad_norm": 4.14899759337832,
+      "learning_rate": 1.9472446775229486e-05,
+      "loss": 1.2402,
+      "step": 675
+    },
+    {
+      "epoch": 0.19495412844036697,
+      "grad_norm": 3.7259985492864804,
+      "learning_rate": 1.9456289437713578e-05,
+      "loss": 1.1713,
+      "step": 680
+    },
+    {
+      "epoch": 0.1963876146788991,
+      "grad_norm": 4.525451494816849,
+      "learning_rate": 1.9439895302774007e-05,
+      "loss": 1.2179,
+      "step": 685
+    },
+    {
+      "epoch": 0.1978211009174312,
+      "grad_norm": 3.9639891914703185,
+      "learning_rate": 1.9423264780940602e-05,
+      "loss": 1.2163,
+      "step": 690
+    },
+    {
+      "epoch": 0.1992545871559633,
+      "grad_norm": 3.8210369748297612,
+      "learning_rate": 1.940639828866262e-05,
+      "loss": 1.2462,
+      "step": 695
+    },
+    {
+      "epoch": 0.2006880733944954,
+      "grad_norm": 4.499363334347953,
+      "learning_rate": 1.938929624829832e-05,
+      "loss": 1.2572,
+      "step": 700
+    },
+    {
+      "epoch": 0.20212155963302753,
+      "grad_norm": 32.95893389851377,
+      "learning_rate": 1.937195908810438e-05,
+      "loss": 1.2332,
+      "step": 705
+    },
+    {
+      "epoch": 0.20355504587155962,
+      "grad_norm": 4.056487188350767,
+      "learning_rate": 1.935438724222517e-05,
+      "loss": 1.1797,
+      "step": 710
+    },
+    {
+      "epoch": 0.20498853211009174,
+      "grad_norm": 3.940310991182425,
+      "learning_rate": 1.93365811506819e-05,
+      "loss": 1.2315,
+      "step": 715
+    },
+    {
+      "epoch": 0.20642201834862386,
+      "grad_norm": 3.92517292280482,
+      "learning_rate": 1.9318541259361573e-05,
+      "loss": 1.1824,
+      "step": 720
+    },
+    {
+      "epoch": 0.20785550458715596,
+      "grad_norm": 4.109451546798357,
+      "learning_rate": 1.9300268020005832e-05,
+      "loss": 1.2375,
+      "step": 725
+    },
+    {
+      "epoch": 0.20928899082568808,
+      "grad_norm": 4.168329616635684,
+      "learning_rate": 1.9281761890199666e-05,
+      "loss": 1.1812,
+      "step": 730
+    },
+    {
+      "epoch": 0.21072247706422018,
+      "grad_norm": 3.8618338414153435,
+      "learning_rate": 1.9263023333359918e-05,
+      "loss": 1.1903,
+      "step": 735
+    },
+    {
+      "epoch": 0.2121559633027523,
+      "grad_norm": 4.070643005805358,
+      "learning_rate": 1.9244052818723706e-05,
+      "loss": 1.2539,
+      "step": 740
+    },
+    {
+      "epoch": 0.2135894495412844,
+      "grad_norm": 3.9207051720216004,
+      "learning_rate": 1.9224850821336664e-05,
+      "loss": 1.2375,
+      "step": 745
+    },
+    {
+      "epoch": 0.21502293577981652,
+      "grad_norm": 4.14495928553583,
+      "learning_rate": 1.920541782204104e-05,
+      "loss": 1.253,
+      "step": 750
+    },
+    {
+      "epoch": 0.21645642201834864,
+      "grad_norm": 4.210904171053472,
+      "learning_rate": 1.918575430746367e-05,
+      "loss": 1.1987,
+      "step": 755
+    },
+    {
+      "epoch": 0.21788990825688073,
+      "grad_norm": 3.6674431622452883,
+      "learning_rate": 1.9165860770003774e-05,
+      "loss": 1.2427,
+      "step": 760
+    },
+    {
+      "epoch": 0.21932339449541285,
+      "grad_norm": 3.7503035449224127,
+      "learning_rate": 1.914573770782065e-05,
+      "loss": 1.1564,
+      "step": 765
+    },
+    {
+      "epoch": 0.22075688073394495,
+      "grad_norm": 4.206820549333469,
+      "learning_rate": 1.9125385624821162e-05,
+      "loss": 1.2026,
+      "step": 770
+    },
+    {
+      "epoch": 0.22219036697247707,
+      "grad_norm": 4.079050524248544,
+      "learning_rate": 1.9104805030647164e-05,
+      "loss": 1.2521,
+      "step": 775
+    },
+    {
+      "epoch": 0.22362385321100917,
+      "grad_norm": 3.8439532493920585,
+      "learning_rate": 1.908399644066272e-05,
+      "loss": 1.2247,
+      "step": 780
+    },
+    {
+      "epoch": 0.2250573394495413,
+      "grad_norm": 4.419153034660572,
+      "learning_rate": 1.906296037594117e-05,
+      "loss": 1.2827,
+      "step": 785
+    },
+    {
+      "epoch": 0.22649082568807338,
+      "grad_norm": 4.802038277224569,
+      "learning_rate": 1.904169736325215e-05,
+      "loss": 1.262,
+      "step": 790
+    },
+    {
+      "epoch": 0.2279243119266055,
+      "grad_norm": 4.284188903637777,
+      "learning_rate": 1.9020207935048317e-05,
+      "loss": 1.1532,
+      "step": 795
+    },
+    {
+      "epoch": 0.22935779816513763,
+      "grad_norm": 3.639715866595117,
+      "learning_rate": 1.8998492629452087e-05,
+      "loss": 1.2381,
+      "step": 800
+    },
+    {
+      "epoch": 0.23079128440366972,
+      "grad_norm": 3.741005239883841,
+      "learning_rate": 1.8976551990242122e-05,
+      "loss": 1.216,
+      "step": 805
+    },
+    {
+      "epoch": 0.23222477064220184,
+      "grad_norm": 3.6139511602383276,
+      "learning_rate": 1.895438656683972e-05,
+      "loss": 1.2473,
+      "step": 810
+    },
+    {
+      "epoch": 0.23365825688073394,
+      "grad_norm": 3.8300619382359846,
+      "learning_rate": 1.8931996914295065e-05,
+      "loss": 1.1882,
+      "step": 815
+    },
+    {
+      "epoch": 0.23509174311926606,
+      "grad_norm": 3.943889343109655,
+      "learning_rate": 1.8909383593273317e-05,
+      "loss": 1.2444,
+      "step": 820
+    },
+    {
+      "epoch": 0.23652522935779816,
+      "grad_norm": 4.115538048586424,
+      "learning_rate": 1.8886547170040575e-05,
+      "loss": 1.233,
+      "step": 825
+    },
+    {
+      "epoch": 0.23795871559633028,
+      "grad_norm": 5.787154468595948,
+      "learning_rate": 1.8863488216449702e-05,
+      "loss": 1.2236,
+      "step": 830
+    },
+    {
+      "epoch": 0.23939220183486237,
+      "grad_norm": 3.778052893153268,
+      "learning_rate": 1.8840207309926003e-05,
+      "loss": 1.2286,
+      "step": 835
+    },
+    {
+      "epoch": 0.2408256880733945,
+      "grad_norm": 3.435004162216508,
+      "learning_rate": 1.881670503345277e-05,
+      "loss": 1.1868,
+      "step": 840
+    },
+    {
+      "epoch": 0.24225917431192662,
+      "grad_norm": 5.261326157501783,
+      "learning_rate": 1.879298197555666e-05,
+      "loss": 1.2486,
+      "step": 845
+    },
+    {
+      "epoch": 0.2436926605504587,
+      "grad_norm": 3.651324240819765,
+      "learning_rate": 1.8769038730292993e-05,
+      "loss": 1.2383,
+      "step": 850
+    },
+    {
+      "epoch": 0.24512614678899083,
+      "grad_norm": 4.834151575695704,
+      "learning_rate": 1.8744875897230853e-05,
+      "loss": 1.2592,
+      "step": 855
+    },
+    {
+      "epoch": 0.24655963302752293,
+      "grad_norm": 3.9673796319727606,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 1.2135,
+      "step": 860
+    },
+    {
+      "epoch": 0.24799311926605505,
+      "grad_norm": 3.9495356227550293,
+      "learning_rate": 1.869589389346611e-05,
+      "loss": 1.2443,
+      "step": 865
+    },
+    {
+      "epoch": 0.24942660550458715,
+      "grad_norm": 3.8171283870857917,
+      "learning_rate": 1.8671075949334713e-05,
+      "loss": 1.1502,
+      "step": 870
+    },
+    {
+      "epoch": 0.25086009174311924,
+      "grad_norm": 3.594654406802359,
+      "learning_rate": 1.8646040870516526e-05,
+      "loss": 1.1831,
+      "step": 875
+    },
+    {
+      "epoch": 0.25229357798165136,
+      "grad_norm": 3.5772784131514337,
+      "learning_rate": 1.862078928392153e-05,
+      "loss": 1.2116,
+      "step": 880
+    },
+    {
+      "epoch": 0.2537270642201835,
+      "grad_norm": 4.016497035159711,
+      "learning_rate": 1.8595321821881322e-05,
+      "loss": 1.2156,
+      "step": 885
+    },
+    {
+      "epoch": 0.2551605504587156,
+      "grad_norm": 3.6584048014823622,
+      "learning_rate": 1.8569639122133304e-05,
+      "loss": 1.1687,
+      "step": 890
+    },
+    {
+      "epoch": 0.25659403669724773,
+      "grad_norm": 4.092581816708313,
+      "learning_rate": 1.8543741827804685e-05,
+      "loss": 1.2433,
+      "step": 895
+    },
+    {
+      "epoch": 0.2580275229357798,
+      "grad_norm": 10.965320915487338,
+      "learning_rate": 1.8517630587396413e-05,
+      "loss": 1.2365,
+      "step": 900
+    },
+    {
+      "epoch": 0.2594610091743119,
+      "grad_norm": 3.744425931996878,
+      "learning_rate": 1.8491306054766907e-05,
+      "loss": 1.2315,
+      "step": 905
+    },
+    {
+      "epoch": 0.26089449541284404,
+      "grad_norm": 4.446159019549649,
+      "learning_rate": 1.8464768889115684e-05,
+      "loss": 1.165,
+      "step": 910
+    },
+    {
+      "epoch": 0.26232798165137616,
+      "grad_norm": 3.4122914752905227,
+      "learning_rate": 1.8438019754966877e-05,
+      "loss": 1.1715,
+      "step": 915
+    },
+    {
+      "epoch": 0.26376146788990823,
+      "grad_norm": 3.681090000606899,
+      "learning_rate": 1.841105932215256e-05,
+      "loss": 1.2197,
+      "step": 920
+    },
+    {
+      "epoch": 0.26519495412844035,
+      "grad_norm": 3.723633399780894,
+      "learning_rate": 1.838388826579601e-05,
+      "loss": 1.2308,
+      "step": 925
+    },
+    {
+      "epoch": 0.2666284403669725,
+      "grad_norm": 3.79765489569063,
+      "learning_rate": 1.835650726629477e-05,
+      "loss": 1.2388,
+      "step": 930
+    },
+    {
+      "epoch": 0.2680619266055046,
+      "grad_norm": 3.636387647388923,
+      "learning_rate": 1.8328917009303634e-05,
+      "loss": 1.2296,
+      "step": 935
+    },
+    {
+      "epoch": 0.2694954128440367,
+      "grad_norm": 3.696962989830932,
+      "learning_rate": 1.830111818571745e-05,
+      "loss": 1.2048,
+      "step": 940
+    },
+    {
+      "epoch": 0.2709288990825688,
+      "grad_norm": 3.896582872078181,
+      "learning_rate": 1.8273111491653867e-05,
+      "loss": 1.2522,
+      "step": 945
+    },
+    {
+      "epoch": 0.2723623853211009,
+      "grad_norm": 3.7244583842983454,
+      "learning_rate": 1.824489762843584e-05,
+      "loss": 1.1896,
+      "step": 950
+    },
+    {
+      "epoch": 0.27379587155963303,
+      "grad_norm": 3.66702244239725,
+      "learning_rate": 1.821647730257413e-05,
+      "loss": 1.1777,
+      "step": 955
+    },
+    {
+      "epoch": 0.27522935779816515,
+      "grad_norm": 6.778664388382065,
+      "learning_rate": 1.818785122574956e-05,
+      "loss": 1.2239,
+      "step": 960
+    },
+    {
+      "epoch": 0.2766628440366973,
+      "grad_norm": 3.5798872160414095,
+      "learning_rate": 1.8159020114795226e-05,
+      "loss": 1.1532,
+      "step": 965
+    },
+    {
+      "epoch": 0.27809633027522934,
+      "grad_norm": 4.080407231530825,
+      "learning_rate": 1.8129984691678547e-05,
+      "loss": 1.2046,
+      "step": 970
+    },
+    {
+      "epoch": 0.27952981651376146,
+      "grad_norm": 3.9179230686635234,
+      "learning_rate": 1.8100745683483168e-05,
+      "loss": 1.254,
+      "step": 975
+    },
+    {
+      "epoch": 0.2809633027522936,
+      "grad_norm": 4.97899730667325,
+      "learning_rate": 1.807130382239075e-05,
+      "loss": 1.2084,
+      "step": 980
+    },
+    {
+      "epoch": 0.2823967889908257,
+      "grad_norm": 3.8642747207026242,
+      "learning_rate": 1.8041659845662663e-05,
+      "loss": 1.2014,
+      "step": 985
+    },
+    {
+      "epoch": 0.2838302752293578,
+      "grad_norm": 3.4483181116099964,
+      "learning_rate": 1.8011814495621506e-05,
+      "loss": 1.1435,
+      "step": 990
+    },
+    {
+      "epoch": 0.2852637614678899,
+      "grad_norm": 3.8042720182477434,
+      "learning_rate": 1.798176851963251e-05,
+      "loss": 1.2047,
+      "step": 995
+    },
+    {
+      "epoch": 0.286697247706422,
+      "grad_norm": 3.483685408627929,
+      "learning_rate": 1.7951522670084847e-05,
+      "loss": 1.171,
+      "step": 1000
+    },
+    {
+      "epoch": 0.28813073394495414,
+      "grad_norm": 3.487052716444198,
+      "learning_rate": 1.792107770437276e-05,
+      "loss": 1.2396,
+      "step": 1005
+    },
+    {
+      "epoch": 0.28956422018348627,
+      "grad_norm": 3.553005650055509,
+      "learning_rate": 1.789043438487662e-05,
+      "loss": 1.1149,
+      "step": 1010
+    },
+    {
+      "epoch": 0.29099770642201833,
+      "grad_norm": 3.5693421017081306,
+      "learning_rate": 1.785959347894383e-05,
+      "loss": 1.2387,
+      "step": 1015
+    },
+    {
+      "epoch": 0.29243119266055045,
+      "grad_norm": 3.535948224388696,
+      "learning_rate": 1.7828555758869602e-05,
+      "loss": 1.232,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2938646788990826,
+      "grad_norm": 3.522215138188648,
+      "learning_rate": 1.7797322001877625e-05,
