diff --git "a/checkpoint-22453/trainer_state.json" "b/checkpoint-22453/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-22453/trainer_state.json"
@@ -0,0 +1,6341 @@
+{
+  "best_metric": 0.10316640883684158,
+  "best_model_checkpoint": "autotrain-ai-image-detect-20250107-2358/checkpoint-22453",
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 22453,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011134369571994833,
+      "grad_norm": 0.0017522681737318635,
+      "learning_rate": 3.999987764250782e-05,
+      "loss": 0.0449,
+      "step": 25
+    },
+    {
+      "epoch": 0.0022268739143989666,
+      "grad_norm": 2.625354051589966,
+      "learning_rate": 3.99995105715284e-05,
+      "loss": 0.0883,
+      "step": 50
+    },
+    {
+      "epoch": 0.00334031087159845,
+      "grad_norm": 0.9239026308059692,
+      "learning_rate": 3.9998898791553145e-05,
+      "loss": 0.1545,
+      "step": 75
+    },
+    {
+      "epoch": 0.004453747828797933,
+      "grad_norm": 0.01456692535430193,
+      "learning_rate": 3.9998042310067626e-05,
+      "loss": 0.0866,
+      "step": 100
+    },
+    {
+      "epoch": 0.005567184785997417,
+      "grad_norm": 0.08153713494539261,
+      "learning_rate": 3.999694113755154e-05,
+      "loss": 0.0635,
+      "step": 125
+    },
+    {
+      "epoch": 0.0066806217431969,
+      "grad_norm": 0.8362884521484375,
+      "learning_rate": 3.9995595287478564e-05,
+      "loss": 0.109,
+      "step": 150
+    },
+    {
+      "epoch": 0.007794058700396384,
+      "grad_norm": 0.1706877499818802,
+      "learning_rate": 3.999400477631618e-05,
+      "loss": 0.1077,
+      "step": 175
+    },
+    {
+      "epoch": 0.008907495657595866,
+      "grad_norm": 0.009436409920454025,
+      "learning_rate": 3.999216962352548e-05,
+      "loss": 0.0612,
+      "step": 200
+    },
+    {
+      "epoch": 0.01002093261479535,
+      "grad_norm": 0.10162385553121567,
+      "learning_rate": 3.9990089851560926e-05,
+      "loss": 0.0925,
+      "step": 225
+    },
+    {
+      "epoch": 0.011134369571994833,
+      "grad_norm": 0.8786261081695557,
+      "learning_rate": 3.99877654858701e-05,
+      "loss": 0.0724,
+      "step": 250
+    },
+    {
+      "epoch": 0.012247806529194317,
+      "grad_norm": 24.078943252563477,
+      "learning_rate": 3.998519655489336e-05,
+      "loss": 0.1674,
+      "step": 275
+    },
+    {
+      "epoch": 0.0133612434863938,
+      "grad_norm": 0.01407669484615326,
+      "learning_rate": 3.998238309006349e-05,
+      "loss": 0.0549,
+      "step": 300
+    },
+    {
+      "epoch": 0.014474680443593284,
+      "grad_norm": 0.011299420148134232,
+      "learning_rate": 3.997932512580533e-05,
+      "loss": 0.0264,
+      "step": 325
+    },
+    {
+      "epoch": 0.015588117400792767,
+      "grad_norm": 0.676752507686615,
+      "learning_rate": 3.997602269953539e-05,
+      "loss": 0.0436,
+      "step": 350
+    },
+    {
+      "epoch": 0.01670155435799225,
+      "grad_norm": 0.010241251438856125,
+      "learning_rate": 3.997247585166133e-05,
+      "loss": 0.063,
+      "step": 375
+    },
+    {
+      "epoch": 0.017814991315191733,
+      "grad_norm": 0.03709862753748894,
+      "learning_rate": 3.996868462558146e-05,
+      "loss": 0.1051,
+      "step": 400
+    },
+    {
+      "epoch": 0.018928428272391218,
+      "grad_norm": 0.1498664766550064,
+      "learning_rate": 3.996464906768429e-05,
+      "loss": 0.0495,
+      "step": 425
+    },
+    {
+      "epoch": 0.0200418652295907,
+      "grad_norm": 0.34423527121543884,
+      "learning_rate": 3.9960369227347905e-05,
+      "loss": 0.0948,
+      "step": 450
+    },
+    {
+      "epoch": 0.021155302186790185,
+      "grad_norm": 141.0285186767578,
+      "learning_rate": 3.995584515693934e-05,
+      "loss": 0.073,
+      "step": 475
+    },
+    {
+      "epoch": 0.022268739143989667,
+      "grad_norm": 1.1084500551223755,
+      "learning_rate": 3.995107691181399e-05,
+      "loss": 0.0735,
+      "step": 500
+    },
+    {
+      "epoch": 0.023382176101189152,
+      "grad_norm": 0.8122372031211853,
+      "learning_rate": 3.994606455031492e-05,
+      "loss": 0.0643,
+      "step": 525
+    },
+    {
+      "epoch": 0.024495613058388634,
+      "grad_norm": 0.2024221420288086,
+      "learning_rate": 3.9940808133772114e-05,
+      "loss": 0.0483,
+      "step": 550
+    },
+    {
+      "epoch": 0.02560905001558812,
+      "grad_norm": 0.052694253623485565,
+      "learning_rate": 3.9935307726501774e-05,
+      "loss": 0.1225,
+      "step": 575
+    },
+    {
+      "epoch": 0.0267224869727876,
+      "grad_norm": 0.2362474501132965,
+      "learning_rate": 3.99295633958055e-05,
+      "loss": 0.0988,
+      "step": 600
+    },
+    {
+      "epoch": 0.027835923929987082,
+      "grad_norm": 0.009827379137277603,
+      "learning_rate": 3.992357521196948e-05,
+      "loss": 0.0966,
+      "step": 625
+    },
+    {
+      "epoch": 0.028949360887186568,
+      "grad_norm": 8.952296257019043,
+      "learning_rate": 3.991734324826364e-05,
+      "loss": 0.1072,
+      "step": 650
+    },
+    {
+      "epoch": 0.03006279784438605,
+      "grad_norm": 7.652793884277344,
+      "learning_rate": 3.9910867580940705e-05,
+      "loss": 0.117,
+      "step": 675
+    },
+    {
+      "epoch": 0.031176234801585535,
+      "grad_norm": 27.530536651611328,
+      "learning_rate": 3.990414828923533e-05,
+      "loss": 0.157,
+      "step": 700
+    },
+    {
+      "epoch": 0.03228967175878502,
+      "grad_norm": 0.032850898802280426,
+      "learning_rate": 3.989718545536309e-05,
+      "loss": 0.054,
+      "step": 725
+    },
+    {
+      "epoch": 0.0334031087159845,
+      "grad_norm": 3.0997722148895264,
+      "learning_rate": 3.988997916451946e-05,
+      "loss": 0.078,
+      "step": 750
+    },
+    {
+      "epoch": 0.03451654567318398,
+      "grad_norm": 70.86727905273438,
+      "learning_rate": 3.9882529504878815e-05,
+      "loss": 0.129,
+      "step": 775
+    },
+    {
+      "epoch": 0.035629982630383465,
+      "grad_norm": 0.058145686984062195,
+      "learning_rate": 3.987483656759332e-05,
+      "loss": 0.0496,
+      "step": 800
+    },
+    {
+      "epoch": 0.036743419587582954,
+      "grad_norm": 0.1561250239610672,
+      "learning_rate": 3.986690044679183e-05,
+      "loss": 0.0652,
+      "step": 825
+    },
+    {
+      "epoch": 0.037856856544782436,
+      "grad_norm": 5.164768695831299,
+      "learning_rate": 3.9858721239578725e-05,
+      "loss": 0.0951,
+      "step": 850
+    },
+    {
+      "epoch": 0.03897029350198192,
+      "grad_norm": 0.022487835958600044,
+      "learning_rate": 3.9850299046032735e-05,
+      "loss": 0.0954,
+      "step": 875
+    },
+    {
+      "epoch": 0.0400837304591814,
+      "grad_norm": 15.329583168029785,
+      "learning_rate": 3.984163396920571e-05,
+      "loss": 0.0913,
+      "step": 900
+    },
+    {
+      "epoch": 0.04119716741638089,
+      "grad_norm": 0.03666286543011665,
+      "learning_rate": 3.983272611512135e-05,
+      "loss": 0.0761,
+      "step": 925
+    },
+    {
+      "epoch": 0.04231060437358037,
+      "grad_norm": 0.03395090997219086,
+      "learning_rate": 3.982357559277393e-05,
+      "loss": 0.1114,
+      "step": 950
+    },
+    {
+      "epoch": 0.04342404133077985,
+      "grad_norm": 0.05037323012948036,
+      "learning_rate": 3.981418251412694e-05,
+      "loss": 0.1333,
+      "step": 975
+    },
+    {
+      "epoch": 0.04453747828797933,
+      "grad_norm": 0.0885915756225586,
+      "learning_rate": 3.980454699411175e-05,
+      "loss": 0.0932,
+      "step": 1000
+    },
+    {
+      "epoch": 0.045650915245178815,
+      "grad_norm": 0.27377429604530334,
+      "learning_rate": 3.979466915062615e-05,
+      "loss": 0.096,
+      "step": 1025
+    },
+    {
+      "epoch": 0.046764352202378304,
+      "grad_norm": 0.03340262174606323,
+      "learning_rate": 3.978454910453296e-05,
+      "loss": 0.0788,
+      "step": 1050
+    },
+    {
+      "epoch": 0.047877789159577785,
+      "grad_norm": 1.4525907039642334,
+      "learning_rate": 3.9774186979658525e-05,
+      "loss": 0.0303,
+      "step": 1075
+    },
+    {
+      "epoch": 0.04899122611677727,
+      "grad_norm": 0.0045611136592924595,
+      "learning_rate": 3.976358290279121e-05,
+      "loss": 0.0807,
+      "step": 1100
+    },
+    {
+      "epoch": 0.05010466307397675,
+      "grad_norm": 0.43597421050071716,
+      "learning_rate": 3.975273700367985e-05,
+      "loss": 0.0809,
+      "step": 1125
+    },
+    {
+      "epoch": 0.05121810003117624,
+      "grad_norm": 0.049976006150245667,
+      "learning_rate": 3.9741649415032124e-05,
+      "loss": 0.1176,
+      "step": 1150
+    },
+    {
+      "epoch": 0.05233153698837572,
+      "grad_norm": 0.09874877333641052,
+      "learning_rate": 3.9730320272513006e-05,
+      "loss": 0.0726,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0534449739455752,
+      "grad_norm": 0.05045215040445328,
+      "learning_rate": 3.9718749714743034e-05,
+      "loss": 0.1028,
+      "step": 1200
+    },
+    {
+      "epoch": 0.05455841090277468,
+      "grad_norm": 0.13014136254787445,
+      "learning_rate": 3.970693788329665e-05,
+      "loss": 0.1244,
+      "step": 1225
+    },
+    {
+      "epoch": 0.055671847859974165,
+      "grad_norm": 0.05505981668829918,
+      "learning_rate": 3.969488492270046e-05,
+      "loss": 0.1089,
+      "step": 1250
+    },
+    {
+      "epoch": 0.056785284817173654,
+      "grad_norm": 0.01733974926173687,
+      "learning_rate": 3.968259098043147e-05,
+      "loss": 0.1082,
+      "step": 1275
+    },
+    {
+      "epoch": 0.057898721774373135,
+      "grad_norm": 7.767026901245117,
+      "learning_rate": 3.967005620691528e-05,
+      "loss": 0.1032,
+      "step": 1300
+    },
+    {
+      "epoch": 0.05901215873157262,
+      "grad_norm": 10.336471557617188,
+      "learning_rate": 3.9657280755524227e-05,
+      "loss": 0.0797,
+      "step": 1325
+    },
+    {
+      "epoch": 0.0601255956887721,
+      "grad_norm": 0.03750215098261833,
+      "learning_rate": 3.964426478257554e-05,
+      "loss": 0.0697,
+      "step": 1350
+    },
+    {
+      "epoch": 0.06123903264597159,
+      "grad_norm": 15.773834228515625,
+      "learning_rate": 3.963100844732939e-05,
+      "loss": 0.0839,
+      "step": 1375
+    },
+    {
+      "epoch": 0.06235246960317107,
+      "grad_norm": 78.63215637207031,
+      "learning_rate": 3.9617511911986966e-05,
+      "loss": 0.1368,
+      "step": 1400
+    },
+    {
+      "epoch": 0.06346590656037056,
+      "grad_norm": 0.0823257565498352,
+      "learning_rate": 3.960377534168851e-05,
+      "loss": 0.1027,
+      "step": 1425
+    },
+    {
+      "epoch": 0.06457934351757004,
+      "grad_norm": 56.04353332519531,
+      "learning_rate": 3.9589798904511234e-05,
+      "loss": 0.0504,
+      "step": 1450
+    },
+    {
+      "epoch": 0.06569278047476952,
+      "grad_norm": 2.2531938552856445,
+      "learning_rate": 3.957558277146732e-05,
+      "loss": 0.1318,
+      "step": 1475
+    },
+    {
+      "epoch": 0.066806217431969,
+      "grad_norm": 0.08561531454324722,
+      "learning_rate": 3.9561127116501816e-05,
+      "loss": 0.0387,
+      "step": 1500
+    },
+    {
+      "epoch": 0.06791965438916849,
+      "grad_norm": 0.03525933623313904,
+      "learning_rate": 3.954643211649048e-05,
+      "loss": 0.1259,
+      "step": 1525
+    },
+    {
+      "epoch": 0.06903309134636797,
+      "grad_norm": 26.943761825561523,
+      "learning_rate": 3.953149795123766e-05,
+      "loss": 0.1124,
+      "step": 1550
+    },
+    {
+      "epoch": 0.07014652830356745,
+      "grad_norm": 17.139528274536133,
+      "learning_rate": 3.951632480347404e-05,
+      "loss": 0.0953,
+      "step": 1575
+    },
+    {
+      "epoch": 0.07125996526076693,
+      "grad_norm": 13.074729919433594,
+      "learning_rate": 3.950091285885447e-05,
+      "loss": 0.1227,
+      "step": 1600
+    },
+    {
+      "epoch": 0.07237340221796641,
+      "grad_norm": 0.05087602138519287,
+      "learning_rate": 3.948526230595562e-05,
+      "loss": 0.0878,
+      "step": 1625
+    },
+    {
+      "epoch": 0.07348683917516591,
+      "grad_norm": 11.556482315063477,
+      "learning_rate": 3.9469373336273745e-05,
+      "loss": 0.0345,
+      "step": 1650
+    },
+    {
+      "epoch": 0.07460027613236539,
+      "grad_norm": 0.24067726731300354,
+      "learning_rate": 3.9453246144222283e-05,
+      "loss": 0.0578,
+      "step": 1675
+    },
+    {
+      "epoch": 0.07571371308956487,
+      "grad_norm": 1.8344037532806396,
+      "learning_rate": 3.9436880927129515e-05,
+      "loss": 0.0968,
+      "step": 1700
+    },
+    {
+      "epoch": 0.07682715004676435,
+      "grad_norm": 23.347782135009766,
+      "learning_rate": 3.942027788523614e-05,
+      "loss": 0.1451,
+      "step": 1725
+    },
+    {
+      "epoch": 0.07794058700396383,
+      "grad_norm": 45.14012145996094,
+      "learning_rate": 3.9403437221692806e-05,
+      "loss": 0.1348,
+      "step": 1750
+    },
+    {
+      "epoch": 0.07905402396116332,
+      "grad_norm": 35.514225006103516,
+      "learning_rate": 3.938635914255766e-05,
+      "loss": 0.147,
+      "step": 1775
+    },
+    {
+      "epoch": 0.0801674609183628,
+      "grad_norm": 0.09194282442331314,
+      "learning_rate": 3.9369043856793775e-05,
+      "loss": 0.0951,
+      "step": 1800
+    },
+    {
+      "epoch": 0.08128089787556228,
+      "grad_norm": 33.31999969482422,
+      "learning_rate": 3.9351491576266675e-05,
+      "loss": 0.0485,
+      "step": 1825
+    },
+    {
+      "epoch": 0.08239433483276178,
+      "grad_norm": 0.5410212278366089,
+      "learning_rate": 3.933370251574163e-05,
+      "loss": 0.0664,
+      "step": 1850
+    },
+    {
+      "epoch": 0.08350777178996126,
+      "grad_norm": 25.731027603149414,
+      "learning_rate": 3.931567689288115e-05,
+      "loss": 0.1271,
+      "step": 1875
+    },
+    {
+      "epoch": 0.08462120874716074,
+      "grad_norm": 0.052043963223695755,
+      "learning_rate": 3.929741492824222e-05,
+      "loss": 0.0841,
+      "step": 1900
+    },
+    {
+      "epoch": 0.08573464570436022,
+      "grad_norm": 0.07185567915439606,
+      "learning_rate": 3.9278916845273665e-05,
+      "loss": 0.1116,
+      "step": 1925
+    },
+    {
+      "epoch": 0.0868480826615597,
+      "grad_norm": 54.373497009277344,
+      "learning_rate": 3.926018287031339e-05,
+      "loss": 0.0396,
+      "step": 1950
+    },
+    {
+      "epoch": 0.08796151961875918,
+      "grad_norm": 0.1579923927783966,
+      "learning_rate": 3.9241213232585607e-05,
+      "loss": 0.1455,
+      "step": 1975
+    },
+    {
+      "epoch": 0.08907495657595867,
+      "grad_norm": 0.19795724749565125,
+      "learning_rate": 3.922200816419806e-05,
+      "loss": 0.1301,
+      "step": 2000
+    },
+    {
+      "epoch": 0.09018839353315815,
+      "grad_norm": 10.653286933898926,
+      "learning_rate": 3.9202567900139134e-05,
+      "loss": 0.0711,
+      "step": 2025
+    },
+    {
+      "epoch": 0.09130183049035763,
+      "grad_norm": 0.07394363731145859,
+      "learning_rate": 3.918289267827504e-05,
+      "loss": 0.0728,
+      "step": 2050
+    },
+    {
+      "epoch": 0.09241526744755713,
+      "grad_norm": 0.03366103768348694,
+      "learning_rate": 3.916298273934685e-05,
+      "loss": 0.1524,
+      "step": 2075
+    },
+    {
+      "epoch": 0.09352870440475661,
+      "grad_norm": 0.06265176832675934,
+      "learning_rate": 3.9142838326967584e-05,
+      "loss": 0.086,
+      "step": 2100
+    },
+    {
+      "epoch": 0.09464214136195609,
+      "grad_norm": 0.4850364327430725,
+      "learning_rate": 3.9122459687619224e-05,
+      "loss": 0.0464,
+      "step": 2125
+    },
+    {
+      "epoch": 0.09575557831915557,
+      "grad_norm": 0.05570212006568909,
+      "learning_rate": 3.910184707064968e-05,
+      "loss": 0.0935,
+      "step": 2150
+    },
+    {
+      "epoch": 0.09686901527635505,
+      "grad_norm": 0.46077781915664673,
+      "learning_rate": 3.908100072826977e-05,
+      "loss": 0.0563,
+      "step": 2175
+    },
+    {
+      "epoch": 0.09798245223355453,
+      "grad_norm": 0.29431354999542236,
+      "learning_rate": 3.905992091555012e-05,
+      "loss": 0.1255,
+      "step": 2200
+    },
+    {
+      "epoch": 0.09909588919075402,
+      "grad_norm": 12.313212394714355,
+      "learning_rate": 3.903860789041802e-05,
+      "loss": 0.1565,
+      "step": 2225
+    },
+    {
+      "epoch": 0.1002093261479535,
+      "grad_norm": 2.2681148052215576,
+      "learning_rate": 3.9017061913654314e-05,
+      "loss": 0.1254,
+      "step": 2250
+    },
+    {
+      "epoch": 0.10132276310515298,
+      "grad_norm": 0.29880329966545105,
+      "learning_rate": 3.8995283248890145e-05,
+      "loss": 0.055,
+      "step": 2275
+    },
+    {
+      "epoch": 0.10243620006235248,
+      "grad_norm": 171.1862335205078,
+      "learning_rate": 3.897327216260383e-05,
+      "loss": 0.0868,
+      "step": 2300
+    },
+    {
+      "epoch": 0.10354963701955196,
+      "grad_norm": 10.678813934326172,
+      "learning_rate": 3.895102892411747e-05,
+      "loss": 0.1091,
+      "step": 2325
+    },
+    {
+      "epoch": 0.10466307397675144,
+      "grad_norm": 11.668232917785645,
+      "learning_rate": 3.892855380559376e-05,
+      "loss": 0.1167,
+      "step": 2350
+    },
+    {
+      "epoch": 0.10577651093395092,
+      "grad_norm": 0.5463631749153137,
+      "learning_rate": 3.8905847082032625e-05,
+      "loss": 0.1234,
+      "step": 2375
+    },
+    {
+      "epoch": 0.1068899478911504,
+      "grad_norm": 0.034784119576215744,
+      "learning_rate": 3.888290903126783e-05,
+      "loss": 0.197,
+      "step": 2400
+    },
+    {
+      "epoch": 0.10800338484834988,
+      "grad_norm": 0.06954146921634674,
+      "learning_rate": 3.885973993396362e-05,
+      "loss": 0.0356,
+      "step": 2425
+    },
+    {
+      "epoch": 0.10911682180554937,
+      "grad_norm": 0.052475456148386,
+      "learning_rate": 3.8836340073611245e-05,
+      "loss": 0.0923,
+      "step": 2450
+    },
+    {
+      "epoch": 0.11023025876274885,
+      "grad_norm": 0.033263199031353,
+      "learning_rate": 3.881270973652554e-05,
+      "loss": 0.1079,
+      "step": 2475
+    },
+    {
+      "epoch": 0.11134369571994833,
+      "grad_norm": 0.024731379002332687,
+      "learning_rate": 3.8788849211841386e-05,
+      "loss": 0.0463,
+      "step": 2500
+    },
+    {
+      "epoch": 0.11245713267714783,
+      "grad_norm": 0.13399425148963928,
+      "learning_rate": 3.8764758791510165e-05,
+      "loss": 0.0258,
+      "step": 2525
+    },
+    {
+      "epoch": 0.11357056963434731,
+      "grad_norm": 0.03367041051387787,
+      "learning_rate": 3.874043877029623e-05,
+      "loss": 0.1288,
+      "step": 2550
+    },
+    {
+      "epoch": 0.11468400659154679,
+      "grad_norm": 0.12984095513820648,
+      "learning_rate": 3.8715889445773263e-05,
+      "loss": 0.1032,
+      "step": 2575
+    },
+    {
+      "epoch": 0.11579744354874627,
+      "grad_norm": 4.606603622436523,
+      "learning_rate": 3.869111111832063e-05,
+      "loss": 0.0668,
+      "step": 2600
+    },
+    {
+      "epoch": 0.11691088050594575,
+      "grad_norm": 0.2087305188179016,
+      "learning_rate": 3.866610409111974e-05,
+      "loss": 0.1088,
+      "step": 2625
+    },
+    {
+      "epoch": 0.11802431746314523,
+      "grad_norm": 0.05369157716631889,
+      "learning_rate": 3.864086867015031e-05,