+      "loss": 1.2004,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2952981651376147,
+      "grad_norm": 3.9401114869933362,
+      "learning_rate": 1.7765892990100593e-05,
+      "loss": 1.1954,
+      "step": 1030
+    },
+    {
+      "epoch": 0.29673165137614677,
+      "grad_norm": 3.9452874322873424,
+      "learning_rate": 1.773426951056064e-05,
+      "loss": 1.1752,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2981651376146789,
+      "grad_norm": 3.77900474922427,
+      "learning_rate": 1.7702452355149606e-05,
+      "loss": 1.2023,
+      "step": 1040
+    },
+    {
+      "epoch": 0.299598623853211,
+      "grad_norm": 4.194810233115317,
+      "learning_rate": 1.7670442320609226e-05,
+      "loss": 1.1762,
+      "step": 1045
+    },
+    {
+      "epoch": 0.30103211009174313,
+      "grad_norm": 3.6317516711250524,
+      "learning_rate": 1.7638240208511162e-05,
+      "loss": 1.2036,
+      "step": 1050
+    },
+    {
+      "epoch": 0.30246559633027525,
+      "grad_norm": 3.613998915069475,
+      "learning_rate": 1.760584682523696e-05,
+      "loss": 1.1794,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3038990825688073,
+      "grad_norm": 3.168795755449482,
+      "learning_rate": 1.7573262981957814e-05,
+      "loss": 1.1812,
+      "step": 1060
+    },
+    {
+      "epoch": 0.30533256880733944,
+      "grad_norm": 3.883990610964808,
+      "learning_rate": 1.7540489494614294e-05,
+      "loss": 1.2273,
+      "step": 1065
+    },
+    {
+      "epoch": 0.30676605504587157,
+      "grad_norm": 4.068560981758666,
+      "learning_rate": 1.7507527183895893e-05,
+      "loss": 1.2297,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3081995412844037,
+      "grad_norm": 3.812795569574441,
+      "learning_rate": 1.747437687522047e-05,
+      "loss": 1.2037,
+      "step": 1075
+    },
+    {
+      "epoch": 0.30963302752293576,
+      "grad_norm": 3.7886870743963077,
+      "learning_rate": 1.744103939871361e-05,
+      "loss": 1.2235,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3110665137614679,
+      "grad_norm": 3.3064001495116555,
+      "learning_rate": 1.7407515589187793e-05,
+      "loss": 1.213,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 3.999062074927365,
+      "learning_rate": 1.7373806286121532e-05,
+      "loss": 1.2586,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3139334862385321,
+      "grad_norm": 4.371331122546795,
+      "learning_rate": 1.7339912333638322e-05,
+      "loss": 1.1731,
+      "step": 1095
+    },
+    {
+      "epoch": 0.31536697247706424,
+      "grad_norm": 3.8167405663507,
+      "learning_rate": 1.730583458048552e-05,
+      "loss": 1.2098,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3168004587155963,
+      "grad_norm": 3.398825843337014,
+      "learning_rate": 1.727157388001307e-05,
+      "loss": 1.187,
+      "step": 1105
+    },
+    {
+      "epoch": 0.31823394495412843,
+      "grad_norm": 3.323984794217734,
+      "learning_rate": 1.723713109015217e-05,
+      "loss": 1.229,
+      "step": 1110
+    },
+    {
+      "epoch": 0.31966743119266056,
+      "grad_norm": 3.388696780319304,
+      "learning_rate": 1.720250707339374e-05,
+      "loss": 1.1616,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3211009174311927,
+      "grad_norm": 3.7284978608483934,
+      "learning_rate": 1.7167702696766877e-05,
+      "loss": 1.173,
+      "step": 1120
+    },
+    {
+      "epoch": 0.32253440366972475,
+      "grad_norm": 4.037999124478957,
+      "learning_rate": 1.7132718831817093e-05,
+      "loss": 1.1695,
+      "step": 1125
+    },
+    {
+      "epoch": 0.32396788990825687,
+      "grad_norm": 3.2060752585491086,
+      "learning_rate": 1.7097556354584526e-05,
+      "loss": 1.1464,
+      "step": 1130
+    },
+    {
+      "epoch": 0.325401376146789,
+      "grad_norm": 3.72775795176933,
+      "learning_rate": 1.7062216145581997e-05,
+      "loss": 1.1237,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3268348623853211,
+      "grad_norm": 3.783961044208216,
+      "learning_rate": 1.7026699089772937e-05,
+      "loss": 1.1899,
+      "step": 1140
+    },
+    {
+      "epoch": 0.32826834862385323,
+      "grad_norm": 3.5936905948051323,
+      "learning_rate": 1.699100607654926e-05,
+      "loss": 1.176,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3297018348623853,
+      "grad_norm": 3.5296084880895156,
+      "learning_rate": 1.6955137999709075e-05,
+      "loss": 1.1445,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3311353211009174,
+      "grad_norm": 3.278850491610551,
+      "learning_rate": 1.6919095757434288e-05,
+      "loss": 1.2269,
+      "step": 1155
+    },
+    {
+      "epoch": 0.33256880733944955,
+      "grad_norm": 3.6404322855144247,
+      "learning_rate": 1.6882880252268156e-05,
+      "loss": 1.1836,
+      "step": 1160
+    },
+    {
+      "epoch": 0.33400229357798167,
+      "grad_norm": 3.437933045156408,
+      "learning_rate": 1.6846492391092625e-05,
+      "loss": 1.2295,
+      "step": 1165
+    },
+    {
+      "epoch": 0.33543577981651373,
+      "grad_norm": 3.5611595737623167,
+      "learning_rate": 1.680993308510568e-05,
+      "loss": 1.1862,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33686926605504586,
+      "grad_norm": 4.312054532006292,
+      "learning_rate": 1.6773203249798482e-05,
+      "loss": 1.1602,
+      "step": 1175
+    },
+    {
+      "epoch": 0.338302752293578,
+      "grad_norm": 3.598611274354031,
+      "learning_rate": 1.6736303804932475e-05,
+      "loss": 1.1825,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3397362385321101,
+      "grad_norm": 3.779933724951351,
+      "learning_rate": 1.6699235674516334e-05,
+      "loss": 1.1847,
+      "step": 1185
+    },
+    {
+      "epoch": 0.3411697247706422,
+      "grad_norm": 3.590038112136874,
+      "learning_rate": 1.666199978678283e-05,
+      "loss": 1.1984,
+      "step": 1190
+    },
+    {
+      "epoch": 0.3426032110091743,
+      "grad_norm": 4.110223116861788,
+      "learning_rate": 1.6624597074165597e-05,
+      "loss": 1.2468,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3440366972477064,
+      "grad_norm": 3.603557931052651,
+      "learning_rate": 1.6587028473275772e-05,
+      "loss": 1.1962,
+      "step": 1200
+    },
+    {
+      "epoch": 0.34547018348623854,
+      "grad_norm": 3.209591671498054,
+      "learning_rate": 1.6549294924878532e-05,
+      "loss": 1.1714,
+      "step": 1205
+    },
+    {
+      "epoch": 0.34690366972477066,
+      "grad_norm": 3.6243010245754337,
+      "learning_rate": 1.651139737386957e-05,
+      "loss": 1.1733,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3483371559633027,
+      "grad_norm": 3.9078886802263697,
+      "learning_rate": 1.6473336769251388e-05,
+      "loss": 1.1715,
+      "step": 1215
+    },
+    {
+      "epoch": 0.34977064220183485,
+      "grad_norm": 3.6344742284976306,
+      "learning_rate": 1.6435114064109575e-05,
+      "loss": 1.1244,
+      "step": 1220
+    },
+    {
+      "epoch": 0.35120412844036697,
+      "grad_norm": 3.6952536271050964,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 1.2089,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3526376146788991,
+      "grad_norm": 3.429886933846167,
+      "learning_rate": 1.6358186184869417e-05,
+      "loss": 1.1892,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3540711009174312,
+      "grad_norm": 3.093455767598654,
+      "learning_rate": 1.631948293714227e-05,
+      "loss": 1.2106,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3555045871559633,
+      "grad_norm": 4.038152481713509,
+      "learning_rate": 1.6280621441585647e-05,
+      "loss": 1.1697,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3569380733944954,
+      "grad_norm": 3.6999557746923095,
+      "learning_rate": 1.6241602671340448e-05,
+      "loss": 1.2376,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3583715596330275,
+      "grad_norm": 3.6041781324392868,
+      "learning_rate": 1.6202427603485933e-05,
+      "loss": 1.1327,
+      "step": 1250
+    },
+    {
+      "epoch": 0.35980504587155965,
+      "grad_norm": 3.4145472213007473,
+      "learning_rate": 1.6163097219015245e-05,
+      "loss": 1.1372,
+      "step": 1255
+    },
+    {
+      "epoch": 0.36123853211009177,
+      "grad_norm": 3.429021637188865,
+      "learning_rate": 1.6123612502810865e-05,
+      "loss": 1.1646,
+      "step": 1260
+    },
+    {
+      "epoch": 0.36267201834862384,
+      "grad_norm": 3.6122171856293708,
+      "learning_rate": 1.6083974443619922e-05,
+      "loss": 1.2031,
+      "step": 1265
+    },
+    {
+      "epoch": 0.36410550458715596,
+      "grad_norm": 3.2252592769247617,
+      "learning_rate": 1.6044184034029445e-05,
+      "loss": 1.1669,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3655389908256881,
+      "grad_norm": 3.513524941132868,
+      "learning_rate": 1.6004242270441523e-05,
+      "loss": 1.1871,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3669724770642202,
+      "grad_norm": 3.5159805187675115,
+      "learning_rate": 1.596415015304833e-05,
+      "loss": 1.1851,
+      "step": 1280
+    },
+    {
+      "epoch": 0.36840596330275227,
+      "grad_norm": 3.3054440445349837,
+      "learning_rate": 1.5923908685807087e-05,
+      "loss": 1.1074,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3698394495412844,
+      "grad_norm": 3.6946632308744136,
+      "learning_rate": 1.588351887641494e-05,
+      "loss": 1.1596,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3712729357798165,
+      "grad_norm": 3.3924561467755785,
+      "learning_rate": 1.5842981736283686e-05,
+      "loss": 1.2262,
+      "step": 1295
+    },
+    {
+      "epoch": 0.37270642201834864,
+      "grad_norm": 3.623083608828368,
+      "learning_rate": 1.5802298280514487e-05,
+      "loss": 1.1823,
+      "step": 1300
+    },
+    {
+      "epoch": 0.37413990825688076,
+      "grad_norm": 3.0148932024272437,
+      "learning_rate": 1.5761469527872427e-05,
+      "loss": 1.1513,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3755733944954128,
+      "grad_norm": 10.198288104380973,
+      "learning_rate": 1.572049650076101e-05,
+      "loss": 1.1498,
+      "step": 1310
+    },
+    {
+      "epoch": 0.37700688073394495,
+      "grad_norm": 3.623884808969302,
+      "learning_rate": 1.5679380225196546e-05,
+      "loss": 1.1857,
+      "step": 1315
+    },
+    {
+      "epoch": 0.37844036697247707,
+      "grad_norm": 21.814389727799938,
+      "learning_rate": 1.5638121730782486e-05,
+      "loss": 1.1748,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3798738532110092,
+      "grad_norm": 3.7586731538464684,
+      "learning_rate": 1.5596722050683598e-05,
+      "loss": 1.2024,
+      "step": 1325
+    },
+    {
+      "epoch": 0.38130733944954126,
+      "grad_norm": 3.914930197959616,
+      "learning_rate": 1.555518222160013e-05,
+      "loss": 1.1873,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3827408256880734,
+      "grad_norm": 3.645421580899915,
+      "learning_rate": 1.551350328374184e-05,
+      "loss": 1.1252,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3841743119266055,
+      "grad_norm": 3.1973889011870376,
+      "learning_rate": 1.5471686280801933e-05,
+      "loss": 1.1839,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3856077981651376,
+      "grad_norm": 3.5501087765790915,
+      "learning_rate": 1.5429732259930955e-05,
+      "loss": 1.1232,
+      "step": 1345
+    },
+    {
+      "epoch": 0.38704128440366975,
+      "grad_norm": 3.3946063622500016,
+      "learning_rate": 1.538764227171054e-05,
+      "loss": 1.1768,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3884747706422018,
+      "grad_norm": 3.3089407247339975,
+      "learning_rate": 1.5345417370127123e-05,
+      "loss": 1.1697,
+      "step": 1355
+    },
+    {
+      "epoch": 0.38990825688073394,
+      "grad_norm": 4.435494305269373,
+      "learning_rate": 1.5303058612545534e-05,
+      "loss": 1.1301,
+      "step": 1360
+    },
+    {
+      "epoch": 0.39134174311926606,
+      "grad_norm": 3.1427532195841343,
+      "learning_rate": 1.5260567059682535e-05,
+      "loss": 1.1857,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3927752293577982,
+      "grad_norm": 3.5070367063445684,
+      "learning_rate": 1.521794377558024e-05,
+      "loss": 1.1294,
+      "step": 1370
+    },
+    {
+      "epoch": 0.39420871559633025,
+      "grad_norm": 3.6077040833104896,
+      "learning_rate": 1.5175189827579489e-05,
+      "loss": 1.1469,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3956422018348624,
+      "grad_norm": 3.401651617393664,