+      "loss": 0.048,
+      "step": 2650
+    },
+    {
+      "epoch": 0.11913775442034472,
+      "grad_norm": 0.21587111055850983,
+      "learning_rate": 3.861540516418663e-05,
+      "loss": 0.0999,
+      "step": 2675
+    },
+    {
+      "epoch": 0.1202511913775442,
+      "grad_norm": 0.8712963461875916,
+      "learning_rate": 3.858971388479375e-05,
+      "loss": 0.1256,
+      "step": 2700
+    },
+    {
+      "epoch": 0.1213646283347437,
+      "grad_norm": 0.18013902008533478,
+      "learning_rate": 3.8563795146323733e-05,
+      "loss": 0.1284,
+      "step": 2725
+    },
+    {
+      "epoch": 0.12247806529194318,
+      "grad_norm": 265.93426513671875,
+      "learning_rate": 3.8537649265911766e-05,
+      "loss": 0.1636,
+      "step": 2750
+    },
+    {
+      "epoch": 0.12359150224914266,
+      "grad_norm": 7.281203746795654,
+      "learning_rate": 3.851127656347228e-05,
+      "loss": 0.1206,
+      "step": 2775
+    },
+    {
+      "epoch": 0.12470493920634214,
+      "grad_norm": 0.5756829977035522,
+      "learning_rate": 3.8484677361695054e-05,
+      "loss": 0.158,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1258183761635416,
+      "grad_norm": 67.49952697753906,
+      "learning_rate": 3.845785198604125e-05,
+      "loss": 0.0772,
+      "step": 2825
+    },
+    {
+      "epoch": 0.12693181312074112,
+      "grad_norm": 30.2692813873291,
+      "learning_rate": 3.843080076473943e-05,
+      "loss": 0.1661,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1280452500779406,
+      "grad_norm": 0.4516218602657318,
+      "learning_rate": 3.840352402878156e-05,
+      "loss": 0.0842,
+      "step": 2875
+    },
+    {
+      "epoch": 0.12915868703514008,
+      "grad_norm": 10.178808212280273,
+      "learning_rate": 3.837602211191894e-05,
+      "loss": 0.1355,
+      "step": 2900
+    },
+    {
+      "epoch": 0.13027212399233956,
+      "grad_norm": 0.2553464472293854,
+      "learning_rate": 3.834829535065812e-05,
+      "loss": 0.1288,
+      "step": 2925
+    },
+    {
+      "epoch": 0.13138556094953904,
+      "grad_norm": 15.92351245880127,
+      "learning_rate": 3.832034408425682e-05,
+      "loss": 0.1105,
+      "step": 2950
+    },
+    {
+      "epoch": 0.13249899790673852,
+      "grad_norm": 0.41279423236846924,
+      "learning_rate": 3.82921686547197e-05,
+      "loss": 0.0394,
+      "step": 2975
+    },
+    {
+      "epoch": 0.133612434863938,
+      "grad_norm": 0.03520906716585159,
+      "learning_rate": 3.826376940679426e-05,
+      "loss": 0.118,
+      "step": 3000
+    },
+    {
+      "epoch": 0.1347258718211375,
+      "grad_norm": 156.47885131835938,
+      "learning_rate": 3.823514668796657e-05,
+      "loss": 0.0569,
+      "step": 3025
+    },
+    {
+      "epoch": 0.13583930877833697,
+      "grad_norm": 0.1534019112586975,
+      "learning_rate": 3.8206300848457054e-05,
+      "loss": 0.1154,
+      "step": 3050
+    },
+    {
+      "epoch": 0.13695274573553645,
+      "grad_norm": 14.097362518310547,
+      "learning_rate": 3.817723224121616e-05,
+      "loss": 0.0839,
+      "step": 3075
+    },
+    {
+      "epoch": 0.13806618269273593,
+      "grad_norm": 0.008445854298770428,
+      "learning_rate": 3.814794122192008e-05,
+      "loss": 0.0128,
+      "step": 3100
+    },
+    {
+      "epoch": 0.13917961964993542,
+      "grad_norm": 0.1724538505077362,
+      "learning_rate": 3.811842814896637e-05,
+      "loss": 0.099,
+      "step": 3125
+    },
+    {
+      "epoch": 0.1402930566071349,
+      "grad_norm": 9.441699981689453,
+      "learning_rate": 3.8088693383469604e-05,
+      "loss": 0.1023,
+      "step": 3150
+    },
+    {
+      "epoch": 0.14140649356433438,
+      "grad_norm": 22.488603591918945,
+      "learning_rate": 3.805873728925691e-05,
+      "loss": 0.1259,
+      "step": 3175
+    },
+    {
+      "epoch": 0.14251993052153386,
+      "grad_norm": 1.3374933004379272,
+      "learning_rate": 3.802856023286354e-05,
+      "loss": 0.0282,
+      "step": 3200
+    },
+    {
+      "epoch": 0.14363336747873334,
+      "grad_norm": 0.08540292829275131,
+      "learning_rate": 3.799816258352839e-05,
+      "loss": 0.1088,
+      "step": 3225
+    },
+    {
+      "epoch": 0.14474680443593282,
+      "grad_norm": 0.20437736809253693,
+      "learning_rate": 3.796754471318948e-05,
+      "loss": 0.1092,
+      "step": 3250
+    },
+    {
+      "epoch": 0.14586024139313233,
+      "grad_norm": 9.652408599853516,
+      "learning_rate": 3.7936706996479383e-05,
+      "loss": 0.1386,
+      "step": 3275
+    },
+    {
+      "epoch": 0.14697367835033182,
+      "grad_norm": 0.07140795141458511,
+      "learning_rate": 3.7905649810720685e-05,
+      "loss": 0.0751,
+      "step": 3300
+    },
+    {
+      "epoch": 0.1480871153075313,
+      "grad_norm": 0.33767127990722656,
+      "learning_rate": 3.78743735359213e-05,
+      "loss": 0.0276,
+      "step": 3325
+    },
+    {
+      "epoch": 0.14920055226473078,
+      "grad_norm": 12.312952995300293,
+      "learning_rate": 3.78428785547699e-05,
+      "loss": 0.1186,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15031398922193026,
+      "grad_norm": 0.3609814941883087,
+      "learning_rate": 3.781116525263117e-05,
+      "loss": 0.0902,
+      "step": 3375
+    },
+    {
+      "epoch": 0.15142742617912974,
+      "grad_norm": 2.5911900997161865,
+      "learning_rate": 3.7779234017541116e-05,
+      "loss": 0.0772,
+      "step": 3400
+    },
+    {
+      "epoch": 0.15254086313632922,
+      "grad_norm": 0.28325122594833374,
+      "learning_rate": 3.7747085240202324e-05,
+      "loss": 0.0747,
+      "step": 3425
+    },
+    {
+      "epoch": 0.1536543000935287,
+      "grad_norm": 0.08380604535341263,
+      "learning_rate": 3.771471931397918e-05,
+      "loss": 0.0432,
+      "step": 3450
+    },
+    {
+      "epoch": 0.1547677370507282,
+      "grad_norm": 14.243409156799316,
+      "learning_rate": 3.768213663489304e-05,
+      "loss": 0.0615,
+      "step": 3475
+    },
+    {
+      "epoch": 0.15588117400792767,
+      "grad_norm": 56.65135955810547,
+      "learning_rate": 3.764933760161739e-05,
+      "loss": 0.0659,
+      "step": 3500
+    },
+    {
+      "epoch": 0.15699461096512715,
+      "grad_norm": 0.15652716159820557,
+      "learning_rate": 3.761632261547297e-05,
+      "loss": 0.1269,
+      "step": 3525
+    },
+    {
+      "epoch": 0.15810804792232663,
+      "grad_norm": 7.383078575134277,
+      "learning_rate": 3.758309208042288e-05,
+      "loss": 0.1109,
+      "step": 3550
+    },
+    {
+      "epoch": 0.15922148487952612,
+      "grad_norm": 95.15608978271484,
+      "learning_rate": 3.754964640306761e-05,
+      "loss": 0.146,
+      "step": 3575
+    },
+    {
+      "epoch": 0.1603349218367256,
+      "grad_norm": 0.1545642465353012,
+      "learning_rate": 3.751598599264008e-05,
+      "loss": 0.0986,
+      "step": 3600
+    },
+    {
+      "epoch": 0.16144835879392508,
+      "grad_norm": 0.13380536437034607,
+      "learning_rate": 3.748211126100063e-05,
+      "loss": 0.1089,
+      "step": 3625
+    },
+    {
+      "epoch": 0.16256179575112456,
+      "grad_norm": 22.723392486572266,
+      "learning_rate": 3.744802262263198e-05,
+      "loss": 0.1362,
+      "step": 3650
+    },
+    {
+      "epoch": 0.16367523270832404,
+      "grad_norm": 19.4218807220459,
+      "learning_rate": 3.741372049463417e-05,
+      "loss": 0.1095,
+      "step": 3675
+    },
+    {
+      "epoch": 0.16478866966552355,
+      "grad_norm": 17.95904541015625,
+      "learning_rate": 3.737920529671942e-05,
+      "loss": 0.0712,
+      "step": 3700
+    },
+    {
+      "epoch": 0.16590210662272303,
+      "grad_norm": 1.368179440498352,
+      "learning_rate": 3.734447745120704e-05,
+      "loss": 0.1154,
+      "step": 3725
+    },
+    {
+      "epoch": 0.16701554357992252,
+      "grad_norm": 0.02144320122897625,
+      "learning_rate": 3.730953738301826e-05,
+      "loss": 0.0832,
+      "step": 3750
+    },
+    {
+      "epoch": 0.168128980537122,
+      "grad_norm": 1.7905651330947876,
+      "learning_rate": 3.727438551967096e-05,
+      "loss": 0.0824,
+      "step": 3775
+    },
+    {
+      "epoch": 0.16924241749432148,
+      "grad_norm": 0.22187261283397675,
+      "learning_rate": 3.7239022291274546e-05,
+      "loss": 0.1902,
+      "step": 3800
+    },
+    {
+      "epoch": 0.17035585445152096,
+      "grad_norm": 0.06837914884090424,
+      "learning_rate": 3.72034481305246e-05,
+      "loss": 0.0784,
+      "step": 3825
+    },
+    {
+      "epoch": 0.17146929140872044,
+      "grad_norm": 0.0925784558057785,
+      "learning_rate": 3.7167663472697645e-05,
+      "loss": 0.0687,
+      "step": 3850
+    },
+    {
+      "epoch": 0.17258272836591992,
+      "grad_norm": 0.2474130392074585,
+      "learning_rate": 3.713166875564577e-05,
+      "loss": 0.1039,
+      "step": 3875
+    },
+    {
+      "epoch": 0.1736961653231194,
+      "grad_norm": 0.6533392071723938,
+      "learning_rate": 3.70954644197913e-05,
+      "loss": 0.0853,
+      "step": 3900
+    },
+    {
+      "epoch": 0.1748096022803189,
+      "grad_norm": 13.996048927307129,
+      "learning_rate": 3.705905090812143e-05,
+      "loss": 0.0873,
+      "step": 3925
+    },
+    {
+      "epoch": 0.17592303923751837,
+      "grad_norm": 15.874128341674805,
+      "learning_rate": 3.7022428666182736e-05,
+      "loss": 0.0953,
+      "step": 3950
+    },
+    {
+      "epoch": 0.17703647619471785,
+      "grad_norm": 189.27093505859375,
+      "learning_rate": 3.69855981420758e-05,
+      "loss": 0.1689,
+      "step": 3975
+    },
+    {
+      "epoch": 0.17814991315191733,
+      "grad_norm": 2.6344797611236572,
+      "learning_rate": 3.694855978644967e-05,
+      "loss": 0.0615,
+      "step": 4000
+    },
+    {
+      "epoch": 0.17926335010911681,
+      "grad_norm": 0.16542290151119232,
+      "learning_rate": 3.6911314052496376e-05,
+      "loss": 0.0705,
+      "step": 4025
+    },
+    {
+      "epoch": 0.1803767870663163,
+      "grad_norm": 0.18003667891025543,
+      "learning_rate": 3.687386139594539e-05,
+      "loss": 0.0997,
+      "step": 4050
+    },
+    {
+      "epoch": 0.18149022402351578,
+      "grad_norm": 16.793657302856445,
+      "learning_rate": 3.683620227505801e-05,
+      "loss": 0.0762,
+      "step": 4075
+    },
+    {
+      "epoch": 0.18260366098071526,
+      "grad_norm": 19.98533821105957,
+      "learning_rate": 3.67983371506218e-05,
+      "loss": 0.1019,
+      "step": 4100
+    },
+    {
+      "epoch": 0.18371709793791474,
+      "grad_norm": 11.568763732910156,
+      "learning_rate": 3.676026648594494e-05,
+      "loss": 0.0991,
+      "step": 4125
+    },
+    {
+      "epoch": 0.18483053489511425,
+      "grad_norm": 0.13966935873031616,
+      "learning_rate": 3.6721990746850516e-05,
+      "loss": 0.0766,
+      "step": 4150
+    },
+    {
+      "epoch": 0.18594397185231373,
+      "grad_norm": 9.846770286560059,
+      "learning_rate": 3.668351040167088e-05,
+      "loss": 0.1473,
+      "step": 4175
+    },
+    {
+      "epoch": 0.18705740880951321,
+      "grad_norm": 0.05613240599632263,
+      "learning_rate": 3.664482592124189e-05,
+      "loss": 0.0633,
+      "step": 4200
+    },
+    {
+      "epoch": 0.1881708457667127,
+      "grad_norm": 0.24261285364627838,
+      "learning_rate": 3.6605937778897145e-05,
+      "loss": 0.0994,
+      "step": 4225
+    },
+    {
+      "epoch": 0.18928428272391218,
+      "grad_norm": 0.037253621965646744,
+      "learning_rate": 3.6566846450462194e-05,
+      "loss": 0.073,
+      "step": 4250
+    },
+    {
+      "epoch": 0.19039771968111166,
+      "grad_norm": 0.6551414728164673,
+      "learning_rate": 3.652755241424875e-05,
+      "loss": 0.155,
+      "step": 4275
+    },
+    {
+      "epoch": 0.19151115663831114,
+      "grad_norm": 1.646484136581421,
+      "learning_rate": 3.6488056151048756e-05,
+      "loss": 0.0776,
+      "step": 4300
+    },
+    {
+      "epoch": 0.19262459359551062,
+      "grad_norm": 306.798095703125,
+      "learning_rate": 3.6448358144128606e-05,
+      "loss": 0.1294,
+      "step": 4325
+    },
+    {
+      "epoch": 0.1937380305527101,
+      "grad_norm": 0.2583596110343933,
+      "learning_rate": 3.640845887922315e-05,
+      "loss": 0.0403,
+      "step": 4350
+    },
+    {
+      "epoch": 0.1948514675099096,
+      "grad_norm": 0.03823433443903923,
+      "learning_rate": 3.636835884452978e-05,
+      "loss": 0.1147,
+      "step": 4375
+    },
+    {
+      "epoch": 0.19596490446710907,
+      "grad_norm": 0.09104909747838974,
+      "learning_rate": 3.6328058530702476e-05,
+      "loss": 0.0611,
+      "step": 4400
+    },
+    {
+      "epoch": 0.19707834142430855,
+      "grad_norm": 13.756211280822754,
+      "learning_rate": 3.6287558430845775e-05,
+      "loss": 0.1169,
+      "step": 4425
+    },
+    {
+      "epoch": 0.19819177838150803,
+      "grad_norm": 0.013652303256094456,
+      "learning_rate": 3.624685904050873e-05,
+      "loss": 0.0528,
+      "step": 4450
+    },
+    {
+      "epoch": 0.19930521533870751,
+      "grad_norm": 0.36829623579978943,
+      "learning_rate": 3.620596085767887e-05,
+      "loss": 0.0996,
+      "step": 4475
+    },
+    {
+      "epoch": 0.200418652295907,
+      "grad_norm": 0.16198916733264923,
+      "learning_rate": 3.616486438277612e-05,
+      "loss": 0.074,
+      "step": 4500
+    },
+    {
+      "epoch": 0.20153208925310648,
+      "grad_norm": 0.01803933084011078,
+      "learning_rate": 3.6123570118646635e-05,
+      "loss": 0.1303,
+      "step": 4525
+    },
+    {
+      "epoch": 0.20264552621030596,
+      "grad_norm": 11.962908744812012,
+      "learning_rate": 3.608207857055667e-05,
+      "loss": 0.1169,
+      "step": 4550
+    },
+    {
+      "epoch": 0.20375896316750547,
+      "grad_norm": 0.08494125306606293,
+      "learning_rate": 3.60403902461864e-05,
+      "loss": 0.1349,
+      "step": 4575
+    },
+    {
+      "epoch": 0.20487240012470495,
+      "grad_norm": 0.7877364158630371,
+      "learning_rate": 3.599850565562372e-05,
+      "loss": 0.108,
+      "step": 4600
+    },
+    {
+      "epoch": 0.20598583708190443,
+      "grad_norm": 90.93081665039062,
+      "learning_rate": 3.595642531135796e-05,
+      "loss": 0.1446,
+      "step": 4625
+    },
+    {
+      "epoch": 0.20709927403910391,
+      "grad_norm": 0.3855713903903961,
+      "learning_rate": 3.591414972827368e-05,
+      "loss": 0.0711,
+      "step": 4650
+    },
+    {
+      "epoch": 0.2082127109963034,
+      "grad_norm": 0.07253967225551605,
+      "learning_rate": 3.587167942364429e-05,
+      "loss": 0.0954,
+      "step": 4675
+    },
+    {
+      "epoch": 0.20932614795350288,
+      "grad_norm": 11.73831558227539,
+      "learning_rate": 3.58290149171258e-05,
+      "loss": 0.0758,
+      "step": 4700
+    },
+    {
+      "epoch": 0.21043958491070236,
+      "grad_norm": 0.1233116090297699,
+      "learning_rate": 3.5786156730750414e-05,
+      "loss": 0.0584,
+      "step": 4725
+    },
+    {
+      "epoch": 0.21155302186790184,
+      "grad_norm": 0.07078699767589569,
+      "learning_rate": 3.5743105388920145e-05,
+      "loss": 0.0915,
+      "step": 4750
+    },
+    {
+      "epoch": 0.21266645882510132,
+      "grad_norm": 19.74798583984375,
+      "learning_rate": 3.5699861418400415e-05,
+      "loss": 0.1626,
+      "step": 4775
+    },
+    {
+      "epoch": 0.2137798957823008,
+      "grad_norm": 0.04195041209459305,
+      "learning_rate": 3.565642534831361e-05,
+      "loss": 0.0012,
+      "step": 4800
+    },
+    {
+      "epoch": 0.2148933327395003,
+      "grad_norm": 75.83612060546875,
+      "learning_rate": 3.5612797710132586e-05,
+      "loss": 0.1313,
+      "step": 4825
+    },
+    {
+      "epoch": 0.21600676969669977,
+      "grad_norm": 0.07214100658893585,
+      "learning_rate": 3.556897903767419e-05,
+      "loss": 0.1026,
+      "step": 4850
+    },
+    {
+      "epoch": 0.21712020665389925,
+      "grad_norm": 0.12980283796787262,
+      "learning_rate": 3.552496986709269e-05,
+      "loss": 0.0828,
+      "step": 4875
+    },
+    {
+      "epoch": 0.21823364361109873,
+      "grad_norm": 0.013344530016183853,
+      "learning_rate": 3.5480770736873275e-05,
+      "loss": 0.083,
+      "step": 4900
+    },
+    {
+      "epoch": 0.2193470805682982,
+      "grad_norm": 3.2380993366241455,
+      "learning_rate": 3.5436382187825425e-05,
+      "loss": 0.1372,
+      "step": 4925
+    },
+    {
+      "epoch": 0.2204605175254977,
+      "grad_norm": 0.045397862792015076,
+      "learning_rate": 3.5391804763076275e-05,
+      "loss": 0.0526,
+      "step": 4950
+    },
+    {
+      "epoch": 0.22157395448269718,
+      "grad_norm": 0.09334590286016464,
+      "learning_rate": 3.534703900806404e-05,
+      "loss": 0.1112,
+      "step": 4975
+    },
+    {
+      "epoch": 0.22268739143989666,
+      "grad_norm": 0.02267163060605526,
+      "learning_rate": 3.530208547053124e-05,
+      "loss": 0.121,
+      "step": 5000
+    },
+    {
+      "epoch": 0.22380082839709617,
+      "grad_norm": 0.24515791237354279,
+      "learning_rate": 3.5256944700518106e-05,
+      "loss": 0.0859,
+      "step": 5025
+    },
+    {
+      "epoch": 0.22491426535429565,
+      "grad_norm": 0.4689369201660156,
+      "learning_rate": 3.521161725035579e-05,
+      "loss": 0.107,
+      "step": 5050
+    },
+    {
+      "epoch": 0.22602770231149513,
+      "grad_norm": 0.13337279856204987,
+      "learning_rate": 3.5166103674659584e-05,
+      "loss": 0.071,
+      "step": 5075
+    },
+    {
+      "epoch": 0.22714113926869461,
+      "grad_norm": 67.71549224853516,
+      "learning_rate": 3.512040453032219e-05,
+      "loss": 0.0726,
+      "step": 5100
+    },
+    {
+      "epoch": 0.2282545762258941,
+      "grad_norm": 0.20219382643699646,
+      "learning_rate": 3.507452037650689e-05,
+      "loss": 0.1498,
+      "step": 5125
+    },
+    {
+      "epoch": 0.22936801318309358,
+      "grad_norm": 0.007606521248817444,
+      "learning_rate": 3.502845177464068e-05,
+      "loss": 0.0418,
+      "step": 5150
+    },
+    {
+      "epoch": 0.23048145014029306,
+      "grad_norm": 0.3652168810367584,
+      "learning_rate": 3.49821992884074e-05,
+      "loss": 0.0834,
+      "step": 5175
+    },
+    {
+      "epoch": 0.23159488709749254,
+      "grad_norm": 11.301214218139648,
+      "learning_rate": 3.49357634837409e-05,
+      "loss": 0.0726,
+      "step": 5200
+    },
+    {
+      "epoch": 0.23270832405469202,
+      "grad_norm": 9.451020240783691,
+      "learning_rate": 3.4889144928818015e-05,
+      "loss": 0.1142,
+      "step": 5225
+    },
+    {
+      "epoch": 0.2338217610118915,
+      "grad_norm": 0.5295103192329407,
+      "learning_rate": 3.4842344194051715e-05,
+      "loss": 0.099,
+      "step": 5250
+    },
+    {
+      "epoch": 0.234935197969091,
+      "grad_norm": 0.1344202160835266,
+      "learning_rate": 3.479536185208404e-05,
+      "loss": 0.1088,
+      "step": 5275
+    },
+    {
+      "epoch": 0.23604863492629047,
+      "grad_norm": 0.0730552077293396,
+      "learning_rate": 3.474819847777915e-05,
+      "loss": 0.0902,
+      "step": 5300
+    },
+    {
+      "epoch": 0.23716207188348995,
+      "grad_norm": 22.340232849121094,
+      "learning_rate": 3.470085464821626e-05,
+      "loss": 0.0695,
+      "step": 5325
+    },
+    {
+      "epoch": 0.23827550884068943,
+      "grad_norm": 0.012485889717936516,
+      "learning_rate": 3.46533309426826e-05,