+      "learning_rate": 1.5132306286293096e-05,
+      "loss": 1.1383,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3970756880733945,
+      "grad_norm": 3.3826428424139925,
+      "learning_rate": 1.5089294225579077e-05,
+      "loss": 1.1887,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3985091743119266,
+      "grad_norm": 3.2830968866458345,
+      "learning_rate": 1.5046154722513718e-05,
+      "loss": 1.15,
+      "step": 1390
+    },
+    {
+      "epoch": 0.39994266055045874,
+      "grad_norm": 3.2832815737637113,
+      "learning_rate": 1.5002888857364624e-05,
+      "loss": 1.1811,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4013761467889908,
+      "grad_norm": 3.222829419424945,
+      "learning_rate": 1.4959497713563677e-05,
+      "loss": 1.1696,
+      "step": 1400
+    },
+    {
+      "epoch": 0.40280963302752293,
+      "grad_norm": 3.739181624579295,
+      "learning_rate": 1.4915982377679885e-05,
+      "loss": 1.1978,
+      "step": 1405
+    },
+    {
+      "epoch": 0.40424311926605505,
+      "grad_norm": 3.383948957829039,
+      "learning_rate": 1.4872343939392189e-05,
+      "loss": 1.1724,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4056766055045872,
+      "grad_norm": 3.7382791698251427,
+      "learning_rate": 1.482858349146216e-05,
+      "loss": 1.1471,
+      "step": 1415
+    },
+    {
+      "epoch": 0.40711009174311924,
+      "grad_norm": 3.510878161789182,
+      "learning_rate": 1.4784702129706655e-05,
+      "loss": 1.1337,
+      "step": 1420
+    },
+    {
+      "epoch": 0.40854357798165136,
+      "grad_norm": 3.505598802015759,
+      "learning_rate": 1.474070095297036e-05,
+      "loss": 1.187,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4099770642201835,
+      "grad_norm": 3.2069715268799204,
+      "learning_rate": 1.469658106309828e-05,
+      "loss": 1.1136,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4114105504587156,
+      "grad_norm": 4.1812266615604345,
+      "learning_rate": 1.465234356490815e-05,
+      "loss": 1.1718,
+      "step": 1435
+    },
+    {
+      "epoch": 0.41284403669724773,
+      "grad_norm": 3.3985858466934338,
+      "learning_rate": 1.4607989566162761e-05,
+      "loss": 1.1414,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4142775229357798,
+      "grad_norm": 3.6381472550431093,
+      "learning_rate": 1.4563520177542226e-05,
+      "loss": 1.2166,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4157110091743119,
+      "grad_norm": 3.268767608874384,
+      "learning_rate": 1.451893651261617e-05,
+      "loss": 1.1695,
+      "step": 1450
+    },
+    {
+      "epoch": 0.41714449541284404,
+      "grad_norm": 4.519674348990799,
+      "learning_rate": 1.4474239687815838e-05,
+      "loss": 1.189,
+      "step": 1455
+    },
+    {
+      "epoch": 0.41857798165137616,
+      "grad_norm": 3.594106437257434,
+      "learning_rate": 1.4429430822406138e-05,
+      "loss": 1.1956,
+      "step": 1460
+    },
+    {
+      "epoch": 0.42001146788990823,
+      "grad_norm": 3.838114634468203,
+      "learning_rate": 1.4384511038457624e-05,
+      "loss": 1.1348,
+      "step": 1465
+    },
+    {
+      "epoch": 0.42144495412844035,
+      "grad_norm": 3.3721934098468083,
+      "learning_rate": 1.4339481460818385e-05,
+      "loss": 1.135,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4228784403669725,
+      "grad_norm": 3.3513743473211512,
+      "learning_rate": 1.429434321708588e-05,
+      "loss": 1.1443,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4243119266055046,
+      "grad_norm": 3.3279635877417304,
+      "learning_rate": 1.4249097437578712e-05,
+      "loss": 1.1574,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4257454128440367,
+      "grad_norm": 3.5340814143703896,
+      "learning_rate": 1.4203745255308306e-05,
+      "loss": 1.1607,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4271788990825688,
+      "grad_norm": 3.3296679892148995,
+      "learning_rate": 1.4158287805950557e-05,
+      "loss": 1.199,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4286123853211009,
+      "grad_norm": 3.401879333323161,
+      "learning_rate": 1.411272622781737e-05,
+      "loss": 1.1364,
+      "step": 1495
+    },
+    {
+      "epoch": 0.43004587155963303,
+      "grad_norm": 3.131916232702695,
+      "learning_rate": 1.4067061661828176e-05,
+      "loss": 1.1472,
+      "step": 1500
+    },
+    {
+      "epoch": 0.43147935779816515,
+      "grad_norm": 3.543349954734912,
+      "learning_rate": 1.4021295251481347e-05,
+      "loss": 1.1813,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4329128440366973,
+      "grad_norm": 3.309506087624542,
+      "learning_rate": 1.3975428142825562e-05,
+      "loss": 1.1581,
+      "step": 1510
+    },
+    {
+      "epoch": 0.43434633027522934,
+      "grad_norm": 3.9281940551258097,
+      "learning_rate": 1.392946148443112e-05,
+      "loss": 1.1808,
+      "step": 1515
+    },
+    {
+      "epoch": 0.43577981651376146,
+      "grad_norm": 3.8214812468244346,
+      "learning_rate": 1.3883396427361169e-05,
+      "loss": 1.1616,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4372133027522936,
+      "grad_norm": 3.997270424750147,
+      "learning_rate": 1.383723412514288e-05,
+      "loss": 1.1113,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4386467889908257,
+      "grad_norm": 3.437008178425516,
+      "learning_rate": 1.3790975733738576e-05,
+      "loss": 1.1713,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4400802752293578,
+      "grad_norm": 3.3811402901395984,
+      "learning_rate": 1.3744622411516758e-05,
+      "loss": 1.1347,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4415137614678899,
+      "grad_norm": 3.3962664452333993,
+      "learning_rate": 1.3698175319223133e-05,
+      "loss": 1.1347,
+      "step": 1540
+    },
+    {
+      "epoch": 0.442947247706422,
+      "grad_norm": 3.6162635705687127,
+      "learning_rate": 1.3651635619951509e-05,
+      "loss": 1.1574,
+      "step": 1545
+    },
+    {
+      "epoch": 0.44438073394495414,
+      "grad_norm": 3.502705575568886,
+      "learning_rate": 1.360500447911471e-05,
+      "loss": 1.0691,
+      "step": 1550
+    },
+    {
+      "epoch": 0.44581422018348627,
+      "grad_norm": 3.473102068128584,
+      "learning_rate": 1.3558283064415357e-05,
+      "loss": 1.1018,
+      "step": 1555
+    },
+    {
+      "epoch": 0.44724770642201833,
+      "grad_norm": 3.5587976176438656,
+      "learning_rate": 1.3511472545816648e-05,
+      "loss": 1.1293,
+      "step": 1560
+    },
+    {
+      "epoch": 0.44868119266055045,
+      "grad_norm": 3.613085860404909,
+      "learning_rate": 1.3464574095513058e-05,
+      "loss": 1.1685,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4501146788990826,
+      "grad_norm": 4.06622018524846,
+      "learning_rate": 1.3417588887900978e-05,
+      "loss": 1.1067,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4515481651376147,
+      "grad_norm": 3.4649254770088698,
+      "learning_rate": 1.3370518099549315e-05,
+      "loss": 1.1026,
+      "step": 1575
+    },
+    {
+      "epoch": 0.45298165137614677,
+      "grad_norm": 2.952840725725241,
+      "learning_rate": 1.3323362909170018e-05,
+      "loss": 1.1283,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4544151376146789,
+      "grad_norm": 3.253136251920175,
+      "learning_rate": 1.3276124497588585e-05,
+      "loss": 1.2414,
+      "step": 1585
+    },
+    {
+      "epoch": 0.455848623853211,
+      "grad_norm": 3.162846986211554,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 1.1169,
+      "step": 1590
+    },
+    {
+      "epoch": 0.45728211009174313,
+      "grad_norm": 3.3272179350821505,
+      "learning_rate": 1.3181402744511446e-05,
+      "loss": 1.1756,
+      "step": 1595
+    },
+    {
+      "epoch": 0.45871559633027525,
+      "grad_norm": 3.1731644889698183,
+      "learning_rate": 1.3133921774968001e-05,
+      "loss": 1.1305,
+      "step": 1600
+    },
+    {
+      "epoch": 0.4601490825688073,
+      "grad_norm": 3.1249681327250904,
+      "learning_rate": 1.3086362328067536e-05,
+      "loss": 1.1603,
+      "step": 1605
+    },
+    {
+      "epoch": 0.46158256880733944,
+      "grad_norm": 3.376924777276381,
+      "learning_rate": 1.3038725594758632e-05,
+      "loss": 1.118,
+      "step": 1610
+    },
+    {
+      "epoch": 0.46301605504587157,
+      "grad_norm": 3.193633263846461,
+      "learning_rate": 1.2991012767925224e-05,
+      "loss": 1.1337,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4644495412844037,
+      "grad_norm": 3.0745547065753915,
+      "learning_rate": 1.2943225042356714e-05,
+      "loss": 1.1252,
+      "step": 1620
+    },
+    {
+      "epoch": 0.46588302752293576,
+      "grad_norm": 7.556185912909887,
+      "learning_rate": 1.2895363614718082e-05,
+      "loss": 1.1517,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4673165137614679,
+      "grad_norm": 3.3618721338615787,
+      "learning_rate": 1.2847429683519879e-05,
+      "loss": 1.1321,
+      "step": 1630
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 3.1654091942580904,
+      "learning_rate": 1.2799424449088246e-05,
+      "loss": 1.1475,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4701834862385321,
+      "grad_norm": 3.3736619152907386,
+      "learning_rate": 1.2751349113534856e-05,
+      "loss": 1.1297,
+      "step": 1640
+    },
+    {
+      "epoch": 0.47161697247706424,
+      "grad_norm": 3.711148861718645,
+      "learning_rate": 1.2703204880726788e-05,
+      "loss": 1.1654,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4730504587155963,
+      "grad_norm": 3.0399296608693946,
+      "learning_rate": 1.2654992956256397e-05,
+      "loss": 1.1072,
+      "step": 1650
+    },
+    {
+      "epoch": 0.47448394495412843,
+      "grad_norm": 3.299482801946242,
+      "learning_rate": 1.2606714547411138e-05,
+      "loss": 1.1338,
+      "step": 1655
+    },
+    {
+      "epoch": 0.47591743119266056,
+      "grad_norm": 4.109998693762478,
+      "learning_rate": 1.2558370863143298e-05,
+      "loss": 1.1517,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4773509174311927,
+      "grad_norm": 3.6028647178766167,
+      "learning_rate": 1.250996311403976e-05,
+      "loss": 1.2193,
+      "step": 1665
+    },
+    {
+      "epoch": 0.47878440366972475,
+      "grad_norm": 3.088463522622337,
+      "learning_rate": 1.246149251229166e-05,
+      "loss": 1.1481,
+      "step": 1670
+    },
+    {
+      "epoch": 0.48021788990825687,
+      "grad_norm": 3.1966754732870033,
+      "learning_rate": 1.2412960271664046e-05,
+      "loss": 1.0923,
+      "step": 1675
+    },
+    {
+      "epoch": 0.481651376146789,
+      "grad_norm": 3.5520419027791044,
+      "learning_rate": 1.2364367607465483e-05,
+      "loss": 1.131,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4830848623853211,
+      "grad_norm": 3.1911366000417103,
+      "learning_rate": 1.2315715736517624e-05,
+      "loss": 1.1524,
+      "step": 1685
+    },
+    {
+      "epoch": 0.48451834862385323,
+      "grad_norm": 3.1232221206590847,
+      "learning_rate": 1.2267005877124721e-05,
+      "loss": 1.1336,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4859518348623853,
+      "grad_norm": 3.1919229672837606,
+      "learning_rate": 1.2218239249043143e-05,
+      "loss": 1.1163,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4873853211009174,
+      "grad_norm": 3.6466393865332223,
+      "learning_rate": 1.2169417073450805e-05,
+      "loss": 1.1113,
+      "step": 1700
+    },
+    {
+      "epoch": 0.48881880733944955,
+      "grad_norm": 3.275240485720321,
+      "learning_rate": 1.2120540572916617e-05,
+      "loss": 1.1516,
+      "step": 1705
+    },
+    {
+      "epoch": 0.49025229357798167,
+      "grad_norm": 3.350383352712665,
+      "learning_rate": 1.2071610971369842e-05,
+      "loss": 1.1564,
+      "step": 1710
+    },
+    {
+      "epoch": 0.49168577981651373,
+      "grad_norm": 3.5074589174762236,
+      "learning_rate": 1.2022629494069466e-05,
+      "loss": 1.1004,
+      "step": 1715
+    },
+    {
+      "epoch": 0.49311926605504586,
+      "grad_norm": 3.4164962172597035,
+      "learning_rate": 1.1973597367573509e-05,
+      "loss": 1.0966,
+      "step": 1720
+    },
+    {
+      "epoch": 0.494552752293578,
+      "grad_norm": 3.3927103725177825,
+      "learning_rate": 1.19245158197083e-05,
+      "loss": 1.1635,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4959862385321101,
+      "grad_norm": 2.9590270815185193,
+      "learning_rate": 1.1875386079537762e-05,
+      "loss": 1.1415,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4974197247706422,
+      "grad_norm": 3.257588158306314,
+      "learning_rate": 1.1826209377332593e-05,
+      "loss": 1.1326,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4988532110091743,
+      "grad_norm": 3.3051541372419737,
+      "learning_rate": 1.1776986944539498e-05,
+      "loss": 1.1278,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5002866972477065,
+      "grad_norm": 3.523346784327648,
+      "learning_rate": 1.1727720013750319e-05,
+      "loss": 1.1099,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5017201834862385,
+      "grad_norm": 3.3893012078265015,
+      "learning_rate": 1.1678409818671192e-05,
+      "loss": 1.1546,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5031536697247706,
+      "grad_norm": 3.3226857594304713,
+      "learning_rate": 1.1629057594091639e-05,
+      "loss": 1.1061,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5045871559633027,
+      "grad_norm": 3.507323841567208,
+      "learning_rate": 1.1579664575853667e-05,
+      "loss": 1.1551,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5060206422018348,
+      "grad_norm": 3.1700645014865856,
+      "learning_rate": 1.1530232000820791e-05,
+      "loss": 1.1151,
+      "step": 1765
+    },
+    {
+      "epoch": 0.507454128440367,
+      "grad_norm": 3.397843452612661,
+      "learning_rate": 1.1480761106847088e-05,
+      "loss": 1.0967,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5088876146788991,
+      "grad_norm": 3.296593763189619,
+      "learning_rate": 1.1431253132746187e-05,
+      "loss": 1.1229,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5103211009174312,
+      "grad_norm": 3.1860797359537574,
+      "learning_rate": 1.138170931826025e-05,
+      "loss": 1.1305,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5117545871559633,
+      "grad_norm": 3.2917554463603333,
+      "learning_rate": 1.133213090402893e-05,
+      "loss": 1.0911,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5131880733944955,
+      "grad_norm": 3.158139008213761,
+      "learning_rate": 1.1282519131558302e-05,
+      "loss": 1.1095,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5146215596330275,
+      "grad_norm": 3.22490301605298,
+      "learning_rate": 1.1232875243189765e-05,
+      "loss": 1.1695,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5160550458715596,
+      "grad_norm": 3.586519697086409,
+      "learning_rate": 1.1183200482068949e-05,
+      "loss": 1.0989,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5174885321100917,
+      "grad_norm": 3.265891742392178,
+      "learning_rate": 1.1133496092114576e-05,
+      "loss": 1.0851,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5189220183486238,
+      "grad_norm": 3.1099004425125334,
+      "learning_rate": 1.1083763317987304e-05,
+      "loss": 1.1119,
+      "step": 1810
+    },
+    {
+      "epoch": 0.520355504587156,
+      "grad_norm": 4.5404304007913305,
+      "learning_rate": 1.103400340505858e-05,
+      "loss": 1.1143,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5217889908256881,
+      "grad_norm": 3.4230865219558764,
+      "learning_rate": 1.0984217599379425e-05,
+      "loss": 1.1182,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5232224770642202,
+      "grad_norm": 3.1427663664508456,
+      "learning_rate": 1.093440714764926e-05,
+      "loss": 1.1317,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5246559633027523,
+      "grad_norm": 2.9784943507834476,
+      "learning_rate": 1.088457329718467e-05,
+      "loss": 1.0364,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5260894495412844,
+      "grad_norm": 3.3656116327984806,
+      "learning_rate": 1.0834717295888168e-05,
+      "loss": 1.1432,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5275229357798165,
+      "grad_norm": 2.8639878500754774,
+      "learning_rate": 1.0784840392216961e-05,
+      "loss": 1.1421,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5289564220183486,
+      "grad_norm": 3.340151227567289,
+      "learning_rate": 1.0734943835151674e-05,
+      "loss": 1.0931,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5303899082568807,
+      "grad_norm": 3.8035180968079993,
+      "learning_rate": 1.0685028874165075e-05,
+      "loss": 1.1179,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5318233944954128,
+      "grad_norm": 3.070748628370571,
+      "learning_rate": 1.0635096759190792e-05,
+      "loss": 1.123,
+      "step": 1855
+    },
+    {
+      "epoch": 0.533256880733945,
+      "grad_norm": 3.226775333715127,
+      "learning_rate": 1.0585148740592013e-05,
+      "loss": 1.1159,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5346903669724771,
+      "grad_norm": 3.236504164334084,
+      "learning_rate": 1.053518606913017e-05,
+      "loss": 1.1494,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5361238532110092,
+      "grad_norm": 3.2302791306950702,
+      "learning_rate": 1.048520999593362e-05,
+      "loss": 1.1444,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5375573394495413,
+      "grad_norm": 3.2832052841745916,
+      "learning_rate": 1.0435221772466318e-05,
+      "loss": 1.1469,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5389908256880734,
+      "grad_norm": 3.2524360195076496,
+      "learning_rate": 1.0385222650496479e-05,
+      "loss": 1.1313,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5404243119266054,
+      "grad_norm": 2.919300243143409,
+      "learning_rate": 1.0335213882065225e-05,
+      "loss": 1.1112,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5418577981651376,
+      "grad_norm": 3.7356105416504546,
+      "learning_rate": 1.0285196719455242e-05,
+      "loss": 1.1113,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5432912844036697,
+      "grad_norm": 3.1163300953894306,
+      "learning_rate": 1.0235172415159418e-05,
+      "loss": 1.0671,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5447247706422018,
+      "grad_norm": 3.6338243642002976,
+      "learning_rate": 1.0185142221849469e-05,
+      "loss": 1.1405,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5461582568807339,
+      "grad_norm": 3.4265846748133613,
+      "learning_rate": 1.0135107392344594e-05,
+      "loss": 1.1052,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5475917431192661,
+      "grad_norm": 3.6433669703963774,
+      "learning_rate": 1.0085069179580076e-05,
+      "loss": 1.0958,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5490252293577982,
+      "grad_norm": 3.070116617818365,
+      "learning_rate": 1.0035028836575922e-05,
+      "loss": 1.0856,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5504587155963303,
+      "grad_norm": 3.3094892848191373,
+      "learning_rate": 9.984987616405486e-06,
+      "loss": 1.1775,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5518922018348624,
+      "grad_norm": 3.1253255381505465,
+      "learning_rate": 9.934946772164082e-06,
+      "loss": 1.0962,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5533256880733946,
+      "grad_norm": 3.2006466765888377,
+      "learning_rate": 9.884907556937619e-06,
+      "loss": 1.0732,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5547591743119266,
+      "grad_norm": 3.1292256633487647,
+      "learning_rate": 9.834871223771204e-06,
+      "loss": 1.0887,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5561926605504587,
+      "grad_norm": 3.1916225285491886,
+      "learning_rate": 9.78483902563778e-06,
+      "loss": 1.1239,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5576261467889908,
+      "grad_norm": 3.0332428019262623,
+      "learning_rate": 9.73481221540674e-06,
+      "loss": 1.1231,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5590596330275229,
+      "grad_norm": 3.045815826748736,
+      "learning_rate": 9.684792045812555e-06,
+      "loss": 1.1098,
+      "step": 1950
+    },
+    {
+      "epoch": 0.560493119266055,
+      "grad_norm": 3.0391892370593334,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 1.1044,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5619266055045872,
+      "grad_norm": 3.055569668736528,
+      "learning_rate": 9.584776638609841e-06,
+      "loss": 1.0894,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5633600917431193,
+      "grad_norm": 3.5406439314630167,
+      "learning_rate": 9.534783905513355e-06,
+      "loss": 1.1109,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5647935779816514,
+      "grad_norm": 3.345905555913882,
+      "learning_rate": 9.484802822015087e-06,
+      "loss": 1.1138,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5662270642201835,
+      "grad_norm": 3.296141105361695,
+      "learning_rate": 9.434834639704464e-06,
+      "loss": 1.1059,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5676605504587156,
+      "grad_norm": 2.8031036854561697,
+      "learning_rate": 9.384880609847838e-06,
+      "loss": 1.0806,
+      "step": 1980
+    },
+    {
+      "epoch": 0.5690940366972477,
+      "grad_norm": 3.0977315843393165,
+      "learning_rate": 9.33494198335717e-06,
+      "loss": 1.1364,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5705275229357798,
+      "grad_norm": 3.3929218329827604,
+      "learning_rate": 9.285020010758706e-06,
+      "loss": 1.1644,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5719610091743119,
+      "grad_norm": 3.2746802917151396,
+      "learning_rate": 9.235115942161656e-06,
+      "loss": 1.0867,
+      "step": 1995
+    },
+    {
+      "epoch": 0.573394495412844,
+      "grad_norm": 3.4111204084118594,
+      "learning_rate": 9.18523102722688e-06,
+      "loss": 1.1372,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5748279816513762,
+      "grad_norm": 3.0339025073036883,
+      "learning_rate": 9.135366515135617e-06,
+      "loss": 1.1156,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5762614678899083,
+      "grad_norm": 3.152200571534001,
+      "learning_rate": 9.0855236545582e-06,
+      "loss": 1.2007,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5776949541284404,
+      "grad_norm": 5.672098040763039,
+      "learning_rate": 9.035703693622762e-06,
+      "loss": 1.1011,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5791284403669725,
+      "grad_norm": 3.0337007868840207,
+      "learning_rate": 8.985907879884011e-06,
+      "loss": 1.0502,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5805619266055045,
+      "grad_norm": 3.397456497600358,
+      "learning_rate": 8.936137460291985e-06,
+      "loss": 1.0686,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5819954128440367,
+      "grad_norm": 3.1557825825649983,
+      "learning_rate": 8.886393681160804e-06,
+      "loss": 1.1242,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5834288990825688,
+      "grad_norm": 3.014868609795805,
+      "learning_rate": 8.836677788137488e-06,
+      "loss": 1.0823,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5848623853211009,
+      "grad_norm": 3.1697976935253256,
+      "learning_rate": 8.78699102617076e-06,
+      "loss": 1.1477,
+      "step": 2040
+    },
+    {
+      "epoch": 0.586295871559633,
+      "grad_norm": 3.060910090014108,
+      "learning_rate": 8.737334639479843e-06,
+      "loss": 1.1047,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5877293577981652,
+      "grad_norm": 3.3975873769480525,
+      "learning_rate": 8.687709871523346e-06,
+      "loss": 1.1472,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5891628440366973,
+      "grad_norm": 3.103294993064734,
+      "learning_rate": 8.638117964968098e-06,
+      "loss": 1.1147,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5905963302752294,
+      "grad_norm": 3.1038713521837176,
+      "learning_rate": 8.588560161658039e-06,
+      "loss": 1.1201,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5920298165137615,
+      "grad_norm": 3.579612587244007,
+      "learning_rate": 8.539037702583108e-06,
+      "loss": 1.2046,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5934633027522935,
+      "grad_norm": 3.1302516904022766,
+      "learning_rate": 8.489551827848197e-06,
+      "loss": 1.0237,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5948967889908257,
+      "grad_norm": 3.24638481758472,
+      "learning_rate": 8.440103776642074e-06,
+      "loss": 1.0678,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5963302752293578,
+      "grad_norm": 3.4237594408353083,
+      "learning_rate": 8.390694787206349e-06,
+      "loss": 1.1734,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5977637614678899,
+      "grad_norm": 3.0719379679166625,
+      "learning_rate": 8.341326096804489e-06,