+      "loss": 0.1368,
+      "step": 5350
+    },
+    {
+      "epoch": 0.2393889457978889,
+      "grad_norm": 0.07602645456790924,
+      "learning_rate": 3.460562794266631e-05,
+      "loss": 0.0963,
+      "step": 5375
+    },
+    {
+      "epoch": 0.2405023827550884,
+      "grad_norm": 0.028351513668894768,
+      "learning_rate": 3.455774623184933e-05,
+      "loss": 0.0761,
+      "step": 5400
+    },
+    {
+      "epoch": 0.24161581971228788,
+      "grad_norm": 0.04493757709860802,
+      "learning_rate": 3.450968639610028e-05,
+      "loss": 0.0967,
+      "step": 5425
+    },
+    {
+      "epoch": 0.2427292566694874,
+      "grad_norm": 0.04149246960878372,
+      "learning_rate": 3.4461449023467254e-05,
+      "loss": 0.0748,
+      "step": 5450
+    },
+    {
+      "epoch": 0.24384269362668687,
+      "grad_norm": 0.18419454991817474,
+      "learning_rate": 3.441303470417063e-05,
+      "loss": 0.0576,
+      "step": 5475
+    },
+    {
+      "epoch": 0.24495613058388635,
+      "grad_norm": 5.215043067932129,
+      "learning_rate": 3.4364444030595893e-05,
+      "loss": 0.0965,
+      "step": 5500
+    },
+    {
+      "epoch": 0.24606956754108583,
+      "grad_norm": 0.19195443391799927,
+      "learning_rate": 3.4315677597286344e-05,
+      "loss": 0.0912,
+      "step": 5525
+    },
+    {
+      "epoch": 0.2471830044982853,
+      "grad_norm": 0.12233854830265045,
+      "learning_rate": 3.4266736000935816e-05,
+      "loss": 0.0968,
+      "step": 5550
+    },
+    {
+      "epoch": 0.2482964414554848,
+      "grad_norm": 0.017935922369360924,
+      "learning_rate": 3.4217619840381416e-05,
+      "loss": 0.0432,
+      "step": 5575
+    },
+    {
+      "epoch": 0.24940987841268428,
+      "grad_norm": 0.3018956780433655,
+      "learning_rate": 3.4168329716596164e-05,
+      "loss": 0.08,
+      "step": 5600
+    },
+    {
+      "epoch": 0.25052331536988376,
+      "grad_norm": 71.5376968383789,
+      "learning_rate": 3.411886623268166e-05,
+      "loss": 0.1331,
+      "step": 5625
+    },
+    {
+      "epoch": 0.2516367523270832,
+      "grad_norm": 0.12258598953485489,
+      "learning_rate": 3.406922999386069e-05,
+      "loss": 0.074,
+      "step": 5650
+    },
+    {
+      "epoch": 0.2527501892842827,
+      "grad_norm": 20.980422973632812,
+      "learning_rate": 3.401942160746981e-05,
+      "loss": 0.1432,
+      "step": 5675
+    },
+    {
+      "epoch": 0.25386362624148223,
+      "grad_norm": 0.22452101111412048,
+      "learning_rate": 3.396944168295195e-05,
+      "loss": 0.0806,
+      "step": 5700
+    },
+    {
+      "epoch": 0.2549770631986817,
+      "grad_norm": 1.4218170642852783,
+      "learning_rate": 3.391929083184895e-05,
+      "loss": 0.0626,
+      "step": 5725
+    },
+    {
+      "epoch": 0.2560905001558812,
+      "grad_norm": 10.604058265686035,
+      "learning_rate": 3.3868969667794025e-05,
+      "loss": 0.106,
+      "step": 5750
+    },
+    {
+      "epoch": 0.25720393711308065,
+      "grad_norm": 0.07359682023525238,
+      "learning_rate": 3.381847880650433e-05,
+      "loss": 0.0403,
+      "step": 5775
+    },
+    {
+      "epoch": 0.25831737407028016,
+      "grad_norm": 0.10821941494941711,
+      "learning_rate": 3.3767818865773376e-05,
+      "loss": 0.0533,
+      "step": 5800
+    },
+    {
+      "epoch": 0.2594308110274796,
+      "grad_norm": 8.280586242675781,
+      "learning_rate": 3.37169904654635e-05,
+      "loss": 0.0945,
+      "step": 5825
+    },
+    {
+      "epoch": 0.2605442479846791,
+      "grad_norm": 1.2354644536972046,
+      "learning_rate": 3.366599422749825e-05,
+      "loss": 0.1221,
+      "step": 5850
+    },
+    {
+      "epoch": 0.2616576849418786,
+      "grad_norm": 1.7320414781570435,
+      "learning_rate": 3.361483077585482e-05,
+      "loss": 0.0921,
+      "step": 5875
+    },
+    {
+      "epoch": 0.2627711218990781,
+      "grad_norm": 1.0141493082046509,
+      "learning_rate": 3.356350073655636e-05,
+      "loss": 0.0727,
+      "step": 5900
+    },
+    {
+      "epoch": 0.26388455885627754,
+      "grad_norm": 8.444538116455078,
+      "learning_rate": 3.3512004737664376e-05,
+      "loss": 0.1648,
+      "step": 5925
+    },
+    {
+      "epoch": 0.26499799581347705,
+      "grad_norm": 7.873695373535156,
+      "learning_rate": 3.346034340927097e-05,
+      "loss": 0.0527,
+      "step": 5950
+    },
+    {
+      "epoch": 0.2661114327706765,
+      "grad_norm": 1.0561110973358154,
+      "learning_rate": 3.340851738349122e-05,
+      "loss": 0.0841,
+      "step": 5975
+    },
+    {
+      "epoch": 0.267224869727876,
+      "grad_norm": 0.1629607230424881,
+      "learning_rate": 3.335652729445538e-05,
+      "loss": 0.0611,
+      "step": 6000
+    },
+    {
+      "epoch": 0.26833830668507547,
+      "grad_norm": 0.042299434542655945,
+      "learning_rate": 3.330437377830113e-05,
+      "loss": 0.1361,
+      "step": 6025
+    },
+    {
+      "epoch": 0.269451743642275,
+      "grad_norm": 0.07904113829135895,
+      "learning_rate": 3.325205747316582e-05,
+      "loss": 0.0895,
+      "step": 6050
+    },
+    {
+      "epoch": 0.27056518059947443,
+      "grad_norm": 0.04022669047117233,
+      "learning_rate": 3.3199579019178644e-05,
+      "loss": 0.0667,
+      "step": 6075
+    },
+    {
+      "epoch": 0.27167861755667394,
+      "grad_norm": 94.2168960571289,
+      "learning_rate": 3.31469390584528e-05,
+      "loss": 0.1373,
+      "step": 6100
+    },
+    {
+      "epoch": 0.27279205451387345,
+      "grad_norm": 23.157451629638672,
+      "learning_rate": 3.309413823507764e-05,
+      "loss": 0.0442,
+      "step": 6125
+    },
+    {
+      "epoch": 0.2739054914710729,
+      "grad_norm": 0.027611056342720985,
+      "learning_rate": 3.3041177195110816e-05,
+      "loss": 0.0808,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2750189284282724,
+      "grad_norm": 3.2276134490966797,
+      "learning_rate": 3.298805658657031e-05,
+      "loss": 0.132,
+      "step": 6175
+    },
+    {
+      "epoch": 0.27613236538547187,
+      "grad_norm": 0.03948564827442169,
+      "learning_rate": 3.2934777059426584e-05,
+      "loss": 0.1221,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2772458023426714,
+      "grad_norm": 0.2750093638896942,
+      "learning_rate": 3.288133926559456e-05,
+      "loss": 0.103,
+      "step": 6225
+    },
+    {
+      "epoch": 0.27835923929987083,
+      "grad_norm": 0.9560427069664001,
+      "learning_rate": 3.2827743858925685e-05,
+      "loss": 0.0519,
+      "step": 6250
+    },
+    {
+      "epoch": 0.27947267625707034,
+      "grad_norm": 0.11851734668016434,
+      "learning_rate": 3.2773991495199913e-05,
+      "loss": 0.1086,
+      "step": 6275
+    },
+    {
+      "epoch": 0.2805861132142698,
+      "grad_norm": 7.182758808135986,
+      "learning_rate": 3.272008283211769e-05,
+      "loss": 0.056,
+      "step": 6300
+    },
+    {
+      "epoch": 0.2816995501714693,
+      "grad_norm": 0.04427009075880051,
+      "learning_rate": 3.266601852929189e-05,
+      "loss": 0.0858,
+      "step": 6325
+    },
+    {
+      "epoch": 0.28281298712866876,
+      "grad_norm": 0.735414445400238,
+      "learning_rate": 3.261179924823978e-05,
+      "loss": 0.0891,
+      "step": 6350
+    },
+    {
+      "epoch": 0.28392642408586827,
+      "grad_norm": 0.01727837324142456,
+      "learning_rate": 3.255742565237487e-05,
+      "loss": 0.0403,
+      "step": 6375
+    },
+    {
+      "epoch": 0.2850398610430677,
+      "grad_norm": 8.384565353393555,
+      "learning_rate": 3.250289840699885e-05,
+      "loss": 0.1248,
+      "step": 6400
+    },
+    {
+      "epoch": 0.28615329800026723,
+      "grad_norm": 19.30202865600586,
+      "learning_rate": 3.244821817929342e-05,
+      "loss": 0.078,
+      "step": 6425
+    },
+    {
+      "epoch": 0.2872667349574667,
+      "grad_norm": 0.05216224864125252,
+      "learning_rate": 3.239338563831213e-05,
+      "loss": 0.1063,
+      "step": 6450
+    },
+    {
+      "epoch": 0.2883801719146662,
+      "grad_norm": 0.3387221097946167,
+      "learning_rate": 3.23384014549722e-05,
+      "loss": 0.0174,
+      "step": 6475
+    },
+    {
+      "epoch": 0.28949360887186565,
+      "grad_norm": 0.050413765013217926,
+      "learning_rate": 3.228326630204632e-05,
+      "loss": 0.1149,
+      "step": 6500
+    },
+    {
+      "epoch": 0.29060704582906516,
+      "grad_norm": 0.05692288279533386,
+      "learning_rate": 3.2227980854154376e-05,
+      "loss": 0.0229,
+      "step": 6525
+    },
+    {
+      "epoch": 0.29172048278626467,
+      "grad_norm": 0.2883279323577881,
+      "learning_rate": 3.217254578775525e-05,
+      "loss": 0.0801,
+      "step": 6550
+    },
+    {
+      "epoch": 0.2928339197434641,
+      "grad_norm": 0.028399428352713585,
+      "learning_rate": 3.2116961781138525e-05,
+      "loss": 0.0652,
+      "step": 6575
+    },
+    {
+      "epoch": 0.29394735670066363,
+      "grad_norm": 18.74675750732422,
+      "learning_rate": 3.206122951441615e-05,
+      "loss": 0.1657,
+      "step": 6600
+    },
+    {
+      "epoch": 0.2950607936578631,
+      "grad_norm": 0.3412401080131531,
+      "learning_rate": 3.200534966951418e-05,
+      "loss": 0.1038,
+      "step": 6625
+    },
+    {
+      "epoch": 0.2961742306150626,
+      "grad_norm": 0.29099151492118835,
+      "learning_rate": 3.194932293016437e-05,
+      "loss": 0.0817,
+      "step": 6650
+    },
+    {
+      "epoch": 0.29728766757226205,
+      "grad_norm": 0.16885584592819214,
+      "learning_rate": 3.189314998189586e-05,
+      "loss": 0.0756,
+      "step": 6675
+    },
+    {
+      "epoch": 0.29840110452946156,
+      "grad_norm": 0.11662675440311432,
+      "learning_rate": 3.1836831512026746e-05,
+      "loss": 0.1049,
+      "step": 6700
+    },
+    {
+      "epoch": 0.299514541486661,
+      "grad_norm": 10.235050201416016,
+      "learning_rate": 3.1780368209655715e-05,
+      "loss": 0.0625,
+      "step": 6725
+    },
+    {
+      "epoch": 0.3006279784438605,
+      "grad_norm": 0.013582950457930565,
+      "learning_rate": 3.1723760765653566e-05,
+      "loss": 0.0811,
+      "step": 6750
+    },
+    {
+      "epoch": 0.30174141540106,
+      "grad_norm": 0.028899747878313065,
+      "learning_rate": 3.1667009872654804e-05,
+      "loss": 0.1065,
+      "step": 6775
+    },
+    {
+      "epoch": 0.3028548523582595,
+      "grad_norm": 0.2777616083621979,
+      "learning_rate": 3.1610116225049107e-05,
+      "loss": 0.0882,
+      "step": 6800
+    },
+    {
+      "epoch": 0.30396828931545894,
+      "grad_norm": 0.2561856806278229,
+      "learning_rate": 3.155308051897287e-05,
+      "loss": 0.0876,
+      "step": 6825
+    },
+    {
+      "epoch": 0.30508172627265845,
+      "grad_norm": 0.0948658362030983,
+      "learning_rate": 3.149590345230072e-05,
+      "loss": 0.153,
+      "step": 6850
+    },
+    {
+      "epoch": 0.3061951632298579,
+      "grad_norm": 0.28858181834220886,
+      "learning_rate": 3.143858572463688e-05,
+      "loss": 0.125,
+      "step": 6875
+    },
+    {
+      "epoch": 0.3073086001870574,
+      "grad_norm": 1.5904790163040161,
+      "learning_rate": 3.1381128037306706e-05,
+      "loss": 0.0845,
+      "step": 6900
+    },
+    {
+      "epoch": 0.30842203714425687,
+      "grad_norm": 7.804311275482178,
+      "learning_rate": 3.1323531093348036e-05,
+      "loss": 0.0496,
+      "step": 6925
+    },
+    {
+      "epoch": 0.3095354741014564,
+      "grad_norm": 1.7563951015472412,
+      "learning_rate": 3.126579559750265e-05,
+      "loss": 0.1077,
+      "step": 6950
+    },
+    {
+      "epoch": 0.3106489110586559,
+      "grad_norm": 11.621973991394043,
+      "learning_rate": 3.120792225620759e-05,
+      "loss": 0.0678,
+      "step": 6975
+    },
+    {
+      "epoch": 0.31176234801585534,
+      "grad_norm": 54.6142692565918,
+      "learning_rate": 3.1149911777586533e-05,
+      "loss": 0.1449,
+      "step": 7000
+    },
+    {
+      "epoch": 0.31287578497305485,
+      "grad_norm": 0.6821468472480774,
+      "learning_rate": 3.109176487144116e-05,
+      "loss": 0.0686,
+      "step": 7025
+    },
+    {
+      "epoch": 0.3139892219302543,
+      "grad_norm": 0.13249638676643372,
+      "learning_rate": 3.103348224924244e-05,
+      "loss": 0.1183,
+      "step": 7050
+    },
+    {
+      "epoch": 0.3151026588874538,
+      "grad_norm": 0.16593268513679504,
+      "learning_rate": 3.09750646241219e-05,
+      "loss": 0.126,
+      "step": 7075
+    },
+    {
+      "epoch": 0.31621609584465327,
+      "grad_norm": 0.07781495153903961,
+      "learning_rate": 3.091651271086297e-05,
+      "loss": 0.0662,
+      "step": 7100
+    },
+    {
+      "epoch": 0.3173295328018528,
+      "grad_norm": 14.361419677734375,
+      "learning_rate": 3.085782722589217e-05,
+      "loss": 0.049,
+      "step": 7125
+    },
+    {
+      "epoch": 0.31844296975905223,
+      "grad_norm": 0.07331876456737518,
+      "learning_rate": 3.0799008887270376e-05,
+      "loss": 0.1351,
+      "step": 7150
+    },
+    {
+      "epoch": 0.31955640671625174,
+      "grad_norm": 0.1390572488307953,
+      "learning_rate": 3.074005841468403e-05,
+      "loss": 0.0748,
+      "step": 7175
+    },
+    {
+      "epoch": 0.3206698436734512,
+      "grad_norm": 0.049495305866003036,
+      "learning_rate": 3.068097652943633e-05,
+      "loss": 0.1004,
+      "step": 7200
+    },
+    {
+      "epoch": 0.3217832806306507,
+      "grad_norm": 12.76275634765625,
+      "learning_rate": 3.0621763954438417e-05,
+      "loss": 0.1007,
+      "step": 7225
+    },
+    {
+      "epoch": 0.32289671758785016,
+      "grad_norm": 13.835783958435059,
+      "learning_rate": 3.0562421414200495e-05,
+      "loss": 0.0994,
+      "step": 7250
+    },
+    {
+      "epoch": 0.32401015454504967,
+      "grad_norm": 11.560086250305176,
+      "learning_rate": 3.0502949634823004e-05,
+      "loss": 0.1399,
+      "step": 7275
+    },
+    {
+      "epoch": 0.3251235915022491,
+      "grad_norm": 125.53492736816406,
+      "learning_rate": 3.0443349343987728e-05,
+      "loss": 0.0669,
+      "step": 7300
+    },
+    {
+      "epoch": 0.32623702845944863,
+      "grad_norm": 0.15544213354587555,
+      "learning_rate": 3.0383621270948888e-05,
+      "loss": 0.0814,
+      "step": 7325
+    },
+    {
+      "epoch": 0.3273504654166481,
+      "grad_norm": 0.24035079777240753,
+      "learning_rate": 3.032376614652419e-05,
+      "loss": 0.1366,
+      "step": 7350
+    },
+    {
+      "epoch": 0.3284639023738476,
+      "grad_norm": 0.06057662144303322,
+      "learning_rate": 3.0263784703085935e-05,
+      "loss": 0.0574,
+      "step": 7375
+    },
+    {
+      "epoch": 0.3295773393310471,
+      "grad_norm": 0.25817441940307617,
+      "learning_rate": 3.0203677674552022e-05,
+      "loss": 0.0275,
+      "step": 7400
+    },
+    {
+      "epoch": 0.33069077628824656,
+      "grad_norm": 0.24544256925582886,
+      "learning_rate": 3.014344579637698e-05,
+      "loss": 0.0903,
+      "step": 7425
+    },
+    {
+      "epoch": 0.33180421324544607,
+      "grad_norm": 17.432865142822266,
+      "learning_rate": 3.0083089805542963e-05,
+      "loss": 0.1184,
+      "step": 7450
+    },
+    {
+      "epoch": 0.3329176502026455,
+      "grad_norm": 16.43193244934082,
+      "learning_rate": 3.0022610440550733e-05,
+      "loss": 0.1503,
+      "step": 7475
+    },
+    {
+      "epoch": 0.33403108715984503,
+      "grad_norm": 0.01428154855966568,
+      "learning_rate": 2.9962008441410645e-05,
+      "loss": 0.1138,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3351445241170445,
+      "grad_norm": 0.045094311237335205,
+      "learning_rate": 2.9901284549633554e-05,
+      "loss": 0.0755,
+      "step": 7525
+    },
+    {
+      "epoch": 0.336257961074244,
+      "grad_norm": 16.375850677490234,
+      "learning_rate": 2.9840439508221774e-05,
+      "loss": 0.1069,
+      "step": 7550
+    },
+    {
+      "epoch": 0.33737139803144345,
+      "grad_norm": 0.2958405911922455,
+      "learning_rate": 2.9779474061659974e-05,
+      "loss": 0.1006,
+      "step": 7575
+    },
+    {
+      "epoch": 0.33848483498864296,
+      "grad_norm": 0.21808381378650665,
+      "learning_rate": 2.9718388955906063e-05,
+      "loss": 0.1078,
+      "step": 7600
+    },
+    {
+      "epoch": 0.3395982719458424,
+      "grad_norm": 0.09185987710952759,
+      "learning_rate": 2.9657184938382087e-05,
+      "loss": 0.0668,
+      "step": 7625
+    },
+    {
+      "epoch": 0.3407117089030419,
+      "grad_norm": 0.903226912021637,
+      "learning_rate": 2.9595862757965043e-05,
+      "loss": 0.1102,
+      "step": 7650
+    },
+    {
+      "epoch": 0.3418251458602414,
+      "grad_norm": 0.2668018937110901,
+      "learning_rate": 2.9534423164977765e-05,
+      "loss": 0.1009,
+      "step": 7675
+    },
+    {
+      "epoch": 0.3429385828174409,
+      "grad_norm": 0.26890435814857483,
+      "learning_rate": 2.9472866911179695e-05,
+      "loss": 0.06,
+      "step": 7700
+    },
+    {
+      "epoch": 0.34405201977464034,
+      "grad_norm": 17.915388107299805,
+      "learning_rate": 2.941119474975772e-05,
+      "loss": 0.1073,
+      "step": 7725
+    },
+    {
+      "epoch": 0.34516545673183985,
+      "grad_norm": 0.05032012239098549,
+      "learning_rate": 2.934940743531694e-05,
+      "loss": 0.0772,
+      "step": 7750
+    },
+    {
+      "epoch": 0.3462788936890393,
+      "grad_norm": 0.38257360458374023,
+      "learning_rate": 2.928750572387144e-05,
+      "loss": 0.1137,
+      "step": 7775
+    },
+    {
+      "epoch": 0.3473923306462388,
+      "grad_norm": 2.620697498321533,
+      "learning_rate": 2.9225490372835043e-05,
+      "loss": 0.1189,
+      "step": 7800
+    },
+    {
+      "epoch": 0.34850576760343827,
+      "grad_norm": 1.1211553812026978,
+      "learning_rate": 2.9163362141012024e-05,
+      "loss": 0.0608,
+      "step": 7825
+    },
+    {
+      "epoch": 0.3496192045606378,
+      "grad_norm": 11.626333236694336,
+      "learning_rate": 2.9101121788587846e-05,
+      "loss": 0.0781,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3507326415178373,
+      "grad_norm": 0.05085618793964386,
+      "learning_rate": 2.9038770077119865e-05,
+      "loss": 0.0555,
+      "step": 7875
+    },
+    {
+      "epoch": 0.35184607847503674,
+      "grad_norm": 0.007242363411933184,
+      "learning_rate": 2.8976307769527974e-05,
+      "loss": 0.0769,
+      "step": 7900
+    },
+    {
+      "epoch": 0.35295951543223625,
+      "grad_norm": 0.018498949706554413,
+      "learning_rate": 2.8913735630085305e-05,
+      "loss": 0.0516,
+      "step": 7925
+    },
+    {
+      "epoch": 0.3540729523894357,
+      "grad_norm": 0.01721395179629326,
+      "learning_rate": 2.885105442440887e-05,
+      "loss": 0.1599,
+      "step": 7950
+    },
+    {
+      "epoch": 0.3551863893466352,
+      "grad_norm": 12.357562065124512,
+      "learning_rate": 2.878826491945018e-05,
+      "loss": 0.0521,
+      "step": 7975
+    },
+    {
+      "epoch": 0.35629982630383467,
+      "grad_norm": 0.11994218826293945,
+      "learning_rate": 2.872536788348587e-05,
+      "loss": 0.0558,
+      "step": 8000
+    },
+    {
+      "epoch": 0.3574132632610342,
+      "grad_norm": 6.6094536781311035,
+      "learning_rate": 2.8662364086108302e-05,
+      "loss": 0.1597,
+      "step": 8025
+    },
+    {
+      "epoch": 0.35852670021823363,
+      "grad_norm": 1.3313615322113037,
+      "learning_rate": 2.8599254298216136e-05,