+      "loss": 1.0969,
+      "step": 2085
+    },
+    {
+      "epoch": 0.599197247706422,
+      "grad_norm": 3.114117068959359,
+      "learning_rate": 8.291998941690821e-06,
+      "loss": 1.1036,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6006307339449541,
+      "grad_norm": 3.058349370894333,
+      "learning_rate": 8.242714557079563e-06,
+      "loss": 1.0749,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6020642201834863,
+      "grad_norm": 3.036934322182637,
+      "learning_rate": 8.193474177113918e-06,
+      "loss": 1.0583,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6034977064220184,
+      "grad_norm": 3.13799906852602,
+      "learning_rate": 8.144279034835157e-06,
+      "loss": 1.1374,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6049311926605505,
+      "grad_norm": 3.303135805293391,
+      "learning_rate": 8.095130362151737e-06,
+      "loss": 1.0465,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6063646788990825,
+      "grad_norm": 3.3978912865766686,
+      "learning_rate": 8.046029389808457e-06,
+      "loss": 1.055,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6077981651376146,
+      "grad_norm": 3.47121454527395,
+      "learning_rate": 7.996977347355647e-06,
+      "loss": 1.1155,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6092316513761468,
+      "grad_norm": 3.156947339229311,
+      "learning_rate": 7.947975463118361e-06,
+      "loss": 1.0363,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6106651376146789,
+      "grad_norm": 3.2797688368791955,
+      "learning_rate": 7.899024964165634e-06,
+      "loss": 1.1381,
+      "step": 2130
+    },
+    {
+      "epoch": 0.612098623853211,
+      "grad_norm": 2.9662302688805275,
+      "learning_rate": 7.850127076279747e-06,
+      "loss": 1.1032,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6135321100917431,
+      "grad_norm": 3.203748291163447,
+      "learning_rate": 7.801283023925536e-06,
+      "loss": 1.0517,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6149655963302753,
+      "grad_norm": 3.319287430110112,
+      "learning_rate": 7.752494030219724e-06,
+      "loss": 1.115,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6163990825688074,
+      "grad_norm": 3.7703557404425823,
+      "learning_rate": 7.703761316900293e-06,
+      "loss": 1.0813,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6178325688073395,
+      "grad_norm": 3.2303522091544727,
+      "learning_rate": 7.655086104295904e-06,
+      "loss": 1.0668,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6192660550458715,
+      "grad_norm": 3.1818989111713676,
+      "learning_rate": 7.606469611295315e-06,
+      "loss": 1.1368,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6206995412844036,
+      "grad_norm": 3.0516455298261347,
+      "learning_rate": 7.5579130553168815e-06,
+      "loss": 1.1132,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6221330275229358,
+      "grad_norm": 3.3307988467764087,
+      "learning_rate": 7.50941765227805e-06,
+      "loss": 1.0518,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6235665137614679,
+      "grad_norm": 3.3783423662774363,
+      "learning_rate": 7.460984616564929e-06,
+      "loss": 1.1132,
+      "step": 2175
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 3.634134930894941,
+      "learning_rate": 7.412615161001866e-06,
+      "loss": 1.0798,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6264334862385321,
+      "grad_norm": 3.2417955659870796,
+      "learning_rate": 7.364310496821086e-06,
+      "loss": 1.0864,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6278669724770642,
+      "grad_norm": 3.1150031782926395,
+      "learning_rate": 7.316071833632346e-06,
+      "loss": 1.1044,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6293004587155964,
+      "grad_norm": 3.058907033640181,
+      "learning_rate": 7.2679003793926626e-06,
+      "loss": 1.1097,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6307339449541285,
+      "grad_norm": 3.2942538258713348,
+      "learning_rate": 7.2197973403760614e-06,
+      "loss": 1.1005,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6321674311926605,
+      "grad_norm": 3.225308816069398,
+      "learning_rate": 7.171763921143346e-06,
+      "loss": 1.0627,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6336009174311926,
+      "grad_norm": 3.0409935206984566,
+      "learning_rate": 7.123801324511972e-06,
+      "loss": 1.0937,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6350344036697247,
+      "grad_norm": 9.664703329802007,
+      "learning_rate": 7.075910751525895e-06,
+      "loss": 1.1024,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6364678899082569,
+      "grad_norm": 3.057262418668561,
+      "learning_rate": 7.0280934014255195e-06,
+      "loss": 1.0857,
+      "step": 2220
+    },
+    {
+      "epoch": 0.637901376146789,
+      "grad_norm": 3.0509477926802395,
+      "learning_rate": 6.980350471617638e-06,
+      "loss": 1.0727,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6393348623853211,
+      "grad_norm": 3.4647112333044743,
+      "learning_rate": 6.9326831576454835e-06,
+      "loss": 1.1128,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6407683486238532,
+      "grad_norm": 8.615283468155432,
+      "learning_rate": 6.885092653158768e-06,
+      "loss": 1.1205,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6422018348623854,
+      "grad_norm": 3.6518697308480927,
+      "learning_rate": 6.837580149883787e-06,
+      "loss": 1.0998,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6436353211009175,
+      "grad_norm": 3.2292043015948493,
+      "learning_rate": 6.790146837593599e-06,
+      "loss": 1.0467,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6450688073394495,
+      "grad_norm": 3.060662379650319,
+      "learning_rate": 6.7427939040782175e-06,
+      "loss": 1.0426,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6465022935779816,
+      "grad_norm": 3.3641577098688087,
+      "learning_rate": 6.695522535114866e-06,
+      "loss": 1.0968,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6479357798165137,
+      "grad_norm": 3.025310634435396,
+      "learning_rate": 6.64833391443829e-06,
+      "loss": 1.1066,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6493692660550459,
+      "grad_norm": 3.3638409022750477,
+      "learning_rate": 6.601229223711123e-06,
+      "loss": 1.061,
+      "step": 2265
+    },
+    {
+      "epoch": 0.650802752293578,
+      "grad_norm": 3.0591413772292415,
+      "learning_rate": 6.554209642494267e-06,
+      "loss": 1.1305,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6522362385321101,
+      "grad_norm": 3.048895188924846,
+      "learning_rate": 6.507276348217393e-06,
+      "loss": 1.0676,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6536697247706422,
+      "grad_norm": 3.1625182877065,
+      "learning_rate": 6.460430516149433e-06,
+      "loss": 0.9783,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6551032110091743,
+      "grad_norm": 3.0971219975300093,
+      "learning_rate": 6.413673319369145e-06,
+      "loss": 1.0947,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6565366972477065,
+      "grad_norm": 3.0087113841047706,
+      "learning_rate": 6.36700592873576e-06,
+      "loss": 1.1115,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6579701834862385,
+      "grad_norm": 3.240656660427129,
+      "learning_rate": 6.320429512859645e-06,
+      "loss": 1.0604,
+      "step": 2295
+    },
+    {
+      "epoch": 0.6594036697247706,
+      "grad_norm": 2.8631951161021423,
+      "learning_rate": 6.273945238073047e-06,
+      "loss": 1.0521,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6608371559633027,
+      "grad_norm": 3.372695675980184,
+      "learning_rate": 6.227554268400875e-06,
+      "loss": 1.1259,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6622706422018348,
+      "grad_norm": 3.0645684579682353,
+      "learning_rate": 6.1812577655315695e-06,
+      "loss": 1.0146,
+      "step": 2310
+    },
+    {
+      "epoch": 0.663704128440367,
+      "grad_norm": 3.2992667272685456,
+      "learning_rate": 6.135056888788004e-06,
+      "loss": 1.0908,
+      "step": 2315
+    },
+    {
+      "epoch": 0.6651376146788991,
+      "grad_norm": 3.134342552266428,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 1.0408,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6665711009174312,
+      "grad_norm": 3.195978033002455,
+      "learning_rate": 6.042946638967586e-06,
+      "loss": 1.0734,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6680045871559633,
+      "grad_norm": 3.244082712886425,
+      "learning_rate": 5.997039572447658e-06,
+      "loss": 1.0466,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6694380733944955,
+      "grad_norm": 2.9373127993862105,
+      "learning_rate": 5.951232745109552e-06,
+      "loss": 1.0658,
+      "step": 2335
+    },
+    {
+      "epoch": 0.6708715596330275,
+      "grad_norm": 3.1038594855264967,
+      "learning_rate": 5.9055273040140374e-06,
+      "loss": 1.0916,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6723050458715596,
+      "grad_norm": 3.1281256278361256,
+      "learning_rate": 5.859924393683056e-06,
+      "loss": 1.1042,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6737385321100917,
+      "grad_norm": 2.9165714439580825,
+      "learning_rate": 5.8144251560710415e-06,
+      "loss": 1.0735,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6751720183486238,
+      "grad_norm": 3.3690225312568955,
+      "learning_rate": 5.769030730536336e-06,
+      "loss": 1.0574,
+      "step": 2355
+    },
+    {
+      "epoch": 0.676605504587156,
+      "grad_norm": 3.298580191163876,
+      "learning_rate": 5.723742253812658e-06,
+      "loss": 1.1132,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6780389908256881,
+      "grad_norm": 3.004930197106605,
+      "learning_rate": 5.678560859980621e-06,
+      "loss": 1.0691,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6794724770642202,
+      "grad_norm": 3.1318112002832197,
+      "learning_rate": 5.633487680439362e-06,
+      "loss": 1.0305,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6809059633027523,
+      "grad_norm": 3.139769183129407,
+      "learning_rate": 5.588523843878189e-06,
+      "loss": 1.0547,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6823394495412844,
+      "grad_norm": 3.060531894281262,
+      "learning_rate": 5.543670476248327e-06,
+      "loss": 1.0354,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6837729357798165,
+      "grad_norm": 3.1023960334424556,
+      "learning_rate": 5.498928700734713e-06,
+      "loss": 1.0565,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6852064220183486,
+      "grad_norm": 3.1211535834048396,
+      "learning_rate": 5.454299637727885e-06,
+      "loss": 1.0662,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6866399082568807,
+      "grad_norm": 2.99851195359665,
+      "learning_rate": 5.409784404795913e-06,
+      "loss": 1.0467,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6880733944954128,
+      "grad_norm": 3.0166642308548606,
+      "learning_rate": 5.365384116656415e-06,
+      "loss": 1.1033,
+      "step": 2400
+    },
+    {
+      "epoch": 0.689506880733945,
+      "grad_norm": 3.1825150095020724,
+      "learning_rate": 5.321099885148652e-06,
+      "loss": 1.0541,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6909403669724771,
+      "grad_norm": 3.3490852800179174,
+      "learning_rate": 5.2769328192056824e-06,
+      "loss": 1.0624,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6923738532110092,
+      "grad_norm": 3.1423133524388698,
+      "learning_rate": 5.23288402482658e-06,
+      "loss": 1.0582,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6938073394495413,
+      "grad_norm": 3.1573558030480235,
+      "learning_rate": 5.18895460504876e-06,
+      "loss": 1.0325,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6952408256880734,
+      "grad_norm": 3.107763639677989,
+      "learning_rate": 5.145145659920348e-06,
+      "loss": 1.0017,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6966743119266054,
+      "grad_norm": 3.3802618734734793,
+      "learning_rate": 5.101458286472618e-06,
+      "loss": 1.0718,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6981077981651376,
+      "grad_norm": 2.918248752013308,
+      "learning_rate": 5.05789357869255e-06,
+      "loss": 1.0316,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6995412844036697,
+      "grad_norm": 3.189799059727885,
+      "learning_rate": 5.01445262749542e-06,
+      "loss": 1.0851,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7009747706422018,
+      "grad_norm": 3.2475620018827676,
+      "learning_rate": 4.9711365206974716e-06,
+      "loss": 1.0537,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7024082568807339,
+      "grad_norm": 3.0707000997716047,
+      "learning_rate": 4.927946342988699e-06,
+      "loss": 1.0271,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7038417431192661,
+      "grad_norm": 2.9176475930747543,
+      "learning_rate": 4.884883175905671e-06,
+      "loss": 1.0535,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7052752293577982,
+      "grad_norm": 3.0577673743440923,
+      "learning_rate": 4.8419480978044395e-06,
+      "loss": 1.0848,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7067087155963303,
+      "grad_norm": 3.3735135945939136,
+      "learning_rate": 4.799142183833561e-06,
+      "loss": 1.0651,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7081422018348624,
+      "grad_norm": 3.170994136975781,
+      "learning_rate": 4.75646650590715e-06,
+      "loss": 1.0736,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7095756880733946,
+      "grad_norm": 2.993299399999142,
+      "learning_rate": 4.713922132678055e-06,
+      "loss": 1.0415,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7110091743119266,
+      "grad_norm": 3.4225173298945117,
+      "learning_rate": 4.671510129511074e-06,
+      "loss": 1.0363,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7124426605504587,
+      "grad_norm": 3.0344762068899365,
+      "learning_rate": 4.629231558456306e-06,
+      "loss": 1.0414,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7138761467889908,
+      "grad_norm": 3.1844264445491297,
+      "learning_rate": 4.587087478222539e-06,
+      "loss": 1.0714,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7153096330275229,
+      "grad_norm": 3.09219275586818,
+      "learning_rate": 4.545078944150728e-06,
+      "loss": 1.017,
+      "step": 2495
+    },
+    {
+      "epoch": 0.716743119266055,
+      "grad_norm": 3.4728979799373976,
+      "learning_rate": 4.5032070081876e-06,
+      "loss": 1.0612,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7181766055045872,
+      "grad_norm": 3.063801560649391,
+      "learning_rate": 4.4614727188592835e-06,
+      "loss": 1.0798,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7196100917431193,
+      "grad_norm": 3.823900363914515,
+      "learning_rate": 4.419877121245058e-06,
+      "loss": 1.0435,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7210435779816514,
+      "grad_norm": 2.8618913314337107,
+      "learning_rate": 4.378421256951192e-06,
+      "loss": 1.0102,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7224770642201835,
+      "grad_norm": 2.9378204976736466,
+      "learning_rate": 4.337106164084861e-06,
+      "loss": 1.0824,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7239105504587156,
+      "grad_norm": 2.945681604419468,
+      "learning_rate": 4.295932877228132e-06,
+      "loss": 0.9985,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7253440366972477,
+      "grad_norm": 3.106814303596456,
+      "learning_rate": 4.254902427412082e-06,
+      "loss": 1.0317,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7267775229357798,
+      "grad_norm": 3.1281573814995363,
+      "learning_rate": 4.214015842090969e-06,
+      "loss": 1.0491,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7282110091743119,
+      "grad_norm": 2.795954987429199,
+      "learning_rate": 4.173274145116491e-06,
+      "loss": 0.9885,
+      "step": 2540
+    },
+    {
+      "epoch": 0.729644495412844,
+      "grad_norm": 3.3191476249085565,
+      "learning_rate": 4.1326783567121685e-06,
+      "loss": 1.0517,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7310779816513762,
+      "grad_norm": 3.07441160825076,
+      "learning_rate": 4.092229493447788e-06,
+      "loss": 1.0512,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7325114678899083,
+      "grad_norm": 3.3162975946352824,
+      "learning_rate": 4.051928568213942e-06,
+      "loss": 1.0476,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7339449541284404,
+      "grad_norm": 3.222715903509193,
+      "learning_rate": 4.0117765901966635e-06,
+      "loss": 1.0401,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7353784403669725,
+      "grad_norm": 3.1385505256509014,
+      "learning_rate": 3.9717745648521646e-06,
+      "loss": 1.0243,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7368119266055045,
+      "grad_norm": 3.5703129773472506,
+      "learning_rate": 3.931923493881659e-06,
+      "loss": 1.0319,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7382454128440367,
+      "grad_norm": 3.0801290385363393,
+      "learning_rate": 3.892224375206256e-06,
+      "loss": 1.0382,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7396788990825688,
+      "grad_norm": 3.2917177730741676,
+      "learning_rate": 3.8526782029420005e-06,
+      "loss": 1.1191,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7411123853211009,
+      "grad_norm": 3.1425801332889827,
+      "learning_rate": 3.8132859673749688e-06,
+      "loss": 1.0481,
+      "step": 2585
+    },
+    {
+      "epoch": 0.742545871559633,
+      "grad_norm": 3.1741701966057447,
+      "learning_rate": 3.774048654936454e-06,
+      "loss": 1.0072,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7439793577981652,
+      "grad_norm": 3.4923814958157466,
+      "learning_rate": 3.7349672481782894e-06,
+      "loss": 1.01,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7454128440366973,
+      "grad_norm": 3.121310877940095,
+      "learning_rate": 3.6960427257482343e-06,
+      "loss": 1.0804,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7468463302752294,
+      "grad_norm": 3.1172405011336908,
+      "learning_rate": 3.657276062365457e-06,
+      "loss": 1.0313,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7482798165137615,
+      "grad_norm": 3.0906434983571223,
+      "learning_rate": 3.618668228796143e-06,
+      "loss": 1.0378,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7497133027522935,
+      "grad_norm": 3.1863927645425725,
+      "learning_rate": 3.580220191829178e-06,
+      "loss": 0.9929,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7511467889908257,
+      "grad_norm": 3.0251252557702437,
+      "learning_rate": 3.5419329142519433e-06,
+      "loss": 1.0814,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7525802752293578,
+      "grad_norm": 4.943706664249104,
+      "learning_rate": 3.5038073548261888e-06,
+      "loss": 1.0629,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7540137614678899,
+      "grad_norm": 3.16322658334476,
+      "learning_rate": 3.46584446826405e-06,
+      "loss": 0.9881,
+      "step": 2630
+    },
+    {
+      "epoch": 0.755447247706422,
+      "grad_norm": 3.125095249964922,
+      "learning_rate": 3.428045205204125e-06,
+      "loss": 0.9923,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7568807339449541,
+      "grad_norm": 2.973224447074909,
+      "learning_rate": 3.3904105121876764e-06,
+      "loss": 1.1052,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7583142201834863,
+      "grad_norm": 2.996637163056179,
+      "learning_rate": 3.3529413316349145e-06,
+      "loss": 1.1049,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7597477064220184,
+      "grad_norm": 2.935944096948497,
+      "learning_rate": 3.3156386018214193e-06,
+      "loss": 1.0712,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7611811926605505,
+      "grad_norm": 3.038494065615691,
+      "learning_rate": 3.2785032568546304e-06,
+      "loss": 1.0472,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7626146788990825,
+      "grad_norm": 3.2871929054536397,
+      "learning_rate": 3.2415362266504626e-06,
+      "loss": 1.0342,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7640481651376146,
+      "grad_norm": 3.098314519979228,
+      "learning_rate": 3.2047384369100065e-06,
+      "loss": 1.0903,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7654816513761468,
+      "grad_norm": 2.9569161427598787,
+      "learning_rate": 3.16811080909637e-06,
+      "loss": 1.0007,
+      "step": 2670
+    },
+    {
+      "epoch": 0.7669151376146789,
+      "grad_norm": 3.355325001454353,
+      "learning_rate": 3.1316542604115853e-06,
+      "loss": 1.0525,
+      "step": 2675
+    },
+    {
+      "epoch": 0.768348623853211,
+      "grad_norm": 3.7866942452127113,
+      "learning_rate": 3.095369703773652e-06,
+      "loss": 1.0607,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7697821100917431,
+      "grad_norm": 2.8549205403217024,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 1.0612,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7712155963302753,
+      "grad_norm": 3.1911721995721547,
+      "learning_rate": 3.0233201967530647e-06,
+      "loss": 1.0756,
+      "step": 2690
+    },
+    {
+      "epoch": 0.7726490825688074,
+      "grad_norm": 3.254594926845702,
+      "learning_rate": 2.987557050581017e-06,
+      "loss": 1.0928,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7740825688073395,
+      "grad_norm": 3.029915420885045,
+      "learning_rate": 2.9519695048318353e-06,
+      "loss": 0.9834,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7755160550458715,
+      "grad_norm": 3.7010803603107614,
+      "learning_rate": 2.9165584506625864e-06,
+      "loss": 1.0616,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7769495412844036,
+      "grad_norm": 3.0973140862278155,
+      "learning_rate": 2.8813247748107665e-06,
+      "loss": 1.0493,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7783830275229358,
+      "grad_norm": 3.3576787529696217,
+      "learning_rate": 2.8462693595720938e-06,
+      "loss": 1.0743,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7798165137614679,
+      "grad_norm": 3.074976481415647,
+      "learning_rate": 2.8113930827784076e-06,
+      "loss": 1.0157,
+      "step": 2720
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 3.0753397191274945,
+      "learning_rate": 2.776696817775707e-06,
+      "loss": 1.0266,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7826834862385321,
+      "grad_norm": 3.09366006809429,
+      "learning_rate": 2.7421814334022624e-06,
+      "loss": 1.0243,
+      "step": 2730
+    },
+    {
+      "epoch": 0.7841169724770642,
+      "grad_norm": 3.6168913804023326,
+      "learning_rate": 2.7078477939668625e-06,
+      "loss": 1.0652,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7855504587155964,
+      "grad_norm": 4.050361789957049,
+      "learning_rate": 2.673696759227177e-06,
+      "loss": 1.0636,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7869839449541285,
+      "grad_norm": 3.249151536222312,
+      "learning_rate": 2.639729184368226e-06,
+      "loss": 1.0643,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7884174311926605,
+      "grad_norm": 2.9287260819204692,
+      "learning_rate": 2.6059459199809545e-06,
+      "loss": 1.0028,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7898509174311926,
+      "grad_norm": 3.0465964857735663,
+      "learning_rate": 2.5723478120409474e-06,
+      "loss": 1.0421,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7912844036697247,
+      "grad_norm": 3.153016800252537,
+      "learning_rate": 2.5389357018872405e-06,
+      "loss": 1.0201,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7927178899082569,
+      "grad_norm": 3.115248902197113,
+      "learning_rate": 2.505710426201239e-06,
+      "loss": 1.1122,
+      "step": 2765
+    },
+    {
+      "epoch": 0.794151376146789,
+      "grad_norm": 3.456526732392788,
+      "learning_rate": 2.4726728169857885e-06,
+      "loss": 0.9736,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7955848623853211,
+      "grad_norm": 3.078706890825068,
+      "learning_rate": 2.439823701544328e-06,
+      "loss": 1.0208,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7970183486238532,
+      "grad_norm": 3.0852747181382485,
+      "learning_rate": 2.407163902460167e-06,
+      "loss": 1.0112,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7984518348623854,
+      "grad_norm": 2.9795351626224202,
+      "learning_rate": 2.3746942375758986e-06,
+      "loss": 1.0495,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7998853211009175,
+      "grad_norm": 3.662965002906388,
+      "learning_rate": 2.3424155199729206e-06,
+      "loss": 1.0088,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8013188073394495,
+      "grad_norm": 2.9979503759811146,
+      "learning_rate": 2.310328557951065e-06,