+      "loss": 0.1593,
+      "step": 8050
+    },
+    {
+      "epoch": 0.35964013717543314,
+      "grad_norm": 0.2207033932209015,
+      "learning_rate": 2.853603929200491e-05,
+      "loss": 0.0781,
+      "step": 8075
+    },
+    {
+      "epoch": 0.3607535741326326,
+      "grad_norm": 0.11759678274393082,
+      "learning_rate": 2.847271984095759e-05,
+      "loss": 0.0513,
+      "step": 8100
+    },
+    {
+      "epoch": 0.3618670110898321,
+      "grad_norm": 0.1604037880897522,
+      "learning_rate": 2.84092967198351e-05,
+      "loss": 0.1056,
+      "step": 8125
+    },
+    {
+      "epoch": 0.36298044804703156,
+      "grad_norm": 0.05803081393241882,
+      "learning_rate": 2.834577070466684e-05,
+      "loss": 0.0579,
+      "step": 8150
+    },
+    {
+      "epoch": 0.36409388500423107,
+      "grad_norm": 0.8341618180274963,
+      "learning_rate": 2.8282142572741205e-05,
+      "loss": 0.0779,
+      "step": 8175
+    },
+    {
+      "epoch": 0.3652073219614305,
+      "grad_norm": 0.24928626418113708,
+      "learning_rate": 2.821841310259606e-05,
+      "loss": 0.1118,
+      "step": 8200
+    },
+    {
+      "epoch": 0.36632075891863003,
+      "grad_norm": 92.14579010009766,
+      "learning_rate": 2.8154583074009216e-05,
+      "loss": 0.0581,
+      "step": 8225
+    },
+    {
+      "epoch": 0.3674341958758295,
+      "grad_norm": 0.010704785585403442,
+      "learning_rate": 2.80906532679889e-05,
+      "loss": 0.0959,
+      "step": 8250
+    },
+    {
+      "epoch": 0.368547632833029,
+      "grad_norm": 0.05830851569771767,
+      "learning_rate": 2.802662446676418e-05,
+      "loss": 0.0383,
+      "step": 8275
+    },
+    {
+      "epoch": 0.3696610697902285,
+      "grad_norm": 14.7666597366333,
+      "learning_rate": 2.796249745377541e-05,
+      "loss": 0.0381,
+      "step": 8300
+    },
+    {
+      "epoch": 0.37077450674742796,
+      "grad_norm": 0.33050060272216797,
+      "learning_rate": 2.7898273013664657e-05,
+      "loss": 0.1475,
+      "step": 8325
+    },
+    {
+      "epoch": 0.37188794370462747,
+      "grad_norm": 0.006766165606677532,
+      "learning_rate": 2.7833951932266045e-05,
+      "loss": 0.1048,
+      "step": 8350
+    },
+    {
+      "epoch": 0.3730013806618269,
+      "grad_norm": 57.38689422607422,
+      "learning_rate": 2.7769534996596196e-05,
+      "loss": 0.0475,
+      "step": 8375
+    },
+    {
+      "epoch": 0.37411481761902643,
+      "grad_norm": 0.013443752191960812,
+      "learning_rate": 2.7705022994844588e-05,
+      "loss": 0.0451,
+      "step": 8400
+    },
+    {
+      "epoch": 0.3752282545762259,
+      "grad_norm": 21.72231674194336,
+      "learning_rate": 2.7640416716363896e-05,
+      "loss": 0.1466,
+      "step": 8425
+    },
+    {
+      "epoch": 0.3763416915334254,
+      "grad_norm": 7.488071918487549,
+      "learning_rate": 2.7575716951660335e-05,
+      "loss": 0.1379,
+      "step": 8450
+    },
+    {
+      "epoch": 0.37745512849062485,
+      "grad_norm": 0.05108782649040222,
+      "learning_rate": 2.7510924492384003e-05,
+      "loss": 0.0835,
+      "step": 8475
+    },
+    {
+      "epoch": 0.37856856544782436,
+      "grad_norm": 1.109173059463501,
+      "learning_rate": 2.7446040131319188e-05,
+      "loss": 0.1243,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3796820024050238,
+      "grad_norm": 1.1221072673797607,
+      "learning_rate": 2.7381064662374655e-05,
+      "loss": 0.1334,
+      "step": 8525
+    },
+    {
+      "epoch": 0.3807954393622233,
+      "grad_norm": 16.271162033081055,
+      "learning_rate": 2.7315998880573943e-05,
+      "loss": 0.0482,
+      "step": 8550
+    },
+    {
+      "epoch": 0.3819088763194228,
+      "grad_norm": 0.41395649313926697,
+      "learning_rate": 2.7250843582045653e-05,
+      "loss": 0.0814,
+      "step": 8575
+    },
+    {
+      "epoch": 0.3830223132766223,
+      "grad_norm": 0.26396995782852173,
+      "learning_rate": 2.7185599564013666e-05,
+      "loss": 0.0503,
+      "step": 8600
+    },
+    {
+      "epoch": 0.38413575023382174,
+      "grad_norm": 0.15254704654216766,
+      "learning_rate": 2.7120267624787435e-05,
+      "loss": 0.0801,
+      "step": 8625
+    },
+    {
+      "epoch": 0.38524918719102125,
+      "grad_norm": 0.46369659900665283,
+      "learning_rate": 2.7054848563752175e-05,
+      "loss": 0.1062,
+      "step": 8650
+    },
+    {
+      "epoch": 0.3863626241482207,
+      "grad_norm": 0.3597426116466522,
+      "learning_rate": 2.698934318135912e-05,
+      "loss": 0.1508,
+      "step": 8675
+    },
+    {
+      "epoch": 0.3874760611054202,
+      "grad_norm": 0.08787465840578079,
+      "learning_rate": 2.6923752279115693e-05,
+      "loss": 0.0501,
+      "step": 8700
+    },
+    {
+      "epoch": 0.3885894980626197,
+      "grad_norm": 14.235735893249512,
+      "learning_rate": 2.6858076659575724e-05,
+      "loss": 0.0919,
+      "step": 8725
+    },
+    {
+      "epoch": 0.3897029350198192,
+      "grad_norm": 0.059386011213064194,
+      "learning_rate": 2.6792317126329632e-05,
+      "loss": 0.0746,
+      "step": 8750
+    },
+    {
+      "epoch": 0.3908163719770187,
+      "grad_norm": 0.08174929022789001,
+      "learning_rate": 2.672647448399457e-05,
+      "loss": 0.0747,
+      "step": 8775
+    },
+    {
+      "epoch": 0.39192980893421814,
+      "grad_norm": 109.75809478759766,
+      "learning_rate": 2.66605495382046e-05,
+      "loss": 0.0878,
+      "step": 8800
+    },
+    {
+      "epoch": 0.39304324589141765,
+      "grad_norm": 21.897994995117188,
+      "learning_rate": 2.659454309560082e-05,
+      "loss": 0.043,
+      "step": 8825
+    },
+    {
+      "epoch": 0.3941566828486171,
+      "grad_norm": 0.1924513429403305,
+      "learning_rate": 2.652845596382152e-05,
+      "loss": 0.0445,
+      "step": 8850
+    },
+    {
+      "epoch": 0.3952701198058166,
+      "grad_norm": 10.962334632873535,
+      "learning_rate": 2.6462288951492255e-05,
+      "loss": 0.1423,
+      "step": 8875
+    },
+    {
+      "epoch": 0.39638355676301607,
+      "grad_norm": 1.092352271080017,
+      "learning_rate": 2.639604286821601e-05,
+      "loss": 0.0513,
+      "step": 8900
+    },
+    {
+      "epoch": 0.3974969937202156,
+      "grad_norm": 0.07695068418979645,
+      "learning_rate": 2.632971852456323e-05,
+      "loss": 0.0651,
+      "step": 8925
+    },
+    {
+      "epoch": 0.39861043067741503,
+      "grad_norm": 0.3159739077091217,
+      "learning_rate": 2.6263316732061967e-05,
+      "loss": 0.0808,
+      "step": 8950
+    },
+    {
+      "epoch": 0.39972386763461454,
+      "grad_norm": 0.11245528608560562,
+      "learning_rate": 2.619683830318789e-05,
+      "loss": 0.0989,
+      "step": 8975
+    },
+    {
+      "epoch": 0.400837304591814,
+      "grad_norm": 0.005162264686077833,
+      "learning_rate": 2.6130284051354394e-05,
+      "loss": 0.1038,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4019507415490135,
+      "grad_norm": 0.07931370288133621,
+      "learning_rate": 2.6063654790902605e-05,
+      "loss": 0.1288,
+      "step": 9025
+    },
+    {
+      "epoch": 0.40306417850621296,
+      "grad_norm": 0.1483098566532135,
+      "learning_rate": 2.599695133709145e-05,
+      "loss": 0.0582,
+      "step": 9050
+    },
+    {
+      "epoch": 0.40417761546341247,
+      "grad_norm": 5.938066482543945,
+      "learning_rate": 2.593017450608766e-05,
+      "loss": 0.1444,
+      "step": 9075
+    },
+    {
+      "epoch": 0.4052910524206119,
+      "grad_norm": 0.01066511683166027,
+      "learning_rate": 2.586332511495579e-05,
+      "loss": 0.0234,
+      "step": 9100
+    },
+    {
+      "epoch": 0.40640448937781143,
+      "grad_norm": 0.09881523251533508,
+      "learning_rate": 2.5796403981648232e-05,
+      "loss": 0.1064,
+      "step": 9125
+    },
+    {
+      "epoch": 0.40751792633501094,
+      "grad_norm": 61.86650466918945,
+      "learning_rate": 2.5729411924995176e-05,
+      "loss": 0.0709,
+      "step": 9150
+    },
+    {
+      "epoch": 0.4086313632922104,
+      "grad_norm": 0.13846437633037567,
+      "learning_rate": 2.5662349764694644e-05,
+      "loss": 0.0989,
+      "step": 9175
+    },
+    {
+      "epoch": 0.4097448002494099,
+      "grad_norm": 0.09358134865760803,
+      "learning_rate": 2.5595218321302404e-05,
+      "loss": 0.1336,
+      "step": 9200
+    },
+    {
+      "epoch": 0.41085823720660936,
+      "grad_norm": 0.14855454862117767,
+      "learning_rate": 2.5528018416221957e-05,
+      "loss": 0.1036,
+      "step": 9225
+    },
+    {
+      "epoch": 0.41197167416380887,
+      "grad_norm": 1.0118060111999512,
+      "learning_rate": 2.5460750871694496e-05,
+      "loss": 0.1035,
+      "step": 9250
+    },
+    {
+      "epoch": 0.4130851111210083,
+      "grad_norm": 0.2535758316516876,
+      "learning_rate": 2.5393416510788833e-05,
+      "loss": 0.0889,
+      "step": 9275
+    },
+    {
+      "epoch": 0.41419854807820783,
+      "grad_norm": 13.825139045715332,
+      "learning_rate": 2.5326016157391307e-05,
+      "loss": 0.1798,
+      "step": 9300
+    },
+    {
+      "epoch": 0.4153119850354073,
+      "grad_norm": 0.3431245684623718,
+      "learning_rate": 2.5258550636195746e-05,
+      "loss": 0.0251,
+      "step": 9325
+    },
+    {
+      "epoch": 0.4164254219926068,
+      "grad_norm": 0.01225372590124607,
+      "learning_rate": 2.519102077269335e-05,
+      "loss": 0.0519,
+      "step": 9350
+    },
+    {
+      "epoch": 0.41753885894980625,
+      "grad_norm": 0.027721190825104713,
+      "learning_rate": 2.512342739316259e-05,
+      "loss": 0.0958,
+      "step": 9375
+    },
+    {
+      "epoch": 0.41865229590700576,
+      "grad_norm": 0.017482172697782516,
+      "learning_rate": 2.5055771324659105e-05,
+      "loss": 0.0938,
+      "step": 9400
+    },
+    {
+      "epoch": 0.4197657328642052,
+      "grad_norm": 47.62452697753906,
+      "learning_rate": 2.4988053395005592e-05,
+      "loss": 0.0447,
+      "step": 9425
+    },
+    {
+      "epoch": 0.4208791698214047,
+      "grad_norm": 20.639039993286133,
+      "learning_rate": 2.4920274432781647e-05,
+      "loss": 0.0708,
+      "step": 9450
+    },
+    {
+      "epoch": 0.4219926067786042,
+      "grad_norm": 0.03869742900133133,
+      "learning_rate": 2.485243526731366e-05,
+      "loss": 0.0563,
+      "step": 9475
+    },
+    {
+      "epoch": 0.4231060437358037,
+      "grad_norm": 0.23518817126750946,
+      "learning_rate": 2.478453672866464e-05,
+      "loss": 0.0858,
+      "step": 9500
+    },
+    {
+      "epoch": 0.42421948069300314,
+      "grad_norm": 0.17357154190540314,
+      "learning_rate": 2.4716579647624084e-05,
+      "loss": 0.0794,
+      "step": 9525
+    },
+    {
+      "epoch": 0.42533291765020265,
+      "grad_norm": 0.01160862110555172,
+      "learning_rate": 2.464856485569779e-05,
+      "loss": 0.0798,
+      "step": 9550
+    },
+    {
+      "epoch": 0.42644635460740216,
+      "grad_norm": 0.3391244411468506,
+      "learning_rate": 2.4580493185097692e-05,
+      "loss": 0.0106,
+      "step": 9575
+    },
+    {
+      "epoch": 0.4275597915646016,
+      "grad_norm": 0.02090360037982464,
+      "learning_rate": 2.4512365468731692e-05,
+      "loss": 0.0975,
+      "step": 9600
+    },
+    {
+      "epoch": 0.4286732285218011,
+      "grad_norm": 2.563877820968628,
+      "learning_rate": 2.444418254019343e-05,
+      "loss": 0.0701,
+      "step": 9625
+    },
+    {
+      "epoch": 0.4297866654790006,
+      "grad_norm": 0.10990472882986069,
+      "learning_rate": 2.437594523375213e-05,
+      "loss": 0.141,
+      "step": 9650
+    },
+    {
+      "epoch": 0.4309001024362001,
+      "grad_norm": 0.3325424790382385,
+      "learning_rate": 2.430765438434235e-05,
+      "loss": 0.0948,
+      "step": 9675
+    },
+    {
+      "epoch": 0.43201353939339954,
+      "grad_norm": 0.5756065249443054,
+      "learning_rate": 2.423931082755381e-05,
+      "loss": 0.1001,
+      "step": 9700
+    },
+    {
+      "epoch": 0.43312697635059905,
+      "grad_norm": 8.402749061584473,
+      "learning_rate": 2.417091539962112e-05,
+      "loss": 0.0925,
+      "step": 9725
+    },
+    {
+      "epoch": 0.4342404133077985,
+      "grad_norm": 21.99260711669922,
+      "learning_rate": 2.4102468937413593e-05,
+      "loss": 0.0568,
+      "step": 9750
+    },
+    {
+      "epoch": 0.435353850264998,
+      "grad_norm": 0.5746713876724243,
+      "learning_rate": 2.4033972278424965e-05,
+      "loss": 0.0348,
+      "step": 9775
+    },
+    {
+      "epoch": 0.43646728722219746,
+      "grad_norm": 0.02806955948472023,
+      "learning_rate": 2.3965426260763196e-05,
+      "loss": 0.0285,
+      "step": 9800
+    },
+    {
+      "epoch": 0.437580724179397,
+      "grad_norm": 0.06320275366306305,
+      "learning_rate": 2.3896831723140145e-05,
+      "loss": 0.0852,
+      "step": 9825
+    },
+    {
+      "epoch": 0.4386941611365964,
+      "grad_norm": 0.05184788629412651,
+      "learning_rate": 2.382818950486139e-05,
+      "loss": 0.1056,
+      "step": 9850
+    },
+    {
+      "epoch": 0.43980759809379594,
+      "grad_norm": 9.47110366821289,
+      "learning_rate": 2.375950044581589e-05,
+      "loss": 0.0809,
+      "step": 9875
+    },
+    {
+      "epoch": 0.4409210350509954,
+      "grad_norm": 0.024401208385825157,
+      "learning_rate": 2.369076538646575e-05,
+      "loss": 0.0441,
+      "step": 9900
+    },
+    {
+      "epoch": 0.4420344720081949,
+      "grad_norm": 0.12478948384523392,
+      "learning_rate": 2.3621985167835913e-05,
+      "loss": 0.0856,
+      "step": 9925
+    },
+    {
+      "epoch": 0.44314790896539435,
+      "grad_norm": 8.50674819946289,
+      "learning_rate": 2.3553160631503897e-05,
+      "loss": 0.1729,
+      "step": 9950
+    },
+    {
+      "epoch": 0.44426134592259386,
+      "grad_norm": 4.947354316711426,
+      "learning_rate": 2.3484292619589452e-05,
+      "loss": 0.0641,
+      "step": 9975
+    },
+    {
+      "epoch": 0.4453747828797933,
+      "grad_norm": 0.1289278119802475,
+      "learning_rate": 2.341538197474432e-05,
+      "loss": 0.0291,
+      "step": 10000
+    },
+    {
+      "epoch": 0.44648821983699283,
+      "grad_norm": 0.03130419924855232,
+      "learning_rate": 2.334642954014185e-05,
+      "loss": 0.0388,
+      "step": 10025
+    },
+    {
+      "epoch": 0.44760165679419234,
+      "grad_norm": 0.13553370535373688,
+      "learning_rate": 2.3277436159466756e-05,
+      "loss": 0.0721,
+      "step": 10050
+    },
+    {
+      "epoch": 0.4487150937513918,
+      "grad_norm": 0.05339996889233589,
+      "learning_rate": 2.3208402676904735e-05,
+      "loss": 0.0726,
+      "step": 10075
+    },
+    {
+      "epoch": 0.4498285307085913,
+      "grad_norm": 0.03551988676190376,
+      "learning_rate": 2.3139329937132167e-05,
+      "loss": 0.172,
+      "step": 10100
+    },
+    {
+      "epoch": 0.45094196766579075,
+      "grad_norm": 0.36548513174057007,
+      "learning_rate": 2.3070218785305774e-05,
+      "loss": 0.1802,
+      "step": 10125
+    },
+    {
+      "epoch": 0.45205540462299026,
+      "grad_norm": 0.018251005560159683,
+      "learning_rate": 2.3001070067052277e-05,
+      "loss": 0.0896,
+      "step": 10150
+    },
+    {
+      "epoch": 0.4531688415801897,
+      "grad_norm": 0.2378775030374527,
+      "learning_rate": 2.293188462845806e-05,
+      "loss": 0.1419,
+      "step": 10175
+    },
+    {
+      "epoch": 0.45428227853738923,
+      "grad_norm": 0.4987199008464813,
+      "learning_rate": 2.2862663316058783e-05,
+      "loss": 0.1001,
+      "step": 10200
+    },
+    {
+      "epoch": 0.4553957154945887,
+      "grad_norm": 0.28313758969306946,
+      "learning_rate": 2.279340697682908e-05,
+      "loss": 0.0393,
+      "step": 10225
+    },
+    {
+      "epoch": 0.4565091524517882,
+      "grad_norm": 0.08381898701190948,
+      "learning_rate": 2.272411645817214e-05,
+      "loss": 0.0942,
+      "step": 10250
+    },
+    {
+      "epoch": 0.45762258940898765,
+      "grad_norm": 0.036367420107126236,
+      "learning_rate": 2.265479260790938e-05,
+      "loss": 0.012,
+      "step": 10275
+    },
+    {
+      "epoch": 0.45873602636618716,
+      "grad_norm": 0.2814151644706726,
+      "learning_rate": 2.2585436274270038e-05,
+      "loss": 0.1508,
+      "step": 10300
+    },
+    {
+      "epoch": 0.4598494633233866,
+      "grad_norm": 0.3467974364757538,
+      "learning_rate": 2.251604830588083e-05,
+      "loss": 0.0935,
+      "step": 10325
+    },
+    {
+      "epoch": 0.4609629002805861,
+      "grad_norm": 0.05292457342147827,
+      "learning_rate": 2.2446629551755523e-05,
+      "loss": 0.1146,
+      "step": 10350
+    },
+    {
+      "epoch": 0.4620763372377856,
+      "grad_norm": 0.5696282386779785,
+      "learning_rate": 2.2377180861284593e-05,
+      "loss": 0.0728,
+      "step": 10375
+    },
+    {
+      "epoch": 0.4631897741949851,
+      "grad_norm": 0.13021673262119293,
+      "learning_rate": 2.2307703084224794e-05,
+      "loss": 0.0684,
+      "step": 10400
+    },
+    {
+      "epoch": 0.46430321115218454,
+      "grad_norm": 0.04874199256300926,
+      "learning_rate": 2.2238197070688794e-05,
+      "loss": 0.1286,
+      "step": 10425
+    },
+    {
+      "epoch": 0.46541664810938405,
+      "grad_norm": 0.1427091658115387,
+      "learning_rate": 2.2168663671134727e-05,
+      "loss": 0.0964,
+      "step": 10450
+    },
+    {
+      "epoch": 0.46653008506658356,
+      "grad_norm": 0.5847116708755493,
+      "learning_rate": 2.2099103736355846e-05,
+      "loss": 0.0923,
+      "step": 10475
+    },
+    {
+      "epoch": 0.467643522023783,
+      "grad_norm": 13.224454879760742,
+      "learning_rate": 2.2029518117470062e-05,
+      "loss": 0.1031,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4687569589809825,
+      "grad_norm": 0.14730241894721985,
+      "learning_rate": 2.195990766590956e-05,
+      "loss": 0.0915,
+      "step": 10525
+    },
+    {
+      "epoch": 0.469870395938182,
+      "grad_norm": 0.5047757029533386,
+      "learning_rate": 2.1890273233410366e-05,
+      "loss": 0.1317,
+      "step": 10550
+    },
+    {
+      "epoch": 0.4709838328953815,
+      "grad_norm": 21.881359100341797,
+      "learning_rate": 2.1820615672001934e-05,
+      "loss": 0.0431,
+      "step": 10575
+    },
+    {
+      "epoch": 0.47209726985258094,
+      "grad_norm": 0.19091269373893738,
+      "learning_rate": 2.175093583399672e-05,
+      "loss": 0.0497,
+      "step": 10600
+    },
+    {
+      "epoch": 0.47321070680978045,
+      "grad_norm": 0.06911573559045792,
+      "learning_rate": 2.1681234571979734e-05,
+      "loss": 0.0683,
+      "step": 10625
+    },
+    {
+      "epoch": 0.4743241437669799,
+      "grad_norm": 0.12921109795570374,
+      "learning_rate": 2.1611512738798157e-05,
+      "loss": 0.1125,
+      "step": 10650
+    },
+    {
+      "epoch": 0.4754375807241794,
+      "grad_norm": 1.3468788862228394,
+      "learning_rate": 2.154177118755084e-05,
+      "loss": 0.0542,
+      "step": 10675
+    },
+    {
+      "epoch": 0.47655101768137886,
+      "grad_norm": 14.072271347045898,
+      "learning_rate": 2.1472010771577922e-05,
+      "loss": 0.1039,
+      "step": 10700
+    },
+    {
+      "epoch": 0.4776644546385784,
+      "grad_norm": 0.018694249913096428,
+      "learning_rate": 2.1402232344450357e-05,
+      "loss": 0.078,
+      "step": 10725
+    },
+    {
+      "epoch": 0.4787778915957778,