+      "loss": 0.9978,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8027522935779816,
+      "grad_norm": 3.1993479502657856,
+      "learning_rate": 2.2784341550083577e-06,
+      "loss": 1.08,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8041857798165137,
+      "grad_norm": 2.9645752085086716,
+      "learning_rate": 2.2467331098209098e-06,
+      "loss": 1.0019,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8056192660550459,
+      "grad_norm": 3.0604845129074856,
+      "learning_rate": 2.215226216222911e-06,
+      "loss": 0.9847,
+      "step": 2810
+    },
+    {
+      "epoch": 0.807052752293578,
+      "grad_norm": 3.1249395082814235,
+      "learning_rate": 2.1839142631867396e-06,
+      "loss": 1.003,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8084862385321101,
+      "grad_norm": 3.5887794377789746,
+      "learning_rate": 2.1527980348032263e-06,
+      "loss": 1.0915,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8099197247706422,
+      "grad_norm": 2.9913673167068056,
+      "learning_rate": 2.121878310262008e-06,
+      "loss": 1.0346,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8113532110091743,
+      "grad_norm": 3.076522728603463,
+      "learning_rate": 2.0911558638320117e-06,
+      "loss": 1.0254,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8127866972477065,
+      "grad_norm": 3.4952190302417625,
+      "learning_rate": 2.0606314648420757e-06,
+      "loss": 1.0164,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8142201834862385,
+      "grad_norm": 2.987547482751348,
+      "learning_rate": 2.0303058776616847e-06,
+      "loss": 1.0241,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8156536697247706,
+      "grad_norm": 3.0295895740792265,
+      "learning_rate": 2.0001798616818137e-06,
+      "loss": 0.9397,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8170871559633027,
+      "grad_norm": 3.2124562930166376,
+      "learning_rate": 1.970254171295931e-06,
+      "loss": 1.0349,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8185206422018348,
+      "grad_norm": 2.7811469673449403,
+      "learning_rate": 1.940529555881101e-06,
+      "loss": 1.0372,
+      "step": 2855
+    },
+    {
+      "epoch": 0.819954128440367,
+      "grad_norm": 3.0716049369726854,
+      "learning_rate": 1.9110067597792094e-06,
+      "loss": 1.0258,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8213876146788991,
+      "grad_norm": 3.31583577307639,
+      "learning_rate": 1.8816865222783354e-06,
+      "loss": 1.0659,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8228211009174312,
+      "grad_norm": 3.3545193733702616,
+      "learning_rate": 1.8525695775942376e-06,
+      "loss": 0.997,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8242545871559633,
+      "grad_norm": 3.082738506583357,
+      "learning_rate": 1.8236566548519664e-06,
+      "loss": 1.0109,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8256880733944955,
+      "grad_norm": 2.9272236463346326,
+      "learning_rate": 1.7949484780675941e-06,
+      "loss": 1.0096,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8271215596330275,
+      "grad_norm": 2.9288594245035937,
+      "learning_rate": 1.7664457661301103e-06,
+      "loss": 1.0126,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8285550458715596,
+      "grad_norm": 2.9288774104923077,
+      "learning_rate": 1.7381492327834004e-06,
+      "loss": 1.0499,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8299885321100917,
+      "grad_norm": 2.7877742435190065,
+      "learning_rate": 1.7100595866083713e-06,
+      "loss": 1.0041,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8314220183486238,
+      "grad_norm": 3.0286412164914656,
+      "learning_rate": 1.6821775310052212e-06,
+      "loss": 0.9811,
+      "step": 2900
+    },
+    {
+      "epoch": 0.832855504587156,
+      "grad_norm": 2.7532642945829537,
+      "learning_rate": 1.65450376417582e-06,
+      "loss": 1.0096,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8342889908256881,
+      "grad_norm": 2.9265157315568375,
+      "learning_rate": 1.6270389791062146e-06,
+      "loss": 1.0491,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8357224770642202,
+      "grad_norm": 3.216652891176282,
+      "learning_rate": 1.5997838635492936e-06,
+      "loss": 1.0956,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8371559633027523,
+      "grad_norm": 3.242345149185627,
+      "learning_rate": 1.5727391000075542e-06,
+      "loss": 0.9864,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8385894495412844,
+      "grad_norm": 3.2259601086687204,
+      "learning_rate": 1.5459053657160084e-06,
+      "loss": 1.1009,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8400229357798165,
+      "grad_norm": 2.744523780896053,
+      "learning_rate": 1.5192833326252377e-06,
+      "loss": 1.0163,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8414564220183486,
+      "grad_norm": 2.828910888875833,
+      "learning_rate": 1.4928736673845534e-06,
+      "loss": 1.0037,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8428899082568807,
+      "grad_norm": 3.062048982707707,
+      "learning_rate": 1.4666770313253054e-06,
+      "loss": 1.0248,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8443233944954128,
+      "grad_norm": 3.1169358681462174,
+      "learning_rate": 1.4406940804443303e-06,
+      "loss": 1.0094,
+      "step": 2945
+    },
+    {
+      "epoch": 0.845756880733945,
+      "grad_norm": 3.045993870767794,
+      "learning_rate": 1.4149254653875167e-06,
+      "loss": 1.0523,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8471903669724771,
+      "grad_norm": 3.0564501410586637,
+      "learning_rate": 1.389371831433507e-06,
+      "loss": 1.0095,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8486238532110092,
+      "grad_norm": 3.4108278621499917,
+      "learning_rate": 1.3640338184775526e-06,
+      "loss": 1.0366,
+      "step": 2960
+    },
+    {
+      "epoch": 0.8500573394495413,
+      "grad_norm": 3.309723042860519,
+      "learning_rate": 1.3389120610154804e-06,
+      "loss": 1.015,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8514908256880734,
+      "grad_norm": 3.40140275029476,
+      "learning_rate": 1.3140071881278106e-06,
+      "loss": 1.1159,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8529243119266054,
+      "grad_norm": 3.0642718892804344,
+      "learning_rate": 1.2893198234639904e-06,
+      "loss": 0.983,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8543577981651376,
+      "grad_norm": 3.5196505954999315,
+      "learning_rate": 1.2648505852267956e-06,
+      "loss": 1.0422,
+      "step": 2980
+    },
+    {
+      "epoch": 0.8557912844036697,
+      "grad_norm": 3.012388196153649,
+      "learning_rate": 1.240600086156839e-06,
+      "loss": 1.0161,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8572247706422018,
+      "grad_norm": 4.224630882517724,
+      "learning_rate": 1.2165689335172248e-06,
+      "loss": 1.0435,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8586582568807339,
+      "grad_norm": 3.3927019439454558,
+      "learning_rate": 1.1927577290783488e-06,
+      "loss": 1.0549,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8600917431192661,
+      "grad_norm": 2.8680598337924965,
+      "learning_rate": 1.169167069102828e-06,
+      "loss": 1.062,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8615252293577982,
+      "grad_norm": 3.2446852778299453,
+      "learning_rate": 1.1457975443305625e-06,
+      "loss": 1.0439,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8629587155963303,
+      "grad_norm": 3.074025983059374,
+      "learning_rate": 1.1226497399639501e-06,
+      "loss": 1.0202,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8643922018348624,
+      "grad_norm": 2.9805357814293423,
+      "learning_rate": 1.0997242356532335e-06,
+      "loss": 0.9824,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8658256880733946,
+      "grad_norm": 3.377450943710277,
+      "learning_rate": 1.0770216054819782e-06,
+      "loss": 1.0292,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8672591743119266,
+      "grad_norm": 3.203588510488362,
+      "learning_rate": 1.0545424179526963e-06,
+      "loss": 1.0008,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8686926605504587,
+      "grad_norm": 3.3838873878755438,
+      "learning_rate": 1.03228723597262e-06,
+      "loss": 1.0276,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8701261467889908,
+      "grad_norm": 3.0207979512808247,
+      "learning_rate": 1.0102566168395977e-06,
+      "loss": 1.0448,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8715596330275229,
+      "grad_norm": 3.294936816848852,
+      "learning_rate": 9.884511122281427e-07,
+      "loss": 1.0123,
+      "step": 3040
+    },
+    {
+      "epoch": 0.872993119266055,
+      "grad_norm": 3.0263492251087802,
+      "learning_rate": 9.668712681756087e-07,
+      "loss": 0.9943,
+      "step": 3045
+    },
+    {
+      "epoch": 0.8744266055045872,
+      "grad_norm": 2.987557644263372,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 1.0392,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8758600917431193,
+      "grad_norm": 2.827866834943389,
+      "learning_rate": 9.243907176290945e-07,
+      "loss": 1.0009,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8772935779816514,
+      "grad_norm": 3.0479507652012106,
+      "learning_rate": 9.034910749017211e-07,
+      "loss": 1.0362,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8787270642201835,
+      "grad_norm": 3.166615479232945,
+      "learning_rate": 8.828192202398455e-07,
+      "loss": 1.0649,
+      "step": 3065
+    },
+    {
+      "epoch": 0.8801605504587156,
+      "grad_norm": 3.086530733524286,
+      "learning_rate": 8.623756712928022e-07,
+      "loss": 1.0161,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8815940366972477,
+      "grad_norm": 3.2339942958362995,
+      "learning_rate": 8.421609399928621e-07,
+      "loss": 1.0274,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8830275229357798,
+      "grad_norm": 3.1624079222447845,
+      "learning_rate": 8.221755325424152e-07,
+      "loss": 0.963,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8844610091743119,
+      "grad_norm": 6.67242001014344,
+      "learning_rate": 8.024199494012863e-07,
+      "loss": 0.994,
+      "step": 3085
+    },
+    {
+      "epoch": 0.885894495412844,
+      "grad_norm": 3.0370868865276424,
+      "learning_rate": 7.828946852742148e-07,
+      "loss": 1.0551,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8873279816513762,
+      "grad_norm": 3.1653796046419185,
+      "learning_rate": 7.636002290984634e-07,
+      "loss": 1.0085,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8887614678899083,
+      "grad_norm": 3.018418033888261,
+      "learning_rate": 7.445370640315642e-07,
+      "loss": 1.0017,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8901949541284404,
+      "grad_norm": 3.543389563066827,
+      "learning_rate": 7.257056674392359e-07,
+      "loss": 1.002,
+      "step": 3105
+    },
+    {
+      "epoch": 0.8916284403669725,
+      "grad_norm": 3.1380320671176194,
+      "learning_rate": 7.071065108834197e-07,
+      "loss": 1.0311,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8930619266055045,
+      "grad_norm": 3.107368067479966,
+      "learning_rate": 6.887400601104688e-07,
+      "loss": 0.9969,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8944954128440367,
+      "grad_norm": 2.86265063408354,
+      "learning_rate": 6.706067750394951e-07,
+      "loss": 1.0004,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8959288990825688,
+      "grad_norm": 2.921057543794976,
+      "learning_rate": 6.527071097508475e-07,
+      "loss": 1.0568,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8973623853211009,
+      "grad_norm": 3.413457128089623,
+      "learning_rate": 6.350415124747378e-07,
+      "loss": 1.0966,
+      "step": 3130
+    },
+    {
+      "epoch": 0.898795871559633,
+      "grad_norm": 3.2754484907337913,
+      "learning_rate": 6.176104255800175e-07,
+      "loss": 1.0012,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9002293577981652,
+      "grad_norm": 3.096817656277008,
+      "learning_rate": 6.004142855631068e-07,
+      "loss": 1.0409,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9016628440366973,
+      "grad_norm": 3.2107227855247737,
+      "learning_rate": 5.834535230370586e-07,
+      "loss": 1.022,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9030963302752294,
+      "grad_norm": 3.087915467643749,