+      "grad_norm": 0.015503577888011932,
+      "learning_rate": 2.133243675995948e-05,
+      "loss": 0.0599,
+      "step": 10750
+    },
+    {
+      "epoch": 0.47989132855297734,
+      "grad_norm": 0.015802688896656036,
+      "learning_rate": 2.1262624872106557e-05,
+      "loss": 0.1028,
+      "step": 10775
+    },
+    {
+      "epoch": 0.4810047655101768,
+      "grad_norm": 77.8501968383789,
+      "learning_rate": 2.1192797535092346e-05,
+      "loss": 0.102,
+      "step": 10800
+    },
+    {
+      "epoch": 0.4821182024673763,
+      "grad_norm": 0.7204959392547607,
+      "learning_rate": 2.112295560330662e-05,
+      "loss": 0.0704,
+      "step": 10825
+    },
+    {
+      "epoch": 0.48323163942457575,
+      "grad_norm": 21.50542640686035,
+      "learning_rate": 2.1053099931317754e-05,
+      "loss": 0.0671,
+      "step": 10850
+    },
+    {
+      "epoch": 0.48434507638177526,
+      "grad_norm": 10.961987495422363,
+      "learning_rate": 2.0983231373862224e-05,
+      "loss": 0.0689,
+      "step": 10875
+    },
+    {
+      "epoch": 0.4854585133389748,
+      "grad_norm": 0.4164874255657196,
+      "learning_rate": 2.0913350785834175e-05,
+      "loss": 0.0909,
+      "step": 10900
+    },
+    {
+      "epoch": 0.4865719502961742,
+      "grad_norm": 0.20815160870552063,
+      "learning_rate": 2.0843459022274973e-05,
+      "loss": 0.162,
+      "step": 10925
+    },
+    {
+      "epoch": 0.48768538725337374,
+      "grad_norm": 0.03996884077787399,
+      "learning_rate": 2.077355693836269e-05,
+      "loss": 0.0432,
+      "step": 10950
+    },
+    {
+      "epoch": 0.4887988242105732,
+      "grad_norm": 0.06414720416069031,
+      "learning_rate": 2.070364538940171e-05,
+      "loss": 0.1202,
+      "step": 10975
+    },
+    {
+      "epoch": 0.4899122611677727,
+      "grad_norm": 0.052727632224559784,
+      "learning_rate": 2.06337252308122e-05,
+      "loss": 0.0961,
+      "step": 11000
+    },
+    {
+      "epoch": 0.49102569812497215,
+      "grad_norm": 0.1574171483516693,
+      "learning_rate": 2.0563797318119696e-05,
+      "loss": 0.0644,
+      "step": 11025
+    },
+    {
+      "epoch": 0.49213913508217166,
+      "grad_norm": 0.6176048517227173,
+      "learning_rate": 2.0493862506944597e-05,
+      "loss": 0.0232,
+      "step": 11050
+    },
+    {
+      "epoch": 0.4932525720393711,
+      "grad_norm": 0.07887354493141174,
+      "learning_rate": 2.0423921652991713e-05,
+      "loss": 0.0623,
+      "step": 11075
+    },
+    {
+      "epoch": 0.4943660089965706,
+      "grad_norm": 0.379489541053772,
+      "learning_rate": 2.0353975612039797e-05,
+      "loss": 0.075,
+      "step": 11100
+    },
+    {
+      "epoch": 0.4954794459537701,
+      "grad_norm": 0.22683410346508026,
+      "learning_rate": 2.0284025239931064e-05,
+      "loss": 0.068,
+      "step": 11125
+    },
+    {
+      "epoch": 0.4965928829109696,
+      "grad_norm": 0.014947167597711086,
+      "learning_rate": 2.021407139256072e-05,
+      "loss": 0.0706,
+      "step": 11150
+    },
+    {
+      "epoch": 0.49770631986816904,
+      "grad_norm": 0.08996062725782394,
+      "learning_rate": 2.01441149258665e-05,
+      "loss": 0.1368,
+      "step": 11175
+    },
+    {
+      "epoch": 0.49881975682536855,
+      "grad_norm": 1.830735206604004,
+      "learning_rate": 2.0074156695818194e-05,
+      "loss": 0.135,
+      "step": 11200
+    },
+    {
+      "epoch": 0.499933193782568,
+      "grad_norm": 165.89505004882812,
+      "learning_rate": 2.0004197558407154e-05,
+      "loss": 0.1127,
+      "step": 11225
+    },
+    {
+      "epoch": 0.5010466307397675,
+      "grad_norm": 198.1811065673828,
+      "learning_rate": 1.9934238369635837e-05,
+      "loss": 0.0691,
+      "step": 11250
+    },
+    {
+      "epoch": 0.502160067696967,
+      "grad_norm": 17.284914016723633,
+      "learning_rate": 1.9864279985507345e-05,
+      "loss": 0.0968,
+      "step": 11275
+    },
+    {
+      "epoch": 0.5032735046541664,
+      "grad_norm": 3.0509047508239746,
+      "learning_rate": 1.9794323262014915e-05,
+      "loss": 0.0535,
+      "step": 11300
+    },
+    {
+      "epoch": 0.504386941611366,
+      "grad_norm": 0.1367504894733429,
+      "learning_rate": 1.972436905513146e-05,
+      "loss": 0.0834,
+      "step": 11325
+    },
+    {
+      "epoch": 0.5055003785685654,
+      "grad_norm": 0.118153877556324,
+      "learning_rate": 1.965441822079913e-05,
+      "loss": 0.0818,
+      "step": 11350
+    },
+    {
+      "epoch": 0.5066138155257649,
+      "grad_norm": 0.06398922950029373,
+      "learning_rate": 1.9584471614918787e-05,
+      "loss": 0.0542,
+      "step": 11375
+    },
+    {
+      "epoch": 0.5077272524829645,
+      "grad_norm": 0.03594028949737549,
+      "learning_rate": 1.951453009333955e-05,
+      "loss": 0.0065,
+      "step": 11400
+    },
+    {
+      "epoch": 0.5088406894401639,
+      "grad_norm": 0.29045698046684265,
+      "learning_rate": 1.9444594511848346e-05,
+      "loss": 0.0689,
+      "step": 11425
+    },
+    {
+      "epoch": 0.5099541263973634,
+      "grad_norm": 13.441309928894043,
+      "learning_rate": 1.9374665726159415e-05,
+      "loss": 0.0972,
+      "step": 11450
+    },
+    {
+      "epoch": 0.5110675633545628,
+      "grad_norm": 0.055517617613077164,
+      "learning_rate": 1.9304744591903837e-05,
+      "loss": 0.064,
+      "step": 11475
+    },
+    {
+      "epoch": 0.5121810003117624,
+      "grad_norm": 0.018037894740700722,
+      "learning_rate": 1.923483196461907e-05,
+      "loss": 0.1372,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5132944372689618,
+      "grad_norm": 24.66988754272461,
+      "learning_rate": 1.9164928699738492e-05,
+      "loss": 0.0987,
+      "step": 11525
+    },
+    {
+      "epoch": 0.5144078742261613,
+      "grad_norm": 0.01959937810897827,
+      "learning_rate": 1.909503565258093e-05,
+      "loss": 0.1023,
+      "step": 11550
+    },
+    {
+      "epoch": 0.5155213111833608,
+      "grad_norm": 0.07445913553237915,
+      "learning_rate": 1.9025153678340173e-05,
+      "loss": 0.0704,
+      "step": 11575
+    },
+    {
+      "epoch": 0.5166347481405603,
+      "grad_norm": 0.6288459300994873,
+      "learning_rate": 1.895528363207453e-05,
+      "loss": 0.1181,
+      "step": 11600
+    },
+    {
+      "epoch": 0.5177481850977598,
+      "grad_norm": 0.2090267390012741,
+      "learning_rate": 1.8885426368696374e-05,
+      "loss": 0.121,
+      "step": 11625
+    },
+    {
+      "epoch": 0.5188616220549592,
+      "grad_norm": 0.4164348244667053,
+      "learning_rate": 1.8815582742961652e-05,
+      "loss": 0.0685,
+      "step": 11650
+    },
+    {
+      "epoch": 0.5199750590121587,
+      "grad_norm": 0.7623189091682434,
+      "learning_rate": 1.874575360945946e-05,
+      "loss": 0.0942,
+      "step": 11675
+    },
+    {
+      "epoch": 0.5210884959693582,
+      "grad_norm": 0.10410900413990021,
+      "learning_rate": 1.8675939822601558e-05,
+      "loss": 0.1427,
+      "step": 11700
+    },
+    {
+      "epoch": 0.5222019329265577,
+      "grad_norm": 9.168070793151855,
+      "learning_rate": 1.860614223661194e-05,
+      "loss": 0.1384,
+      "step": 11725
+    },
+    {
+      "epoch": 0.5233153698837572,
+      "grad_norm": 0.12809838354587555,
+      "learning_rate": 1.853636170551636e-05,
+      "loss": 0.0664,
+      "step": 11750
+    },
+    {
+      "epoch": 0.5244288068409567,
+      "grad_norm": 2.214749336242676,
+      "learning_rate": 1.8466599083131894e-05,
+      "loss": 0.0525,
+      "step": 11775
+    },
+    {
+      "epoch": 0.5255422437981562,
+      "grad_norm": 0.1398187279701233,
+      "learning_rate": 1.8396855223056505e-05,
+      "loss": 0.1128,
+      "step": 11800
+    },
+    {
+      "epoch": 0.5266556807553556,
+      "grad_norm": 0.05570574477314949,
+      "learning_rate": 1.8327130978658565e-05,
+      "loss": 0.0942,
+      "step": 11825
+    },
+    {
+      "epoch": 0.5277691177125551,
+      "grad_norm": 0.015549311414361,
+      "learning_rate": 1.825742720306644e-05,
+      "loss": 0.0127,
+      "step": 11850
+    },
+    {
+      "epoch": 0.5288825546697546,
+      "grad_norm": 0.048954132944345474,
+      "learning_rate": 1.8187744749158055e-05,
+      "loss": 0.0417,
+      "step": 11875
+    },
+    {
+      "epoch": 0.5299959916269541,
+      "grad_norm": 2.6331558227539062,
+      "learning_rate": 1.8118084469550446e-05,
+      "loss": 0.0608,
+      "step": 11900
+    },
+    {
+      "epoch": 0.5311094285841536,
+      "grad_norm": 0.16815824806690216,
+      "learning_rate": 1.8048447216589316e-05,
+      "loss": 0.0504,
+      "step": 11925
+    },
+    {
+      "epoch": 0.532222865541353,
+      "grad_norm": 31.2311954498291,
+      "learning_rate": 1.7978833842338625e-05,
+      "loss": 0.1608,
+      "step": 11950
+    },
+    {
+      "epoch": 0.5333363024985526,
+      "grad_norm": 0.21393436193466187,
+      "learning_rate": 1.7909245198570165e-05,
+      "loss": 0.087,
+      "step": 11975
+    },
+    {
+      "epoch": 0.534449739455752,
+      "grad_norm": 0.7251951694488525,
+      "learning_rate": 1.7839682136753143e-05,
+      "loss": 0.0518,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5355631764129515,
+      "grad_norm": 13.856010437011719,
+      "learning_rate": 1.7770145508043722e-05,
+      "loss": 0.07,
+      "step": 12025
+    },
+    {
+      "epoch": 0.5366766133701509,
+      "grad_norm": 0.03707493096590042,
+      "learning_rate": 1.7700636163274657e-05,
+      "loss": 0.1027,
+      "step": 12050
+    },
+    {
+      "epoch": 0.5377900503273505,
+      "grad_norm": 0.33082666993141174,
+      "learning_rate": 1.763115495294486e-05,
+      "loss": 0.0741,
+      "step": 12075
+    },
+    {
+      "epoch": 0.53890348728455,
+      "grad_norm": 1.9731577634811401,
+      "learning_rate": 1.7561702727209e-05,
+      "loss": 0.0381,
+      "step": 12100
+    },
+    {
+      "epoch": 0.5400169242417494,
+      "grad_norm": 0.016690315678715706,
+      "learning_rate": 1.749228033586709e-05,
+      "loss": 0.0613,
+      "step": 12125
+    },
+    {
+      "epoch": 0.5411303611989489,
+      "grad_norm": 11.721412658691406,
+      "learning_rate": 1.7422888628354095e-05,
+      "loss": 0.0957,
+      "step": 12150
+    },
+    {
+      "epoch": 0.5422437981561484,
+      "grad_norm": 0.08547116816043854,
+      "learning_rate": 1.7353528453729552e-05,
+      "loss": 0.0852,
+      "step": 12175
+    },
+    {
+      "epoch": 0.5433572351133479,
+      "grad_norm": 0.03922642022371292,
+      "learning_rate": 1.728420066066717e-05,
+      "loss": 0.0678,
+      "step": 12200
+    },
+    {
+      "epoch": 0.5444706720705473,
+      "grad_norm": 20.77050018310547,
+      "learning_rate": 1.7214906097444425e-05,
+      "loss": 0.0576,
+      "step": 12225
+    },
+    {
+      "epoch": 0.5455841090277469,
+      "grad_norm": 0.051048461347818375,
+      "learning_rate": 1.7145645611932218e-05,
+      "loss": 0.0796,
+      "step": 12250
+    },
+    {
+      "epoch": 0.5466975459849464,
+      "grad_norm": 0.5286572575569153,
+      "learning_rate": 1.707642005158449e-05,
+      "loss": 0.0701,
+      "step": 12275
+    },
+    {
+      "epoch": 0.5478109829421458,
+      "grad_norm": 0.10311714559793472,
+      "learning_rate": 1.7007230263427834e-05,
+      "loss": 0.0738,
+      "step": 12300
+    },
+    {
+      "epoch": 0.5489244198993453,
+      "grad_norm": 98.99889373779297,
+      "learning_rate": 1.693807709405114e-05,
+      "loss": 0.067,
+      "step": 12325
+    },
+    {
+      "epoch": 0.5500378568565448,
+      "grad_norm": 10.721565246582031,
+      "learning_rate": 1.686896138959524e-05,
+      "loss": 0.0247,
+      "step": 12350
+    },
+    {
+      "epoch": 0.5511512938137443,
+      "grad_norm": 0.025939449667930603,
+      "learning_rate": 1.679988399574258e-05,
+      "loss": 0.043,
+      "step": 12375
+    },
+    {
+      "epoch": 0.5522647307709437,
+      "grad_norm": 0.05403141677379608,
+      "learning_rate": 1.673084575770682e-05,
+      "loss": 0.0906,
+      "step": 12400
+    },
+    {
+      "epoch": 0.5533781677281432,
+      "grad_norm": 0.1890357881784439,
+      "learning_rate": 1.666184752022252e-05,
+      "loss": 0.1179,
+      "step": 12425
+    },
+    {
+      "epoch": 0.5544916046853428,
+      "grad_norm": 0.15934480726718903,
+      "learning_rate": 1.659289012753481e-05,
+      "loss": 0.0444,
+      "step": 12450
+    },
+    {
+      "epoch": 0.5556050416425422,
+      "grad_norm": 0.10561876744031906,
+      "learning_rate": 1.6523974423389068e-05,
+      "loss": 0.014,
+      "step": 12475
+    },
+    {
+      "epoch": 0.5567184785997417,
+      "grad_norm": 0.5226536989212036,
+      "learning_rate": 1.645510125102056e-05,
+      "loss": 0.0433,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5578319155569411,
+      "grad_norm": 0.10878469049930573,
+      "learning_rate": 1.6386271453144144e-05,
+      "loss": 0.1258,
+      "step": 12525
+    },
+    {
+      "epoch": 0.5589453525141407,
+      "grad_norm": 0.44833049178123474,
+      "learning_rate": 1.6317485871943966e-05,
+      "loss": 0.1303,
+      "step": 12550
+    },
+    {
+      "epoch": 0.5600587894713401,
+      "grad_norm": 0.02392134629189968,
+      "learning_rate": 1.624874534906316e-05,
+      "loss": 0.0989,
+      "step": 12575
+    },
+    {
+      "epoch": 0.5611722264285396,
+      "grad_norm": 0.05233491584658623,
+      "learning_rate": 1.6180050725593516e-05,
+      "loss": 0.0654,
+      "step": 12600
+    },
+    {
+      "epoch": 0.5622856633857392,
+      "grad_norm": 1.7143317461013794,
+      "learning_rate": 1.6111402842065213e-05,
+      "loss": 0.1005,
+      "step": 12625
+    },
+    {
+      "epoch": 0.5633991003429386,
+      "grad_norm": 0.02804090641438961,
+      "learning_rate": 1.604280253843655e-05,
+      "loss": 0.0328,
+      "step": 12650
+    },
+    {
+      "epoch": 0.5645125373001381,
+      "grad_norm": 0.04471449553966522,
+      "learning_rate": 1.5974250654083643e-05,
+      "loss": 0.125,
+      "step": 12675
+    },
+    {
+      "epoch": 0.5656259742573375,
+      "grad_norm": 0.04020120948553085,
+      "learning_rate": 1.5905748027790145e-05,
+      "loss": 0.0254,
+      "step": 12700
+    },
+    {
+      "epoch": 0.5667394112145371,
+      "grad_norm": 9.090320587158203,
+      "learning_rate": 1.5837295497737005e-05,
+      "loss": 0.1001,
+      "step": 12725
+    },
+    {
+      "epoch": 0.5678528481717365,
+      "grad_norm": 0.017459219321608543,
+      "learning_rate": 1.5768893901492238e-05,
+      "loss": 0.0815,
+      "step": 12750
+    },
+    {
+      "epoch": 0.568966285128936,
+      "grad_norm": 0.14465971291065216,
+      "learning_rate": 1.5700544076000604e-05,
+      "loss": 0.0787,
+      "step": 12775
+    },
+    {
+      "epoch": 0.5700797220861354,
+      "grad_norm": 0.27508795261383057,
+      "learning_rate": 1.5632246857573434e-05,
+      "loss": 0.0788,
+      "step": 12800
+    },
+    {
+      "epoch": 0.571193159043335,
+      "grad_norm": 0.668519914150238,
+      "learning_rate": 1.5564003081878355e-05,
+      "loss": 0.0925,
+      "step": 12825
+    },
+    {
+      "epoch": 0.5723065960005345,
+      "grad_norm": 0.02386564202606678,
+      "learning_rate": 1.5495813583929105e-05,
+      "loss": 0.0432,
+      "step": 12850
+    },
+    {
+      "epoch": 0.5734200329577339,
+      "grad_norm": 0.31950095295906067,
+      "learning_rate": 1.5427679198075277e-05,
+      "loss": 0.0472,
+      "step": 12875
+    },
+    {
+      "epoch": 0.5745334699149334,
+      "grad_norm": 0.013991210609674454,
+      "learning_rate": 1.5359600757992127e-05,
+      "loss": 0.1124,
+      "step": 12900
+    },
+    {
+      "epoch": 0.5756469068721329,
+      "grad_norm": 0.05440308153629303,
+      "learning_rate": 1.529157909667037e-05,
+      "loss": 0.0085,
+      "step": 12925
+    },
+    {
+      "epoch": 0.5767603438293324,
+      "grad_norm": 132.2735137939453,
+      "learning_rate": 1.5223615046406007e-05,
+      "loss": 0.0712,
+      "step": 12950
+    },
+    {
+      "epoch": 0.5778737807865318,
+      "grad_norm": 0.03499956429004669,
+      "learning_rate": 1.5155709438790105e-05,
+      "loss": 0.0275,
+      "step": 12975
+    },
+    {
+      "epoch": 0.5789872177437313,
+      "grad_norm": 13.37109088897705,
+      "learning_rate": 1.5087863104698653e-05,
+      "loss": 0.0884,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5801006547009309,
+      "grad_norm": 0.1880478411912918,
+      "learning_rate": 1.502007687428237e-05,
+      "loss": 0.046,
+      "step": 13025
+    },
+    {
+      "epoch": 0.5812140916581303,
+      "grad_norm": 0.01652064546942711,
+      "learning_rate": 1.4952351576956587e-05,
+      "loss": 0.0456,
+      "step": 13050
+    },
+    {
+      "epoch": 0.5823275286153298,
+      "grad_norm": 0.05549848824739456,
+      "learning_rate": 1.4884688041391053e-05,
+      "loss": 0.0455,
+      "step": 13075
+    },
+    {
+      "epoch": 0.5834409655725293,
+      "grad_norm": 0.1668526977300644,
+      "learning_rate": 1.4817087095499809e-05,
+      "loss": 0.0865,
+      "step": 13100
+    },
+    {
+      "epoch": 0.5845544025297288,
+      "grad_norm": 0.06226930022239685,
+      "learning_rate": 1.4749549566431092e-05,
+      "loss": 0.0033,
+      "step": 13125
+    },
+    {
+      "epoch": 0.5856678394869282,
+      "grad_norm": 0.2953890860080719,
+      "learning_rate": 1.4682076280557167e-05,
+      "loss": 0.063,
+      "step": 13150
+    },
+    {
+      "epoch": 0.5867812764441277,
+      "grad_norm": 0.024164235219359398,
+      "learning_rate": 1.4614668063464235e-05,
+      "loss": 0.0394,
+      "step": 13175
+    },
+    {
+      "epoch": 0.5878947134013273,
+      "grad_norm": 0.04861380532383919,
+      "learning_rate": 1.4547325739942327e-05,
+      "loss": 0.1269,
+      "step": 13200
+    },
+    {
+      "epoch": 0.5890081503585267,
+      "grad_norm": 0.2787506878376007,
+      "learning_rate": 1.4480050133975242e-05,
+      "loss": 0.0625,
+      "step": 13225
+    },
+    {
+      "epoch": 0.5901215873157262,
+      "grad_norm": 0.12541650235652924,
+      "learning_rate": 1.4412842068730419e-05,
+      "loss": 0.0756,
+      "step": 13250
+    },
+    {
+      "epoch": 0.5912350242729256,
+      "grad_norm": 0.06487028300762177,
+      "learning_rate": 1.4345702366548884e-05,
+      "loss": 0.0956,
+      "step": 13275
+    },
+    {
+      "epoch": 0.5923484612301252,
+      "grad_norm": 0.023417694494128227,
+      "learning_rate": 1.4278631848935192e-05,
+      "loss": 0.0588,
+      "step": 13300
+    },
+    {
+      "epoch": 0.5934618981873246,
+      "grad_norm": 0.04044019430875778,
+      "learning_rate": 1.4211631336547389e-05,
+      "loss": 0.0581,
+      "step": 13325
+    },
+    {
+      "epoch": 0.5945753351445241,
+      "grad_norm": 0.13640573620796204,
+      "learning_rate": 1.4144701649186937e-05,
+      "loss": 0.0508,
+      "step": 13350
+    },
+    {
+      "epoch": 0.5956887721017236,
+      "grad_norm": 0.17478786408901215,
+      "learning_rate": 1.4077843605788701e-05,
+      "loss": 0.0371,
+      "step": 13375
+    },
+    {
+      "epoch": 0.5968022090589231,
+      "grad_norm": 0.09428933262825012,
+      "learning_rate": 1.4011058024410931e-05,
+      "loss": 0.0125,
+      "step": 13400
+    },
+    {
+      "epoch": 0.5979156460161226,
+      "grad_norm": 0.01731252484023571,
+      "learning_rate": 1.3944345722225268e-05,