+      "learning_rate": 5.66728562720772e-07,
+      "loss": 1.1023,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9045298165137615,
+      "grad_norm": 3.1987341877696034,
+      "learning_rate": 5.502398234283657e-07,
+      "loss": 1.004,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9059633027522935,
+      "grad_norm": 2.8182259878196527,
+      "learning_rate": 5.339877180586872e-07,
+      "loss": 1.0388,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9073967889908257,
+      "grad_norm": 3.060289512500466,
+      "learning_rate": 5.179726535849649e-07,
+      "loss": 1.0461,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9088302752293578,
+      "grad_norm": 3.1698304516587794,
+      "learning_rate": 5.0219503104463e-07,
+      "loss": 1.0247,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9102637614678899,
+      "grad_norm": 3.130302466725309,
+      "learning_rate": 4.866552455292673e-07,
+      "loss": 1.0121,
+      "step": 3175
+    },
+    {
+      "epoch": 0.911697247706422,
+      "grad_norm": 3.42566263140553,
+      "learning_rate": 4.713536861747181e-07,
+      "loss": 1.0241,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9131307339449541,
+      "grad_norm": 2.9858737991061974,
+      "learning_rate": 4.5629073615134466e-07,
+      "loss": 1.0616,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9145642201834863,
+      "grad_norm": 3.285076013359202,
+      "learning_rate": 4.414667726544308e-07,
+      "loss": 0.9933,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9159977064220184,
+      "grad_norm": 3.2338711561101006,
+      "learning_rate": 4.2688216689472984e-07,
+      "loss": 1.0764,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9174311926605505,
+      "grad_norm": 3.1747891589130766,
+      "learning_rate": 4.125372840891817e-07,
+      "loss": 1.0288,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9188646788990825,
+      "grad_norm": 3.4369303388998604,
+      "learning_rate": 3.984324834517583e-07,
+      "loss": 1.0151,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9202981651376146,
+      "grad_norm": 2.9486157404109927,
+      "learning_rate": 3.845681181844718e-07,
+      "loss": 1.0339,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9217316513761468,
+      "grad_norm": 3.026266781773246,
+      "learning_rate": 3.7094453546852706e-07,
+      "loss": 1.0372,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9231651376146789,
+      "grad_norm": 3.6005808302403715,
+      "learning_rate": 3.575620764556331e-07,
+      "loss": 1.0675,
+      "step": 3220
+    },
+    {
+      "epoch": 0.924598623853211,
+      "grad_norm": 3.065531017189199,
+      "learning_rate": 3.4442107625945577e-07,
+      "loss": 0.9969,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9260321100917431,
+      "grad_norm": 3.101619885450492,
+      "learning_rate": 3.3152186394722506e-07,
+      "loss": 1.004,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9274655963302753,
+      "grad_norm": 3.069142098455305,
+      "learning_rate": 3.188647625315011e-07,
+      "loss": 1.0277,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9288990825688074,
+      "grad_norm": 2.981358314660114,
+      "learning_rate": 3.064500889620792e-07,
+      "loss": 1.012,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9303325688073395,
+      "grad_norm": 3.146209185290999,
+      "learning_rate": 2.9427815411805616e-07,
+      "loss": 1.0409,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9317660550458715,
+      "grad_norm": 2.9139523212945,
+      "learning_rate": 2.823492628000435e-07,
+      "loss": 1.0108,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9331995412844036,
+      "grad_norm": 3.2427768967167068,
+      "learning_rate": 2.7066371372253873e-07,
+      "loss": 1.0033,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9346330275229358,
+      "grad_norm": 3.2929773628001073,
+      "learning_rate": 2.5922179950643833e-07,
+      "loss": 1.053,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9360665137614679,
+      "grad_norm": 3.0857731418567473,
+      "learning_rate": 2.480238066717178e-07,
+      "loss": 1.0159,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 2.913155618655732,
+      "learning_rate": 2.370700156302541e-07,
+      "loss": 0.9997,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9389334862385321,
+      "grad_norm": 2.9236470377703134,
+      "learning_rate": 2.2636070067879933e-07,
+      "loss": 1.0315,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9403669724770642,
+      "grad_norm": 3.1616081864710113,
+      "learning_rate": 2.1589612999211697e-07,
+      "loss": 1.0573,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9418004587155964,
+      "grad_norm": 3.2061935710058305,
+      "learning_rate": 2.056765656162685e-07,
+      "loss": 1.0344,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9432339449541285,
+      "grad_norm": 3.1445007223648087,
+      "learning_rate": 1.9570226346204312e-07,
+      "loss": 1.0502,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9446674311926605,
+      "grad_norm": 3.0523955190740226,
+      "learning_rate": 1.8597347329855742e-07,
+      "loss": 1.0277,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9461009174311926,
+      "grad_norm": 3.10853151235872,
+      "learning_rate": 1.764904387469979e-07,
+      "loss": 0.9951,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9475344036697247,
+      "grad_norm": 2.8610073341168327,
+      "learning_rate": 1.672533972745194e-07,
+      "loss": 1.0133,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9489678899082569,
+      "grad_norm": 2.9659296802362154,
+      "learning_rate": 1.5826258018829866e-07,
+      "loss": 1.0146,
+      "step": 3310
+    },
+    {
+      "epoch": 0.950401376146789,
+      "grad_norm": 3.0101090249896187,
+      "learning_rate": 1.4951821262974563e-07,
+      "loss": 1.0064,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9518348623853211,
+      "grad_norm": 2.873431144530414,
+      "learning_rate": 1.4102051356886027e-07,
+      "loss": 0.9988,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9532683486238532,
+      "grad_norm": 2.861730520518137,
+      "learning_rate": 1.3276969579875453e-07,
+      "loss": 1.007,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9547018348623854,
+      "grad_norm": 3.1539277667342964,
+      "learning_rate": 1.24765965930318e-07,
+      "loss": 1.017,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9561353211009175,
+      "grad_norm": 3.182662519707628,
+      "learning_rate": 1.1700952438705171e-07,
+      "loss": 1.0008,
+      "step": 3335
+    },
+    {
+      "epoch": 0.9575688073394495,
+      "grad_norm": 3.078518777565793,
+      "learning_rate": 1.0950056540004029e-07,
+      "loss": 1.0649,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9590022935779816,
+      "grad_norm": 2.8736168807619507,
+      "learning_rate": 1.0223927700309667e-07,
+      "loss": 1.0177,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9604357798165137,
+      "grad_norm": 3.380358687506272,
+      "learning_rate": 9.522584102804599e-08,
+      "loss": 0.9955,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9618692660550459,
+      "grad_norm": 3.209744698526145,
+      "learning_rate": 8.846043310017927e-08,
+      "loss": 1.0051,
+      "step": 3355
+    },
+    {
+      "epoch": 0.963302752293578,
+      "grad_norm": 3.135993664577884,
+      "learning_rate": 8.194322263385013e-08,
+      "loss": 1.0096,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9647362385321101,
+      "grad_norm": 2.9702419226658128,
+      "learning_rate": 7.567437282823386e-08,
+      "loss": 0.9581,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9661697247706422,
+      "grad_norm": 2.9902064692936365,
+      "learning_rate": 6.96540406632451e-08,
+      "loss": 1.0254,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9676032110091743,
+      "grad_norm": 3.136047137814621,
+      "learning_rate": 6.388237689559762e-08,
+      "loss": 1.0386,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9690366972477065,
+      "grad_norm": 3.209855131586944,
+      "learning_rate": 5.8359526055038476e-08,
+      "loss": 1.0493,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9704701834862385,
+      "grad_norm": 3.2935877965995966,
+      "learning_rate": 5.3085626440724235e-08,
+      "loss": 1.0391,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9719036697247706,
+      "grad_norm": 3.406107022140581,
+      "learning_rate": 4.8060810117757096e-08,
+      "loss": 1.0244,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9733371559633027,
+      "grad_norm": 3.3748108058363377,
+      "learning_rate": 4.3285202913881944e-08,
+      "loss": 0.9787,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9747706422018348,
+      "grad_norm": 3.0832449198208978,
+      "learning_rate": 3.875892441633e-08,
+      "loss": 0.9951,
+      "step": 3400
+    },
+    {
+      "epoch": 0.976204128440367,
+      "grad_norm": 3.186455532049428,
+      "learning_rate": 3.4482087968829014e-08,
+      "loss": 1.0315,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9776376146788991,
+      "grad_norm": 2.9104798470989883,
+      "learning_rate": 3.045480066876105e-08,
+      "loss": 1.0092,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9790711009174312,
+      "grad_norm": 3.037035090176394,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.9587,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9805045871559633,
+      "grad_norm": 2.9918210850993785,
+      "learning_rate": 2.3149270652803592e-08,
+      "loss": 0.9973,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9819380733944955,
+      "grad_norm": 3.183820306323636,
+      "learning_rate": 1.9871210876607484e-08,
+      "loss": 1.011,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9833715596330275,
+      "grad_norm": 3.214411554963273,
+      "learning_rate": 1.6843066122649297e-08,
+      "loss": 0.9913,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9848050458715596,
+      "grad_norm": 3.5216487010248088,
+      "learning_rate": 1.4064912219496907e-08,
+      "loss": 1.0442,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9862385321100917,
+      "grad_norm": 3.1877265670501598,
+      "learning_rate": 1.1536818735630172e-08,
+      "loss": 1.0318,
+      "step": 3440
+    },
+    {
+      "epoch": 0.9876720183486238,
+      "grad_norm": 3.087482942801695,
+      "learning_rate": 9.25884897770013e-09,
+      "loss": 1.0607,
+      "step": 3445
+    },
+    {
+      "epoch": 0.989105504587156,
+      "grad_norm": 3.059255678756277,
+      "learning_rate": 7.231059988945799e-09,
+      "loss": 1.0148,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9905389908256881,
+      "grad_norm": 3.239412598410617,
+      "learning_rate": 5.4535025477642224e-09,
+      "loss": 1.0422,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9919724770642202,
+      "grad_norm": 3.144154368562322,
+      "learning_rate": 3.926221166434818e-09,
+      "loss": 1.0226,
+      "step": 3460
+    },
+    {
+      "epoch": 0.9934059633027523,
+      "grad_norm": 3.4686686147707877,
+      "learning_rate": 2.6492540900135976e-09,
+      "loss": 1.0359,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9948394495412844,
+      "grad_norm": 2.9818281052270432,
+      "learning_rate": 1.6226332953661605e-09,
+      "loss": 0.988,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9962729357798165,
+      "grad_norm": 2.940056641516156,
+      "learning_rate": 8.46384490373886e-10,
+      "loss": 1.0299,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9977064220183486,
+      "grad_norm": 3.0373605695085883,
+      "learning_rate": 3.2052711328778297e-10,
+      "loss": 0.9876,
+      "step": 3480
+    },
+    {
+      "epoch": 0.9991399082568807,
+      "grad_norm": 3.0988136193131335,
+      "learning_rate": 4.5074332237771134e-11,
+      "loss": 1.0197,
+      "step": 3485
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.0320301055908203,
+      "eval_runtime": 4528.3515,
+      "eval_samples_per_second": 18.469,
+      "eval_steps_per_second": 1.155,
+      "step": 3488
+    },
+    {
+      "epoch": 1.0,
+      "step": 3488,
+      "total_flos": 182579059752960.0,
+      "train_loss": 0.025541797006895784,
+      "train_runtime": 4827.4873,
+      "train_samples_per_second": 11.557,
+      "train_steps_per_second": 0.723
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 3488,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 182579059752960.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}