+      "loss": 0.0668,
+      "step": 13425
+    },
+    {
+      "epoch": 0.599029082973322,
+      "grad_norm": 0.017465805634856224,
+      "learning_rate": 1.38777075155067e-05,
+      "loss": 0.1256,
+      "step": 13450
+    },
+    {
+      "epoch": 0.6001425199305215,
+      "grad_norm": 0.028824403882026672,
+      "learning_rate": 1.3811144219623613e-05,
+      "loss": 0.1192,
+      "step": 13475
+    },
+    {
+      "epoch": 0.601255956887721,
+      "grad_norm": 0.2588452696800232,
+      "learning_rate": 1.3744656649027795e-05,
+      "loss": 0.0677,
+      "step": 13500
+    },
+    {
+      "epoch": 0.6023693938449205,
+      "grad_norm": 0.2693833112716675,
+      "learning_rate": 1.3678245617244503e-05,
+      "loss": 0.0943,
+      "step": 13525
+    },
+    {
+      "epoch": 0.60348283080212,
+      "grad_norm": 0.055959224700927734,
+      "learning_rate": 1.361191193686246e-05,
+      "loss": 0.0645,
+      "step": 13550
+    },
+    {
+      "epoch": 0.6045962677593195,
+      "grad_norm": 12.030915260314941,
+      "learning_rate": 1.354565641952393e-05,
+      "loss": 0.0828,
+      "step": 13575
+    },
+    {
+      "epoch": 0.605709704716519,
+      "grad_norm": 153.66726684570312,
+      "learning_rate": 1.3479479875914834e-05,
+      "loss": 0.0849,
+      "step": 13600
+    },
+    {
+      "epoch": 0.6068231416737184,
+      "grad_norm": 0.015734700486063957,
+      "learning_rate": 1.3413383115754748e-05,
+      "loss": 0.0872,
+      "step": 13625
+    },
+    {
+      "epoch": 0.6079365786309179,
+      "grad_norm": 0.16204610466957092,
+      "learning_rate": 1.3347366947787054e-05,
+      "loss": 0.1388,
+      "step": 13650
+    },
+    {
+      "epoch": 0.6090500155881174,
+      "grad_norm": 0.10918817669153214,
+      "learning_rate": 1.3281432179769023e-05,
+      "loss": 0.1072,
+      "step": 13675
+    },
+    {
+      "epoch": 0.6101634525453169,
+      "grad_norm": 0.25966718792915344,
+      "learning_rate": 1.3215579618461955e-05,
+      "loss": 0.1107,
+      "step": 13700
+    },
+    {
+      "epoch": 0.6112768895025164,
+      "grad_norm": 0.11825607717037201,
+      "learning_rate": 1.3149810069621265e-05,
+      "loss": 0.0657,
+      "step": 13725
+    },
+    {
+      "epoch": 0.6123903264597158,
+      "grad_norm": 0.08822792768478394,
+      "learning_rate": 1.3084124337986666e-05,
+      "loss": 0.0468,
+      "step": 13750
+    },
+    {
+      "epoch": 0.6135037634169154,
+      "grad_norm": 0.02553175576031208,
+      "learning_rate": 1.301852322727228e-05,
+      "loss": 0.0874,
+      "step": 13775
+    },
+    {
+      "epoch": 0.6146172003741148,
+      "grad_norm": 0.3989555537700653,
+      "learning_rate": 1.295300754015687e-05,
+      "loss": 0.0379,
+      "step": 13800
+    },
+    {
+      "epoch": 0.6157306373313143,
+      "grad_norm": 0.04876472055912018,
+      "learning_rate": 1.2887578078273944e-05,
+      "loss": 0.0972,
+      "step": 13825
+    },
+    {
+      "epoch": 0.6168440742885137,
+      "grad_norm": 0.776457667350769,
+      "learning_rate": 1.2822235642201983e-05,
+      "loss": 0.0757,
+      "step": 13850
+    },
+    {
+      "epoch": 0.6179575112457133,
+      "grad_norm": 1.26872718334198,
+      "learning_rate": 1.275698103145465e-05,
+      "loss": 0.0438,
+      "step": 13875
+    },
+    {
+      "epoch": 0.6190709482029128,
+      "grad_norm": 7.4887309074401855,
+      "learning_rate": 1.2691815044471005e-05,
+      "loss": 0.0944,
+      "step": 13900
+    },
+    {
+      "epoch": 0.6201843851601122,
+      "grad_norm": 0.061836518347263336,
+      "learning_rate": 1.262673847860572e-05,
+      "loss": 0.0507,
+      "step": 13925
+    },
+    {
+      "epoch": 0.6212978221173118,
+      "grad_norm": 31.68821144104004,
+      "learning_rate": 1.2561752130119325e-05,
+      "loss": 0.1315,
+      "step": 13950
+    },
+    {
+      "epoch": 0.6224112590745112,
+      "grad_norm": 1.8287922143936157,
+      "learning_rate": 1.2496856794168493e-05,
+      "loss": 0.0173,
+      "step": 13975
+    },
+    {
+      "epoch": 0.6235246960317107,
+      "grad_norm": 0.07127226889133453,
+      "learning_rate": 1.243205326479628e-05,
+      "loss": 0.1028,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6246381329889101,
+      "grad_norm": 0.01413736306130886,
+      "learning_rate": 1.2367342334922419e-05,
+      "loss": 0.0194,
+      "step": 14025
+    },
+    {
+      "epoch": 0.6257515699461097,
+      "grad_norm": 0.01650845631957054,
+      "learning_rate": 1.2302724796333611e-05,
+      "loss": 0.0083,
+      "step": 14050
+    },
+    {
+      "epoch": 0.6268650069033092,
+      "grad_norm": 0.01144505012780428,
+      "learning_rate": 1.223820143967386e-05,
+      "loss": 0.0715,
+      "step": 14075
+    },
+    {
+      "epoch": 0.6279784438605086,
+      "grad_norm": 43.129913330078125,
+      "learning_rate": 1.2173773054434786e-05,
+      "loss": 0.1128,
+      "step": 14100
+    },
+    {
+      "epoch": 0.6290918808177081,
+      "grad_norm": 0.40030625462532043,
+      "learning_rate": 1.2109440428945937e-05,
+      "loss": 0.0995,
+      "step": 14125
+    },
+    {
+      "epoch": 0.6302053177749076,
+      "grad_norm": 0.14804595708847046,
+      "learning_rate": 1.2045204350365194e-05,
+      "loss": 0.077,
+      "step": 14150
+    },
+    {
+      "epoch": 0.6313187547321071,
+      "grad_norm": 8.412322044372559,
+      "learning_rate": 1.1981065604669102e-05,
+      "loss": 0.0816,
+      "step": 14175
+    },
+    {
+      "epoch": 0.6324321916893065,
+      "grad_norm": 0.00858822651207447,
+      "learning_rate": 1.1917024976643274e-05,
+      "loss": 0.0158,
+      "step": 14200
+    },
+    {
+      "epoch": 0.633545628646506,
+      "grad_norm": 30.407943725585938,
+      "learning_rate": 1.1853083249872768e-05,
+      "loss": 0.033,
+      "step": 14225
+    },
+    {
+      "epoch": 0.6346590656037056,
+      "grad_norm": 0.0964275524020195,
+      "learning_rate": 1.1789241206732519e-05,
+      "loss": 0.0953,
+      "step": 14250
+    },
+    {
+      "epoch": 0.635772502560905,
+      "grad_norm": 0.0621681734919548,
+      "learning_rate": 1.1725499628377759e-05,
+      "loss": 0.0566,
+      "step": 14275
+    },
+    {
+      "epoch": 0.6368859395181045,
+      "grad_norm": 91.5651626586914,
+      "learning_rate": 1.1661859294734454e-05,
+      "loss": 0.1282,
+      "step": 14300
+    },
+    {
+      "epoch": 0.6379993764753039,
+      "grad_norm": 20.94974708557129,
+      "learning_rate": 1.1598320984489767e-05,
+      "loss": 0.079,
+      "step": 14325
+    },
+    {
+      "epoch": 0.6391128134325035,
+      "grad_norm": 0.06739246100187302,
+      "learning_rate": 1.153488547508252e-05,
+      "loss": 0.0762,
+      "step": 14350
+    },
+    {
+      "epoch": 0.6402262503897029,
+      "grad_norm": 0.02148587629199028,
+      "learning_rate": 1.1471553542693715e-05,
+      "loss": 0.0377,
+      "step": 14375
+    },
+    {
+      "epoch": 0.6413396873469024,
+      "grad_norm": 37.625404357910156,
+      "learning_rate": 1.140832596223698e-05,
+      "loss": 0.0565,
+      "step": 14400
+    },
+    {
+      "epoch": 0.642453124304102,
+      "grad_norm": 0.10515021532773972,
+      "learning_rate": 1.134520350734913e-05,
+      "loss": 0.0788,
+      "step": 14425
+    },
+    {
+      "epoch": 0.6435665612613014,
+      "grad_norm": 0.11388204246759415,
+      "learning_rate": 1.1282186950380711e-05,
+      "loss": 0.0452,
+      "step": 14450
+    },
+    {
+      "epoch": 0.6446799982185009,
+      "grad_norm": 0.4367184340953827,
+      "learning_rate": 1.1219277062386497e-05,
+      "loss": 0.0818,
+      "step": 14475
+    },
+    {
+      "epoch": 0.6457934351757003,
+      "grad_norm": 33.38430404663086,
+      "learning_rate": 1.1156474613116106e-05,
+      "loss": 0.0526,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6469068721328999,
+      "grad_norm": 7.543337345123291,
+      "learning_rate": 1.1093780371004548e-05,
+      "loss": 0.1187,
+      "step": 14525
+    },
+    {
+      "epoch": 0.6480203090900993,
+      "grad_norm": 0.032384540885686874,
+      "learning_rate": 1.103119510316286e-05,
+      "loss": 0.0276,
+      "step": 14550
+    },
+    {
+      "epoch": 0.6491337460472988,
+      "grad_norm": 11.106677055358887,
+      "learning_rate": 1.0968719575368679e-05,
+      "loss": 0.0858,
+      "step": 14575
+    },
+    {
+      "epoch": 0.6502471830044982,
+      "grad_norm": 101.50527954101562,
+      "learning_rate": 1.0906354552056895e-05,
+      "loss": 0.0249,
+      "step": 14600
+    },
+    {
+      "epoch": 0.6513606199616978,
+      "grad_norm": 16.35914421081543,
+      "learning_rate": 1.0844100796310287e-05,
+      "loss": 0.023,
+      "step": 14625
+    },
+    {
+      "epoch": 0.6524740569188973,
+      "grad_norm": 0.3573369085788727,
+      "learning_rate": 1.078195906985021e-05,
+      "loss": 0.0342,
+      "step": 14650
+    },
+    {
+      "epoch": 0.6535874938760967,
+      "grad_norm": 0.01951185241341591,
+      "learning_rate": 1.0719930133027238e-05,
+      "loss": 0.0695,
+      "step": 14675
+    },
+    {
+      "epoch": 0.6547009308332962,
+      "grad_norm": 0.006902535445988178,
+      "learning_rate": 1.0658014744811892e-05,
+      "loss": 0.0251,
+      "step": 14700
+    },
+    {
+      "epoch": 0.6558143677904957,
+      "grad_norm": 17.231019973754883,
+      "learning_rate": 1.059621366278532e-05,
+      "loss": 0.1581,
+      "step": 14725
+    },
+    {
+      "epoch": 0.6569278047476952,
+      "grad_norm": 0.09472651034593582,
+      "learning_rate": 1.0534527643130082e-05,
+      "loss": 0.0712,
+      "step": 14750
+    },
+    {
+      "epoch": 0.6580412417048946,
+      "grad_norm": 0.022969521582126617,
+      "learning_rate": 1.0472957440620837e-05,
+      "loss": 0.097,
+      "step": 14775
+    },
+    {
+      "epoch": 0.6591546786620942,
+      "grad_norm": 0.44400086998939514,
+      "learning_rate": 1.0411503808615131e-05,
+      "loss": 0.0358,
+      "step": 14800
+    },
+    {
+      "epoch": 0.6602681156192937,
+      "grad_norm": 0.04555133730173111,
+      "learning_rate": 1.0350167499044216e-05,
+      "loss": 0.0573,
+      "step": 14825
+    },
+    {
+      "epoch": 0.6613815525764931,
+      "grad_norm": 0.021451586857438087,
+      "learning_rate": 1.028894926240378e-05,
+      "loss": 0.0344,
+      "step": 14850
+    },
+    {
+      "epoch": 0.6624949895336926,
+      "grad_norm": 9.742546081542969,
+      "learning_rate": 1.0227849847744817e-05,
+      "loss": 0.0721,
+      "step": 14875
+    },
+    {
+      "epoch": 0.6636084264908921,
+      "grad_norm": 2.3758695125579834,
+      "learning_rate": 1.0166870002664432e-05,
+      "loss": 0.0478,
+      "step": 14900
+    },
+    {
+      "epoch": 0.6647218634480916,
+      "grad_norm": 0.03618834540247917,
+      "learning_rate": 1.0106010473296737e-05,
+      "loss": 0.0711,
+      "step": 14925
+    },
+    {
+      "epoch": 0.665835300405291,
+      "grad_norm": 0.051780980080366135,
+      "learning_rate": 1.004527200430366e-05,
+      "loss": 0.0579,
+      "step": 14950
+    },
+    {
+      "epoch": 0.6669487373624905,
+      "grad_norm": 0.17397186160087585,
+      "learning_rate": 9.984655338865874e-06,
+      "loss": 0.0528,
+      "step": 14975
+    },
+    {
+      "epoch": 0.6680621743196901,
+      "grad_norm": 0.010688379406929016,
+      "learning_rate": 9.924161218673689e-06,
+      "loss": 0.0364,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6691756112768895,
+      "grad_norm": 0.10326721519231796,
+      "learning_rate": 9.863790383918007e-06,
+      "loss": 0.0841,
+      "step": 15025
+    },
+    {
+      "epoch": 0.670289048234089,
+      "grad_norm": 0.006623438559472561,
+      "learning_rate": 9.80354357328121e-06,
+      "loss": 0.1098,
+      "step": 15050
+    },
+    {
+      "epoch": 0.6714024851912884,
+      "grad_norm": 0.1552818864583969,
+      "learning_rate": 9.743421523928172e-06,
+      "loss": 0.0705,
+      "step": 15075
+    },
+    {
+      "epoch": 0.672515922148488,
+      "grad_norm": 0.34292566776275635,
+      "learning_rate": 9.683424971497195e-06,
+      "loss": 0.1153,
+      "step": 15100
+    },
+    {
+      "epoch": 0.6736293591056874,
+      "grad_norm": 0.5139630436897278,
+      "learning_rate": 9.623554650091066e-06,
+      "loss": 0.0384,
+      "step": 15125
+    },
+    {
+      "epoch": 0.6747427960628869,
+      "grad_norm": 0.44938117265701294,
+      "learning_rate": 9.563811292268019e-06,
+      "loss": 0.0338,
+      "step": 15150
+    },
+    {
+      "epoch": 0.6758562330200863,
+      "grad_norm": 0.04680267721414566,
+      "learning_rate": 9.504195629032795e-06,
+      "loss": 0.0282,
+      "step": 15175
+    },
+    {
+      "epoch": 0.6769696699772859,
+      "grad_norm": 0.7540410757064819,
+      "learning_rate": 9.444708389827693e-06,
+      "loss": 0.0763,
+      "step": 15200
+    },
+    {
+      "epoch": 0.6780831069344854,
+      "grad_norm": 0.027341516688466072,
+      "learning_rate": 9.385350302523667e-06,
+      "loss": 0.0413,
+      "step": 15225
+    },
+    {
+      "epoch": 0.6791965438916848,
+      "grad_norm": 39.69243240356445,
+      "learning_rate": 9.326122093411378e-06,
+      "loss": 0.0971,
+      "step": 15250
+    },
+    {
+      "epoch": 0.6803099808488844,
+      "grad_norm": 0.05220544710755348,
+      "learning_rate": 9.267024487192338e-06,
+      "loss": 0.0582,
+      "step": 15275
+    },
+    {
+      "epoch": 0.6814234178060838,
+      "grad_norm": 0.011502075009047985,
+      "learning_rate": 9.208058206970045e-06,
+      "loss": 0.0376,
+      "step": 15300
+    },
+    {
+      "epoch": 0.6825368547632833,
+      "grad_norm": 0.31533384323120117,
+      "learning_rate": 9.149223974241113e-06,
+      "loss": 0.1116,
+      "step": 15325
+    },
+    {
+      "epoch": 0.6836502917204828,
+      "grad_norm": 0.1892848163843155,
+      "learning_rate": 9.090522508886457e-06,
+      "loss": 0.0357,
+      "step": 15350
+    },
+    {
+      "epoch": 0.6847637286776823,
+      "grad_norm": 0.021438557654619217,
+      "learning_rate": 9.031954529162477e-06,
+      "loss": 0.0801,
+      "step": 15375
+    },
+    {
+      "epoch": 0.6858771656348818,
+      "grad_norm": 0.0905388817191124,
+      "learning_rate": 8.973520751692304e-06,
+      "loss": 0.0151,
+      "step": 15400
+    },
+    {
+      "epoch": 0.6869906025920812,
+      "grad_norm": 0.02208125777542591,
+      "learning_rate": 8.915221891456974e-06,
+      "loss": 0.0501,
+      "step": 15425
+    },
+    {
+      "epoch": 0.6881040395492807,
+      "grad_norm": 0.3321913480758667,
+      "learning_rate": 8.857058661786722e-06,
+      "loss": 0.0797,
+      "step": 15450
+    },
+    {
+      "epoch": 0.6892174765064802,
+      "grad_norm": 0.16185687482357025,
+      "learning_rate": 8.799031774352231e-06,
+      "loss": 0.0608,
+      "step": 15475
+    },
+    {
+      "epoch": 0.6903309134636797,
+      "grad_norm": 0.011803328059613705,
+      "learning_rate": 8.741141939155961e-06,
+      "loss": 0.013,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6914443504208792,
+      "grad_norm": 0.013166255317628384,
+      "learning_rate": 8.683389864523409e-06,
+      "loss": 0.0187,
+      "step": 15525
+    },
+    {
+      "epoch": 0.6925577873780786,
+      "grad_norm": 0.06444378197193146,
+      "learning_rate": 8.625776257094477e-06,
+      "loss": 0.0295,
+      "step": 15550
+    },
+    {
+      "epoch": 0.6936712243352782,
+      "grad_norm": 20.706222534179688,
+      "learning_rate": 8.568301821814808e-06,
+      "loss": 0.0995,
+      "step": 15575
+    },
+    {
+      "epoch": 0.6947846612924776,
+      "grad_norm": 51.754207611083984,
+      "learning_rate": 8.510967261927199e-06,
+      "loss": 0.1911,
+      "step": 15600
+    },
+    {
+      "epoch": 0.6958980982496771,
+      "grad_norm": 0.015366621315479279,
+      "learning_rate": 8.453773278962932e-06,
+      "loss": 0.067,
+      "step": 15625
+    },
+    {
+      "epoch": 0.6970115352068765,
+      "grad_norm": 8.943309783935547,
+      "learning_rate": 8.396720572733243e-06,
+      "loss": 0.117,
+      "step": 15650
+    },
+    {
+      "epoch": 0.6981249721640761,
+      "grad_norm": 0.11915289610624313,
+      "learning_rate": 8.339809841320732e-06,
+      "loss": 0.0645,
+      "step": 15675
+    },
+    {
+      "epoch": 0.6992384091212756,
+      "grad_norm": 0.09671024978160858,
+      "learning_rate": 8.283041781070847e-06,
+      "loss": 0.0928,
+      "step": 15700
+    },
+    {
+      "epoch": 0.700351846078475,
+      "grad_norm": 0.03198957443237305,
+      "learning_rate": 8.22641708658333e-06,
+      "loss": 0.1054,
+      "step": 15725
+    },
+    {
+      "epoch": 0.7014652830356746,
+      "grad_norm": 68.7297592163086,
+      "learning_rate": 8.169936450703737e-06,
+      "loss": 0.0969,
+      "step": 15750
+    },
+    {
+      "epoch": 0.702578719992874,
+      "grad_norm": 0.02619253285229206,
+      "learning_rate": 8.113600564514979e-06,
+      "loss": 0.0013,
+      "step": 15775
+    },
+    {
+      "epoch": 0.7036921569500735,
+      "grad_norm": 0.6842272281646729,
+      "learning_rate": 8.057410117328823e-06,
+      "loss": 0.0443,
+      "step": 15800
+    },
+    {
+      "epoch": 0.7048055939072729,
+      "grad_norm": 132.8560333251953,
+      "learning_rate": 8.00136579667749e-06,
+      "loss": 0.0347,
+      "step": 15825
+    },
+    {
+      "epoch": 0.7059190308644725,
+      "grad_norm": 0.31542474031448364,
+      "learning_rate": 7.945468288305226e-06,
+      "loss": 0.0402,
+      "step": 15850
+    },
+    {
+      "epoch": 0.707032467821672,
+      "grad_norm": 0.1650291383266449,
+      "learning_rate": 7.889718276159937e-06,
+      "loss": 0.1061,
+      "step": 15875
+    },
+    {
+      "epoch": 0.7081459047788714,
+      "grad_norm": 0.1369670331478119,
+      "learning_rate": 7.834116442384785e-06,
+      "loss": 0.0693,
+      "step": 15900
+    },
+    {
+      "epoch": 0.7092593417360709,
+      "grad_norm": 17.667673110961914,
+      "learning_rate": 7.778663467309865e-06,
+      "loss": 0.047,
+      "step": 15925
+    },
+    {
+      "epoch": 0.7103727786932704,
+      "grad_norm": 10.790023803710938,
+      "learning_rate": 7.723360029443865e-06,
+      "loss": 0.126,
+      "step": 15950
+    },
+    {
+      "epoch": 0.7114862156504699,
+      "grad_norm": 0.010408015921711922,
+      "learning_rate": 7.668206805465797e-06,
+      "loss": 0.0677,
+      "step": 15975
+    },
+    {
+      "epoch": 0.7125996526076693,
+      "grad_norm": 15.175890922546387,
+      "learning_rate": 7.613204470216671e-06,
+      "loss": 0.0762,
+      "step": 16000
+    },
+    {
+      "epoch": 0.7137130895648688,
+      "grad_norm": 0.09815490990877151,
+      "learning_rate": 7.558353696691267e-06,
+      "loss": 0.033,
+      "step": 16025
+    },
+    {
+      "epoch": 0.7148265265220684,
+      "grad_norm": 0.23362314701080322,
+      "learning_rate": 7.503655156029885e-06,
+      "loss": 0.0456,
+      "step": 16050
+    },
+    {
+      "epoch": 0.7159399634792678,
+      "grad_norm": 0.04253578558564186,
+      "learning_rate": 7.449109517510171e-06,
+      "loss": 0.0295,
+      "step": 16075
+    },
+    {
+      "epoch": 0.7170534004364673,
+      "grad_norm": 0.08314715325832367,
+      "learning_rate": 7.3947174485388666e-06,
+      "loss": 0.0577,
+      "step": 16100
+    },
+    {
+      "epoch": 0.7181668373936668,
+      "grad_norm": 0.19052563607692719,
+      "learning_rate": 7.340479614643681e-06,
+      "loss": 0.0474,
+      "step": 16125
+    },
+    {
+      "epoch": 0.7192802743508663,
+      "grad_norm": 1.0644758939743042,
+      "learning_rate": 7.286396679465164e-06,
+      "loss": 0.1031,
+      "step": 16150
+    },
+    {
+      "epoch": 0.7203937113080657,
+      "grad_norm": 0.02732575498521328,
+      "learning_rate": 7.232469304748539e-06,
+      "loss": 0.0895,
+      "step": 16175
+    },
+    {
+      "epoch": 0.7215071482652652,
+      "grad_norm": 83.3663101196289,
+      "learning_rate": 7.178698150335641e-06,
+      "loss": 0.083,
+      "step": 16200
+    },
+    {
+      "epoch": 0.7226205852224648,
+      "grad_norm": 0.3653166890144348,
+      "learning_rate": 7.125083874156819e-06,
+      "loss": 0.0565,
+      "step": 16225
+    },
+    {
+      "epoch": 0.7237340221796642,
+      "grad_norm": 0.18628665804862976,
+      "learning_rate": 7.071627132222931e-06,
+      "loss": 0.0314,
+      "step": 16250
+    },
+    {
+      "epoch": 0.7248474591368637,
+      "grad_norm": 0.016880426555871964,
+      "learning_rate": 7.018328578617253e-06,
+      "loss": 0.0562,
+      "step": 16275
+    },
+    {
+      "epoch": 0.7259608960940631,
+      "grad_norm": 0.044409096240997314,
+      "learning_rate": 6.965188865487522e-06,
+      "loss": 0.0471,
+      "step": 16300
+    },
+    {
+      "epoch": 0.7270743330512627,
+      "grad_norm": 21.80831527709961,
+      "learning_rate": 6.912208643037937e-06,
+      "loss": 0.1115,
+      "step": 16325
+    },
+    {
+      "epoch": 0.7281877700084621,
+      "grad_norm": 0.017523668706417084,
+      "learning_rate": 6.859388559521225e-06,
+      "loss": 0.0268,
+      "step": 16350
+    },
+    {
+      "epoch": 0.7293012069656616,
+      "grad_norm": 201.114013671875,
+      "learning_rate": 6.806729261230676e-06,
+      "loss": 0.0669,
+      "step": 16375
+    },
+    {
+      "epoch": 0.730414643922861,
+      "grad_norm": 0.01862688735127449,
+      "learning_rate": 6.754231392492254e-06,
+      "loss": 0.0437,
+      "step": 16400
+    },
+    {
+      "epoch": 0.7315280808800606,
+      "grad_norm": 0.026870861649513245,
+      "learning_rate": 6.701895595656715e-06,
+      "loss": 0.0957,
+      "step": 16425
+    },
+    {
+      "epoch": 0.7326415178372601,
+      "grad_norm": 0.07133554667234421,
+      "learning_rate": 6.6497225110917515e-06,
+      "loss": 0.0566,
+      "step": 16450
+    },
+    {
+      "epoch": 0.7337549547944595,
+      "grad_norm": 48.52125930786133,
+      "learning_rate": 6.597712777174141e-06,
+      "loss": 0.0401,
+      "step": 16475
+    },
+    {
+      "epoch": 0.734868391751659,
+      "grad_norm": 20.256444931030273,
+      "learning_rate": 6.54586703028194e-06,
+      "loss": 0.0592,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7359818287088585,
+      "grad_norm": 22.120195388793945,
+      "learning_rate": 6.494185904786703e-06,
+      "loss": 0.1089,
+      "step": 16525
+    },
+    {
+      "epoch": 0.737095265666058,
+      "grad_norm": 0.08727439492940903,
+      "learning_rate": 6.442670033045731e-06,
+      "loss": 0.0479,
+      "step": 16550
+    },
+    {
+      "epoch": 0.7382087026232574,
+      "grad_norm": 0.47929519414901733,
+      "learning_rate": 6.391320045394305e-06,
+      "loss": 0.1064,
+      "step": 16575
+    },
+    {
+      "epoch": 0.739322139580457,
+      "grad_norm": 0.8217176795005798,
+      "learning_rate": 6.340136570137991e-06,
+      "loss": 0.0677,
+      "step": 16600
+    },
+    {
+      "epoch": 0.7404355765376565,
+      "grad_norm": 16.746612548828125,
+      "learning_rate": 6.28912023354497e-06,
+      "loss": 0.0475,
+      "step": 16625
+    },
+    {
+      "epoch": 0.7415490134948559,
+      "grad_norm": 0.14232224225997925,
+      "learning_rate": 6.2382716598383355e-06,
+      "loss": 0.0882,
+      "step": 16650
+    },
+    {
+      "epoch": 0.7426624504520554,
+      "grad_norm": 0.030962107703089714,
+      "learning_rate": 6.187591471188481e-06,
+      "loss": 0.0495,
+      "step": 16675
+    },
+    {
+      "epoch": 0.7437758874092549,
+      "grad_norm": 0.061680518090724945,
+      "learning_rate": 6.137080287705481e-06,
+      "loss": 0.0929,
+      "step": 16700
+    },
+    {
+      "epoch": 0.7448893243664544,
+      "grad_norm": 17.700639724731445,
+      "learning_rate": 6.086738727431523e-06,
+      "loss": 0.0528,
+      "step": 16725
+    },
+    {
+      "epoch": 0.7460027613236538,
+      "grad_norm": 1.2548288106918335,
+      "learning_rate": 6.0365674063333046e-06,
+      "loss": 0.0833,
+      "step": 16750
+    },
+    {
+      "epoch": 0.7471161982808533,
+      "grad_norm": 0.4032031297683716,
+      "learning_rate": 5.986566938294532e-06,
+      "loss": 0.0506,
+      "step": 16775
+    },
+    {
+      "epoch": 0.7482296352380529,
+      "grad_norm": 0.03656847029924393,
+      "learning_rate": 5.936737935108383e-06,
+      "loss": 0.0262,
+      "step": 16800
+    },
+    {
+      "epoch": 0.7493430721952523,
+      "grad_norm": 0.02097596973180771,
+      "learning_rate": 5.887081006470061e-06,
+      "loss": 0.058,
+      "step": 16825
+    },
+    {
+      "epoch": 0.7504565091524518,
+      "grad_norm": 0.11637075990438461,
+      "learning_rate": 5.837596759969281e-06,
+      "loss": 0.0949,
+      "step": 16850
+    },
+    {
+      "epoch": 0.7515699461096512,
+      "grad_norm": 0.06315034627914429,
+      "learning_rate": 5.788285801082878e-06,
+      "loss": 0.0304,
+      "step": 16875
+    },
+    {
+      "epoch": 0.7526833830668508,
+      "grad_norm": 0.11649344861507416,
+      "learning_rate": 5.73914873316737e-06,
+      "loss": 0.0368,
+      "step": 16900
+    },
+    {
+      "epoch": 0.7537968200240502,
+      "grad_norm": 0.03985821083188057,
+      "learning_rate": 5.6901861574516115e-06,
+      "loss": 0.0074,
+      "step": 16925
+    },
+    {
+      "epoch": 0.7549102569812497,
+      "grad_norm": 0.016696982085704803,
+      "learning_rate": 5.64139867302939e-06,
+      "loss": 0.0675,
+      "step": 16950
+    },
+    {
+      "epoch": 0.7560236939384493,
+      "grad_norm": 0.011424514465034008,
+      "learning_rate": 5.592786876852127e-06,
+      "loss": 0.0303,
+      "step": 16975
+    },
+    {
+      "epoch": 0.7571371308956487,
+      "grad_norm": 15.772963523864746,
+      "learning_rate": 5.54435136372158e-06,
+      "loss": 0.0487,
+      "step": 17000
+    },
+    {
+      "epoch": 0.7582505678528482,
+      "grad_norm": 0.7694092392921448,
+      "learning_rate": 5.496092726282538e-06,
+      "loss": 0.0411,
+      "step": 17025
+    },
+    {
+      "epoch": 0.7593640048100476,
+      "grad_norm": 14.076139450073242,
+      "learning_rate": 5.448011555015584e-06,
+      "loss": 0.1227,
+      "step": 17050
+    },
+    {
+      "epoch": 0.7604774417672472,
+      "grad_norm": 0.1597975343465805,
+      "learning_rate": 5.400108438229865e-06,
+      "loss": 0.0503,
+      "step": 17075
+    },
+    {
+      "epoch": 0.7615908787244466,
+      "grad_norm": 0.08557901531457901,
+      "learning_rate": 5.352383962055918e-06,
+      "loss": 0.0788,
+      "step": 17100
+    },
+    {
+      "epoch": 0.7627043156816461,
+      "grad_norm": 19.060636520385742,
+      "learning_rate": 5.304838710438458e-06,
+      "loss": 0.0935,
+      "step": 17125
+    },
+    {
+      "epoch": 0.7638177526388455,
+      "grad_norm": 0.024234607815742493,
+      "learning_rate": 5.2574732651292606e-06,
+      "loss": 0.0575,
+      "step": 17150
+    },
+    {
+      "epoch": 0.7649311895960451,
+      "grad_norm": 0.010592158883810043,
+      "learning_rate": 5.210288205680032e-06,
+      "loss": 0.0992,
+      "step": 17175
+    },
+    {
+      "epoch": 0.7660446265532446,
+      "grad_norm": 0.08044129610061646,
+      "learning_rate": 5.163284109435336e-06,
+      "loss": 0.1277,
+      "step": 17200
+    },
+    {
+      "epoch": 0.767158063510444,
+      "grad_norm": 15.679009437561035,
+      "learning_rate": 5.116461551525503e-06,
+      "loss": 0.0561,
+      "step": 17225
+    },
+    {
+      "epoch": 0.7682715004676435,
+      "grad_norm": 0.5942630171775818,
+      "learning_rate": 5.069821104859607e-06,
+      "loss": 0.0234,
+      "step": 17250
+    },
+    {
+      "epoch": 0.769384937424843,
+      "grad_norm": 0.06479264050722122,
+      "learning_rate": 5.023363340118452e-06,
+      "loss": 0.1106,
+      "step": 17275
+    },
+    {
+      "epoch": 0.7704983743820425,
+      "grad_norm": 10.735722541809082,
+      "learning_rate": 4.977088825747609e-06,
+      "loss": 0.0826,
+      "step": 17300
+    },
+    {
+      "epoch": 0.771611811339242,
+      "grad_norm": 0.09754932671785355,
+      "learning_rate": 4.930998127950424e-06,
+      "loss": 0.0574,
+      "step": 17325
+    },
+    {
+      "epoch": 0.7727252482964414,
+      "grad_norm": 1.6428059339523315,
+      "learning_rate": 4.885091810681118e-06,
+      "loss": 0.0551,
+      "step": 17350
+    },
+    {
+      "epoch": 0.773838685253641,
+      "grad_norm": 0.011265415698289871,
+      "learning_rate": 4.839370435637869e-06,
+      "loss": 0.0587,
+      "step": 17375
+    },
+    {
+      "epoch": 0.7749521222108404,
+      "grad_norm": 0.5027729272842407,
+      "learning_rate": 4.793834562255968e-06,
+      "loss": 0.0549,
+      "step": 17400
+    },
+    {
+      "epoch": 0.7760655591680399,
+      "grad_norm": 0.034433938562870026,
+      "learning_rate": 4.748484747700937e-06,
+      "loss": 0.0857,
+      "step": 17425
+    },
+    {
+      "epoch": 0.7771789961252394,
+      "grad_norm": 0.39721280336380005,
+      "learning_rate": 4.703321546861728e-06,
+      "loss": 0.0136,
+      "step": 17450
+    },
+    {
+      "epoch": 0.7782924330824389,
+      "grad_norm": 11.144375801086426,
+      "learning_rate": 4.658345512343954e-06,
+      "loss": 0.0919,
+      "step": 17475
+    },
+    {
+      "epoch": 0.7794058700396383,
+      "grad_norm": 0.03096114657819271,
+      "learning_rate": 4.613557194463083e-06,
+      "loss": 0.1601,
+      "step": 17500
+    },
+    {
+      "epoch": 0.7805193069968378,
+      "grad_norm": 0.29175180196762085,
+      "learning_rate": 4.5689571412377444e-06,
+      "loss": 0.1204,
+      "step": 17525
+    },
+    {
+      "epoch": 0.7816327439540374,
+      "grad_norm": 0.077189140021801,
+      "learning_rate": 4.524545898382999e-06,
+      "loss": 0.0422,
+      "step": 17550
+    },
+    {
+      "epoch": 0.7827461809112368,
+      "grad_norm": 0.04853476583957672,
+      "learning_rate": 4.4803240093036895e-06,
+      "loss": 0.0418,
+      "step": 17575
+    },
+    {
+      "epoch": 0.7838596178684363,
+      "grad_norm": 0.25351813435554504,
+      "learning_rate": 4.436292015087751e-06,
+      "loss": 0.1125,
+      "step": 17600
+    },
+    {
+      "epoch": 0.7849730548256357,
+      "grad_norm": 0.33566001057624817,
+      "learning_rate": 4.392450454499624e-06,
+      "loss": 0.1044,
+      "step": 17625
+    },
+    {
+      "epoch": 0.7860864917828353,
+      "grad_norm": 0.33860525488853455,
+      "learning_rate": 4.348799863973645e-06,
+      "loss": 0.0398,
+      "step": 17650
+    },
+    {
+      "epoch": 0.7871999287400347,
+      "grad_norm": 0.019481919705867767,
+      "learning_rate": 4.305340777607503e-06,
+      "loss": 0.114,
+      "step": 17675
+    },
+    {
+      "epoch": 0.7883133656972342,
+      "grad_norm": 0.05631513521075249,
+      "learning_rate": 4.262073727155673e-06,
+      "loss": 0.0805,
+      "step": 17700
+    },
+    {
+      "epoch": 0.7894268026544337,
+      "grad_norm": 0.23303325474262238,
+      "learning_rate": 4.218999242022936e-06,
+      "loss": 0.0584,
+      "step": 17725
+    },
+    {
+      "epoch": 0.7905402396116332,
+      "grad_norm": 0.4254063665866852,
+      "learning_rate": 4.176117849257884e-06,
+      "loss": 0.0253,
+      "step": 17750
+    },
+    {
+      "epoch": 0.7916536765688327,
+      "grad_norm": 0.07280249893665314,
+      "learning_rate": 4.133430073546496e-06,
+      "loss": 0.017,
+      "step": 17775
+    },
+    {
+      "epoch": 0.7927671135260321,
+      "grad_norm": 0.09632007777690887,
+      "learning_rate": 4.090936437205683e-06,
+      "loss": 0.0441,
+      "step": 17800
+    },
+    {
+      "epoch": 0.7938805504832316,
+      "grad_norm": 0.028888842090964317,
+      "learning_rate": 4.048637460176925e-06,
+      "loss": 0.0914,
+      "step": 17825
+    },
+    {
+      "epoch": 0.7949939874404311,
+      "grad_norm": 0.27641311287879944,
+      "learning_rate": 4.0065336600198886e-06,
+      "loss": 0.041,
+      "step": 17850
+    },
+    {
+      "epoch": 0.7961074243976306,
+      "grad_norm": 0.08862696588039398,
+      "learning_rate": 3.96462555190613e-06,
+      "loss": 0.0638,
+      "step": 17875
+    },
+    {
+      "epoch": 0.7972208613548301,
+      "grad_norm": 0.021214209496974945,
+      "learning_rate": 3.922913648612738e-06,
+      "loss": 0.0523,
+      "step": 17900
+    },
+    {
+      "epoch": 0.7983342983120296,
+      "grad_norm": 0.4044612646102905,
+      "learning_rate": 3.881398460516106e-06,
+      "loss": 0.0566,
+      "step": 17925
+    },
+    {
+      "epoch": 0.7994477352692291,
+      "grad_norm": 0.13831877708435059,
+      "learning_rate": 3.840080495585659e-06,
+      "loss": 0.0788,
+      "step": 17950
+    },
+    {
+      "epoch": 0.8005611722264285,
+      "grad_norm": 0.05166686698794365,
+      "learning_rate": 3.7989602593776597e-06,
+      "loss": 0.03,
+      "step": 17975
+    },
+    {
+      "epoch": 0.801674609183628,
+      "grad_norm": 0.5509428381919861,
+      "learning_rate": 3.758038255029006e-06,
+      "loss": 0.0587,
+      "step": 18000
+    },
+    {
+      "epoch": 0.8027880461408275,
+      "grad_norm": 0.0240898709744215,
+      "learning_rate": 3.717314983251077e-06,
+      "loss": 0.0805,
+      "step": 18025
+    },
+    {
+      "epoch": 0.803901483098027,
+      "grad_norm": 0.01613386906683445,
+      "learning_rate": 3.6767909423236115e-06,
+      "loss": 0.0789,
+      "step": 18050
+    },
+    {
+      "epoch": 0.8050149200552265,
+      "grad_norm": 0.010899515822529793,
+      "learning_rate": 3.6364666280886194e-06,
+      "loss": 0.0338,
+      "step": 18075
+    },
+    {
+      "epoch": 0.8061283570124259,
+      "grad_norm": 0.04010821506381035,
+      "learning_rate": 3.5963425339442925e-06,
+      "loss": 0.0343,
+      "step": 18100
+    },
+    {
+      "epoch": 0.8072417939696255,
+      "grad_norm": 0.034061819314956665,
+      "learning_rate": 3.5564191508389855e-06,
+      "loss": 0.0384,
+      "step": 18125
+    },
+    {
+      "epoch": 0.8083552309268249,
+      "grad_norm": 0.018084630370140076,
+      "learning_rate": 3.516696967265203e-06,
+      "loss": 0.0813,
+      "step": 18150
+    },
+    {
+      "epoch": 0.8094686678840244,
+      "grad_norm": 0.07554034888744354,
+      "learning_rate": 3.4771764692536135e-06,
+      "loss": 0.0494,
+      "step": 18175
+    },
+    {
+      "epoch": 0.8105821048412238,
+      "grad_norm": 0.5343236923217773,
+      "learning_rate": 3.4378581403671294e-06,
+      "loss": 0.0743,
+      "step": 18200
+    },
+    {
+      "epoch": 0.8116955417984234,
+      "grad_norm": 10.600082397460938,
+      "learning_rate": 3.3987424616949617e-06,
+      "loss": 0.1109,
+      "step": 18225
+    },
+    {
+      "epoch": 0.8128089787556229,
+      "grad_norm": 0.37871259450912476,
+      "learning_rate": 3.3598299118467413e-06,
+      "loss": 0.0192,
+      "step": 18250
+    },
+    {
+      "epoch": 0.8139224157128223,
+      "grad_norm": 0.024805864319205284,
+      "learning_rate": 3.321120966946667e-06,
+      "loss": 0.0009,
+      "step": 18275
+    },
+    {
+      "epoch": 0.8150358526700219,
+      "grad_norm": 0.11892875283956528,
+      "learning_rate": 3.2826161006276913e-06,
+      "loss": 0.071,
+      "step": 18300
+    },
+    {
+      "epoch": 0.8161492896272213,
+      "grad_norm": 0.11794514954090118,
+      "learning_rate": 3.2443157840256957e-06,
+      "loss": 0.0547,
+      "step": 18325
+    },
+    {
+      "epoch": 0.8172627265844208,
+      "grad_norm": 0.048415035009384155,
+      "learning_rate": 3.2062204857737476e-06,
+      "loss": 0.0293,
+      "step": 18350
+    },
+    {
+      "epoch": 0.8183761635416202,
+      "grad_norm": 0.09221319109201431,
+      "learning_rate": 3.1683306719963604e-06,
+      "loss": 0.0468,
+      "step": 18375
+    },
+    {
+      "epoch": 0.8194896004988198,
+      "grad_norm": 0.009933177381753922,
+      "learning_rate": 3.130646806303803e-06,
+      "loss": 0.0501,
+      "step": 18400
+    },
+    {
+      "epoch": 0.8206030374560193,
+      "grad_norm": 0.2183195799589157,
+      "learning_rate": 3.0931693497864023e-06,
+      "loss": 0.0534,
+      "step": 18425
+    },
+    {
+      "epoch": 0.8217164744132187,
+      "grad_norm": 0.04401267319917679,
+      "learning_rate": 3.0558987610089132e-06,
+      "loss": 0.0671,
+      "step": 18450
+    },
+    {
+      "epoch": 0.8228299113704182,
+      "grad_norm": 0.10095870494842529,
+      "learning_rate": 3.0188354960049105e-06,
+      "loss": 0.0267,
+      "step": 18475
+    },
+    {
+      "epoch": 0.8239433483276177,
+      "grad_norm": 37.297325134277344,
+      "learning_rate": 2.9819800082712215e-06,
+      "loss": 0.0325,
+      "step": 18500
+    },
+    {
+      "epoch": 0.8250567852848172,
+      "grad_norm": 0.0290220957249403,
+      "learning_rate": 2.945332748762344e-06,
+      "loss": 0.0447,
+      "step": 18525
+    },
+    {
+      "epoch": 0.8261702222420166,
+      "grad_norm": 17.21735382080078,
+      "learning_rate": 2.908894165884957e-06,
+      "loss": 0.0511,
+      "step": 18550
+    },
+    {
+      "epoch": 0.8272836591992161,
+      "grad_norm": 0.03269251063466072,
+      "learning_rate": 2.872664705492416e-06,
+      "loss": 0.0893,
+      "step": 18575
+    },
+    {
+      "epoch": 0.8283970961564157,
+      "grad_norm": 0.0077833011746406555,
+      "learning_rate": 2.8366448108793233e-06,
+      "loss": 0.0283,
+      "step": 18600
+    },
+    {
+      "epoch": 0.8295105331136151,
+      "grad_norm": 0.43284299969673157,
+      "learning_rate": 2.80083492277607e-06,
+      "loss": 0.0557,
+      "step": 18625
+    },
+    {
+      "epoch": 0.8306239700708146,
+      "grad_norm": 0.02338644489645958,
+      "learning_rate": 2.7652354793434666e-06,
+      "loss": 0.1057,
+      "step": 18650
+    },
+    {
+      "epoch": 0.831737407028014,
+      "grad_norm": 0.03380841761827469,
+      "learning_rate": 2.7298469161673733e-06,
+      "loss": 0.0813,
+      "step": 18675
+    },
+    {
+      "epoch": 0.8328508439852136,
+      "grad_norm": 0.13972997665405273,
+      "learning_rate": 2.6946696662533824e-06,
+      "loss": 0.06,
+      "step": 18700
+    },
+    {
+      "epoch": 0.833964280942413,
+      "grad_norm": 0.06724318861961365,
+      "learning_rate": 2.6597041600214947e-06,
+      "loss": 0.0262,
+      "step": 18725
+    },
+    {
+      "epoch": 0.8350777178996125,
+      "grad_norm": 0.022787462919950485,
+      "learning_rate": 2.6249508253008806e-06,
+      "loss": 0.1003,
+      "step": 18750
+    },
+    {
+      "epoch": 0.8361911548568121,
+      "grad_norm": 0.010312095284461975,
+      "learning_rate": 2.5904100873246195e-06,
+      "loss": 0.1163,
+      "step": 18775
+    },
+    {
+      "epoch": 0.8373045918140115,
+      "grad_norm": 3.376255989074707,
+      "learning_rate": 2.5560823687245282e-06,
+      "loss": 0.0964,
+      "step": 18800
+    },
+    {
+      "epoch": 0.838418028771211,
+      "grad_norm": 0.20705439150333405,
+      "learning_rate": 2.5219680895259612e-06,
+      "loss": 0.0264,
+      "step": 18825
+    },
+    {
+      "epoch": 0.8395314657284104,
+      "grad_norm": 0.05201965197920799,
+      "learning_rate": 2.488067667142682e-06,
+      "loss": 0.0483,
+      "step": 18850
+    },
+    {
+      "epoch": 0.84064490268561,
+      "grad_norm": 15.602285385131836,
+      "learning_rate": 2.454381516371753e-06,
+      "loss": 0.0855,
+      "step": 18875
+    },
+    {
+      "epoch": 0.8417583396428094,
+      "grad_norm": 14.771347045898438,
+      "learning_rate": 2.4209100493884763e-06,
+      "loss": 0.1803,
+      "step": 18900
+    },
+    {
+      "epoch": 0.8428717766000089,
+      "grad_norm": 0.04606569930911064,
+      "learning_rate": 2.3876536757413218e-06,
+      "loss": 0.0739,
+      "step": 18925
+    },
+    {
+      "epoch": 0.8439852135572083,
+      "grad_norm": 0.016485804691910744,
+      "learning_rate": 2.3546128023469383e-06,
+      "loss": 0.0433,
+      "step": 18950
+    },
+    {
+      "epoch": 0.8450986505144079,
+      "grad_norm": 0.07802044600248337,
+      "learning_rate": 2.321787833485163e-06,
+      "loss": 0.0248,
+      "step": 18975
+    },
+    {
+      "epoch": 0.8462120874716074,
+      "grad_norm": 0.017175985500216484,
+      "learning_rate": 2.2891791707940913e-06,
+      "loss": 0.0322,
+      "step": 19000
+    },
+    {
+      "epoch": 0.8473255244288068,
+      "grad_norm": 0.31442588567733765,
+      "learning_rate": 2.2567872132651393e-06,
+      "loss": 0.0566,
+      "step": 19025
+    },
+    {
+      "epoch": 0.8484389613860063,
+      "grad_norm": 0.009995149448513985,
+      "learning_rate": 2.2246123572381718e-06,
+      "loss": 0.0276,
+      "step": 19050
+    },
+    {
+      "epoch": 0.8495523983432058,
+      "grad_norm": 0.025844572111964226,
+      "learning_rate": 2.192654996396659e-06,
+      "loss": 0.079,
+      "step": 19075
+    },
+    {
+      "epoch": 0.8506658353004053,
+      "grad_norm": 0.1997177004814148,
+      "learning_rate": 2.1609155217628584e-06,
+      "loss": 0.0545,
+      "step": 19100
+    },
+    {
+      "epoch": 0.8517792722576047,
+      "grad_norm": 0.11774219572544098,
+      "learning_rate": 2.129394321693021e-06,
+      "loss": 0.0845,
+      "step": 19125
+    },
+    {
+      "epoch": 0.8528927092148043,
+      "grad_norm": 0.8494458198547363,
+      "learning_rate": 2.09809178187264e-06,
+      "loss": 0.0629,
+      "step": 19150
+    },
+    {
+      "epoch": 0.8540061461720038,
+      "grad_norm": 16.696989059448242,
+      "learning_rate": 2.067008285311751e-06,
+      "loss": 0.0306,
+      "step": 19175
+    },
+    {
+      "epoch": 0.8551195831292032,
+      "grad_norm": 0.04906664416193962,
+      "learning_rate": 2.0361442123402164e-06,
+      "loss": 0.0913,
+      "step": 19200
+    },
+    {
+      "epoch": 0.8562330200864027,
+      "grad_norm": 13.342795372009277,
+      "learning_rate": 2.005499940603097e-06,
+      "loss": 0.1194,
+      "step": 19225
+    },
+    {
+      "epoch": 0.8573464570436022,
+      "grad_norm": 0.04243914783000946,
+      "learning_rate": 1.9750758450560116e-06,
+      "loss": 0.0775,
+      "step": 19250
+    },
+    {
+      "epoch": 0.8584598940008017,
+      "grad_norm": 0.024273626506328583,
+      "learning_rate": 1.9448722979605692e-06,
+      "loss": 0.0922,
+      "step": 19275
+    },
+    {
+      "epoch": 0.8595733309580011,
+      "grad_norm": 15.679910659790039,
+      "learning_rate": 1.9148896688797956e-06,
+      "loss": 0.0433,
+      "step": 19300
+    },
+    {
+      "epoch": 0.8606867679152006,
+      "grad_norm": 0.013732674531638622,
+      "learning_rate": 1.8851283246736218e-06,
+      "loss": 0.0515,
+      "step": 19325
+    },
+    {
+      "epoch": 0.8618002048724002,
+      "grad_norm": 0.02653571590781212,
+      "learning_rate": 1.855588629494387e-06,
+      "loss": 0.0831,
+      "step": 19350
+    },
+    {
+      "epoch": 0.8629136418295996,
+      "grad_norm": 18.192138671875,
+      "learning_rate": 1.8262709447824001e-06,
+      "loss": 0.0354,
+      "step": 19375
+    },
+    {
+      "epoch": 0.8640270787867991,
+      "grad_norm": 1.8473023176193237,
+      "learning_rate": 1.7971756292614961e-06,
+      "loss": 0.0811,
+      "step": 19400
+    },
+    {
+      "epoch": 0.8651405157439985,
+      "grad_norm": 15.879515647888184,
+      "learning_rate": 1.7683030389346577e-06,
+      "loss": 0.0214,
+      "step": 19425
+    },
+    {
+      "epoch": 0.8662539527011981,
+      "grad_norm": 12.173957824707031,
+      "learning_rate": 1.739653527079659e-06,
+      "loss": 0.0837,
+      "step": 19450
+    },
+    {
+      "epoch": 0.8673673896583975,
+      "grad_norm": 12.342336654663086,
+      "learning_rate": 1.7112274442447475e-06,
+      "loss": 0.0446,
+      "step": 19475
+    },
+    {
+      "epoch": 0.868480826615597,
+      "grad_norm": 0.009152592159807682,
+      "learning_rate": 1.6830251382443408e-06,
+      "loss": 0.0659,
+      "step": 19500
+    },
+    {
+      "epoch": 0.8695942635727965,
+      "grad_norm": 0.06887594610452652,
+      "learning_rate": 1.655046954154782e-06,
+      "loss": 0.1522,
+      "step": 19525
+    },
+    {
+      "epoch": 0.870707700529996,
+      "grad_norm": 0.007866369560360909,
+      "learning_rate": 1.6272932343101122e-06,
+      "loss": 0.0102,
+      "step": 19550
+    },
+    {
+      "epoch": 0.8718211374871955,
+      "grad_norm": 0.30307716131210327,
+      "learning_rate": 1.599764318297894e-06,
+      "loss": 0.0528,
+      "step": 19575
+    },
+    {
+      "epoch": 0.8729345744443949,
+      "grad_norm": 0.057492926716804504,
+      "learning_rate": 1.5724605429550366e-06,
+      "loss": 0.0162,
+      "step": 19600
+    },
+    {
+      "epoch": 0.8740480114015945,
+      "grad_norm": 0.0504818893969059,
+      "learning_rate": 1.545382242363682e-06,
+      "loss": 0.0132,
+      "step": 19625
+    },
+    {
+      "epoch": 0.875161448358794,
+      "grad_norm": 0.007792500779032707,
+      "learning_rate": 1.5185297478471373e-06,
+      "loss": 0.0239,
+      "step": 19650
+    },
+    {
+      "epoch": 0.8762748853159934,
+      "grad_norm": 0.04944828525185585,
+      "learning_rate": 1.4919033879657852e-06,
+      "loss": 0.0283,
+      "step": 19675
+    },
+    {
+      "epoch": 0.8773883222731929,
+      "grad_norm": 0.08414191752672195,
+      "learning_rate": 1.4655034885130893e-06,
+      "loss": 0.0719,
+      "step": 19700
+    },
+    {
+      "epoch": 0.8785017592303924,
+      "grad_norm": 11.17496395111084,
+      "learning_rate": 1.4393303725115914e-06,
+      "loss": 0.0667,
+      "step": 19725
+    },
+    {
+      "epoch": 0.8796151961875919,
+      "grad_norm": 110.51337432861328,
+      "learning_rate": 1.4133843602089847e-06,
+      "loss": 0.0893,
+      "step": 19750
+    },
+    {
+      "epoch": 0.8807286331447913,
+      "grad_norm": 0.24963940680027008,
+      "learning_rate": 1.387665769074167e-06,
+      "loss": 0.0753,
+      "step": 19775
+    },
+    {
+      "epoch": 0.8818420701019908,
+      "grad_norm": 0.39186498522758484,
+      "learning_rate": 1.362174913793366e-06,
+      "loss": 0.0504,
+      "step": 19800
+    },
+    {
+      "epoch": 0.8829555070591903,
+      "grad_norm": 0.00791653897613287,
+      "learning_rate": 1.3369121062662948e-06,
+      "loss": 0.0943,
+      "step": 19825
+    },
+    {
+      "epoch": 0.8840689440163898,
+      "grad_norm": 0.08412867039442062,
+      "learning_rate": 1.3118776556023338e-06,
+      "loss": 0.0733,
+      "step": 19850
+    },
+    {
+      "epoch": 0.8851823809735893,
+      "grad_norm": 90.87469482421875,
+      "learning_rate": 1.2870718681167427e-06,
+      "loss": 0.02,
+      "step": 19875
+    },
+    {
+      "epoch": 0.8862958179307887,
+      "grad_norm": 11.460103988647461,
+      "learning_rate": 1.2624950473269148e-06,
+      "loss": 0.0946,
+      "step": 19900
+    },
+    {
+      "epoch": 0.8874092548879883,
+      "grad_norm": 0.0059192064218223095,
+      "learning_rate": 1.238147493948665e-06,
+      "loss": 0.0185,
+      "step": 19925
+    },
+    {
+      "epoch": 0.8885226918451877,
+      "grad_norm": 0.22380521893501282,
+      "learning_rate": 1.2140295058925533e-06,
+      "loss": 0.0008,
+      "step": 19950
+    },
+    {
+      "epoch": 0.8896361288023872,
+      "grad_norm": 0.2047863006591797,
+      "learning_rate": 1.1901413782602322e-06,
+      "loss": 0.0572,
+      "step": 19975
+    },
+    {
+      "epoch": 0.8907495657595866,
+      "grad_norm": 0.008318849839270115,
+      "learning_rate": 1.166483403340839e-06,
+      "loss": 0.0353,
+      "step": 20000
+    },
+    {
+      "epoch": 0.8918630027167862,
+      "grad_norm": 0.06174292787909508,
+      "learning_rate": 1.1430558706074212e-06,
+      "loss": 0.0302,
+      "step": 20025
+    },
+    {
+      "epoch": 0.8929764396739857,
+      "grad_norm": 0.0325322151184082,
+      "learning_rate": 1.1198590667134002e-06,
+      "loss": 0.0282,
+      "step": 20050
+    },
+    {
+      "epoch": 0.8940898766311851,
+      "grad_norm": 0.09468420594930649,
+      "learning_rate": 1.096893275489046e-06,
+      "loss": 0.0445,
+      "step": 20075
+    },
+    {
+      "epoch": 0.8952033135883847,
+      "grad_norm": 17.444026947021484,
+      "learning_rate": 1.0741587779380192e-06,
+      "loss": 0.062,
+      "step": 20100
+    },
+    {
+      "epoch": 0.8963167505455841,
+      "grad_norm": 0.02685241959989071,
+      "learning_rate": 1.0516558522339348e-06,
+      "loss": 0.0764,
+      "step": 20125
+    },
+    {
+      "epoch": 0.8974301875027836,
+      "grad_norm": 0.0349482037127018,
+      "learning_rate": 1.029384773716946e-06,
+      "loss": 0.0352,
+      "step": 20150
+    },
+    {
+      "epoch": 0.898543624459983,
+      "grad_norm": 0.06942135840654373,
+      "learning_rate": 1.0073458148903858e-06,
+      "loss": 0.0864,
+      "step": 20175
+    },
+    {
+      "epoch": 0.8996570614171826,
+      "grad_norm": 0.1278313398361206,
+      "learning_rate": 9.85539245417424e-07,
+      "loss": 0.0696,
+      "step": 20200
+    },
+    {
+      "epoch": 0.9007704983743821,
+      "grad_norm": 0.14868886768817902,
+      "learning_rate": 9.639653321177823e-07,
+      "loss": 0.0709,
+      "step": 20225
+    },
+    {
+      "epoch": 0.9018839353315815,
+      "grad_norm": 11.872529029846191,
+      "learning_rate": 9.426243389644463e-07,
+      "loss": 0.0779,
+      "step": 20250
+    },
+    {
+      "epoch": 0.902997372288781,
+      "grad_norm": 0.015093258582055569,
+      "learning_rate": 9.215165270804616e-07,
+      "loss": 0.1371,
+      "step": 20275
+    },
+    {
+      "epoch": 0.9041108092459805,
+      "grad_norm": 0.31670305132865906,
+      "learning_rate": 9.006421547357158e-07,
+      "loss": 0.0841,
+      "step": 20300
+    },
+    {
+      "epoch": 0.90522424620318,
+      "grad_norm": 47.879825592041016,
+      "learning_rate": 8.800014773438015e-07,
+      "loss": 0.0608,
+      "step": 20325
+    },
+    {
+      "epoch": 0.9063376831603794,
+      "grad_norm": 11.414122581481934,
+      "learning_rate": 8.595947474588651e-07,
+      "loss": 0.1249,
+      "step": 20350
+    },
+    {
+      "epoch": 0.9074511201175789,
+      "grad_norm": 0.02606755867600441,
+      "learning_rate": 8.394222147725362e-07,
+      "loss": 0.0606,
+      "step": 20375
+    },
+    {
+      "epoch": 0.9085645570747785,
+      "grad_norm": 0.014529196545481682,
+      "learning_rate": 8.194841261108655e-07,
+      "loss": 0.0362,
+      "step": 20400
+    },
+    {
+      "epoch": 0.9096779940319779,
+      "grad_norm": 11.685364723205566,
+      "learning_rate": 7.997807254313117e-07,
+      "loss": 0.0832,
+      "step": 20425
+    },
+    {
+      "epoch": 0.9107914309891774,
+      "grad_norm": 0.1444302648305893,
+      "learning_rate": 7.803122538197371e-07,
+      "loss": 0.0661,
+      "step": 20450
+    },
+    {
+      "epoch": 0.9119048679463769,
+      "grad_norm": 0.026668483391404152,
+      "learning_rate": 7.610789494874793e-07,
+      "loss": 0.0679,
+      "step": 20475
+    },
+    {
+      "epoch": 0.9130183049035764,
+      "grad_norm": 0.5154994130134583,
+      "learning_rate": 7.420810477684325e-07,
+      "loss": 0.0381,
+      "step": 20500
+    },
+    {
+      "epoch": 0.9141317418607758,
+      "grad_norm": 0.7168388366699219,
+      "learning_rate": 7.233187811161513e-07,
+      "loss": 0.0598,
+      "step": 20525
+    },
+    {
+      "epoch": 0.9152451788179753,
+      "grad_norm": 0.0393916517496109,
+      "learning_rate": 7.047923791010269e-07,
+      "loss": 0.0228,
+      "step": 20550
+    },
+    {
+      "epoch": 0.9163586157751749,
+      "grad_norm": 0.11829303205013275,
+      "learning_rate": 6.865020684074686e-07,
+      "loss": 0.0124,
+      "step": 20575
+    },
+    {
+      "epoch": 0.9174720527323743,
+      "grad_norm": 0.14193271100521088,
+      "learning_rate": 6.684480728311315e-07,
+      "loss": 0.0468,
+      "step": 20600
+    },
+    {
+      "epoch": 0.9185854896895738,
+      "grad_norm": 0.0144277885556221,
+      "learning_rate": 6.506306132761797e-07,
+      "loss": 0.0809,
+      "step": 20625
+    },
+    {
+      "epoch": 0.9196989266467732,
+      "grad_norm": 0.4156523644924164,
+      "learning_rate": 6.330499077525765e-07,
+      "loss": 0.0279,
+      "step": 20650
+    },
+    {
+      "epoch": 0.9208123636039728,
+      "grad_norm": 8.571426391601562,
+      "learning_rate": 6.157061713734269e-07,
+      "loss": 0.0859,
+      "step": 20675
+    },
+    {
+      "epoch": 0.9219258005611722,
+      "grad_norm": 0.005042167380452156,
+      "learning_rate": 5.985996163523422e-07,
+      "loss": 0.0348,
+      "step": 20700
+    },
+    {
+      "epoch": 0.9230392375183717,
+      "grad_norm": 0.023421961814165115,
+      "learning_rate": 5.817304520008371e-07,
+      "loss": 0.0634,
+      "step": 20725
+    },
+    {
+      "epoch": 0.9241526744755711,
+      "grad_norm": 0.03484127297997475,
+      "learning_rate": 5.650988847257743e-07,
+      "loss": 0.0419,
+      "step": 20750
+    },
+    {
+      "epoch": 0.9252661114327707,
+      "grad_norm": 0.03696205094456673,
+      "learning_rate": 5.487051180268443e-07,
+      "loss": 0.0773,
+      "step": 20775
+    },
+    {
+      "epoch": 0.9263795483899702,
+      "grad_norm": 0.3316960334777832,
+      "learning_rate": 5.325493524940628e-07,
+      "loss": 0.0473,
+      "step": 20800
+    },
+    {
+      "epoch": 0.9274929853471696,
+      "grad_norm": 0.13512980937957764,
+      "learning_rate": 5.166317858053261e-07,
+      "loss": 0.0007,
+      "step": 20825
+    },
+    {
+      "epoch": 0.9286064223043691,
+      "grad_norm": 0.04074832424521446,
+      "learning_rate": 5.009526127239883e-07,
+      "loss": 0.0653,
+      "step": 20850
+    },
+    {
+      "epoch": 0.9297198592615686,
+      "grad_norm": 76.54032135009766,
+      "learning_rate": 4.855120250964773e-07,
+      "loss": 0.0371,
+      "step": 20875
+    },
+    {
+      "epoch": 0.9308332962187681,
+      "grad_norm": 0.2280624508857727,
+      "learning_rate": 4.7031021184995585e-07,
+      "loss": 0.0385,
+      "step": 20900
+    },
+    {
+      "epoch": 0.9319467331759675,
+      "grad_norm": 0.10697930306196213,
+      "learning_rate": 4.553473589899926e-07,
+      "loss": 0.0673,
+      "step": 20925
+    },
+    {
+      "epoch": 0.9330601701331671,
+      "grad_norm": 0.012013726867735386,
+      "learning_rate": 4.4062364959830626e-07,
+      "loss": 0.0798,
+      "step": 20950
+    },
+    {
+      "epoch": 0.9341736070903666,
+      "grad_norm": 0.011254810728132725,
+      "learning_rate": 4.2613926383051173e-07,
+      "loss": 0.0832,
+      "step": 20975
+    },
+    {
+      "epoch": 0.935287044047566,
+      "grad_norm": 0.25895559787750244,
+      "learning_rate": 4.1189437891392183e-07,
+      "loss": 0.054,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9364004810047655,
+      "grad_norm": 0.27853846549987793,
+      "learning_rate": 3.9788916914537566e-07,
+      "loss": 0.0545,
+      "step": 21025
+    },
+    {
+      "epoch": 0.937513917961965,
+      "grad_norm": 0.04974246025085449,
+      "learning_rate": 3.8412380588910503e-07,
+      "loss": 0.0477,
+      "step": 21050
+    },
+    {
+      "epoch": 0.9386273549191645,
+      "grad_norm": 0.305239200592041,
+      "learning_rate": 3.7059845757464686e-07,
+      "loss": 0.0209,
+      "step": 21075
+    },
+    {
+      "epoch": 0.939740791876364,
+      "grad_norm": 0.027433935552835464,
+      "learning_rate": 3.573132896947673e-07,
+      "loss": 0.1036,
+      "step": 21100
+    },
+    {
+      "epoch": 0.9408542288335634,
+      "grad_norm": 38.015682220458984,
+      "learning_rate": 3.4426846480345224e-07,
+      "loss": 0.0664,
+      "step": 21125
+    },
+    {
+      "epoch": 0.941967665790763,
+      "grad_norm": 17.695173263549805,
+      "learning_rate": 3.3146414251390423e-07,
+      "loss": 0.0429,
+      "step": 21150
+    },
+    {
+      "epoch": 0.9430811027479624,
+      "grad_norm": 1.5212093591690063,
+      "learning_rate": 3.189004794966022e-07,
+      "loss": 0.0012,
+      "step": 21175
+    },
+    {
+      "epoch": 0.9441945397051619,
+      "grad_norm": 13.568913459777832,
+      "learning_rate": 3.0657762947737815e-07,
+      "loss": 0.0847,
+      "step": 21200
+    },
+    {
+      "epoch": 0.9453079766623613,
+      "grad_norm": 0.4400452971458435,
+      "learning_rate": 2.9449574323552996e-07,
+      "loss": 0.0533,
+      "step": 21225
+    },
+    {
+      "epoch": 0.9464214136195609,
+      "grad_norm": 0.020213905721902847,
+      "learning_rate": 2.826549686019875e-07,
+      "loss": 0.0222,
+      "step": 21250
+    },
+    {
+      "epoch": 0.9475348505767603,
+      "grad_norm": 0.015498008579015732,
+      "learning_rate": 2.710554504575047e-07,
+      "loss": 0.0243,
+      "step": 21275
+    },
+    {
+      "epoch": 0.9486482875339598,
+      "grad_norm": 79.78962707519531,
+      "learning_rate": 2.596973307308726e-07,
+      "loss": 0.0592,
+      "step": 21300
+    },
+    {
+      "epoch": 0.9497617244911594,
+      "grad_norm": 0.05232081189751625,
+      "learning_rate": 2.485807483971958e-07,
+      "loss": 0.0283,
+      "step": 21325
+    },
+    {
+      "epoch": 0.9508751614483588,
+      "grad_norm": 0.22005562484264374,
+      "learning_rate": 2.3770583947619218e-07,
+      "loss": 0.0032,
+      "step": 21350
+    },
+    {
+      "epoch": 0.9519885984055583,
+      "grad_norm": 0.5018379092216492,
+      "learning_rate": 2.270727370305159e-07,
+      "loss": 0.0475,
+      "step": 21375
+    },
+    {
+      "epoch": 0.9531020353627577,
+      "grad_norm": 0.07405375689268112,
+      "learning_rate": 2.1668157116414346e-07,
+      "loss": 0.0568,
+      "step": 21400
+    },
+    {
+      "epoch": 0.9542154723199573,
+      "grad_norm": 13.536460876464844,
+      "learning_rate": 2.0653246902077263e-07,
+      "loss": 0.0938,
+      "step": 21425
+    },
+    {
+      "epoch": 0.9553289092771567,
+      "grad_norm": 0.15782712399959564,
+      "learning_rate": 1.9662555478227484e-07,
+      "loss": 0.0847,
+      "step": 21450
+    },
+    {
+      "epoch": 0.9564423462343562,
+      "grad_norm": 0.08711089193820953,
+      "learning_rate": 1.8696094966716537e-07,
+      "loss": 0.0362,
+      "step": 21475
+    },
+    {
+      "epoch": 0.9575557831915557,
+      "grad_norm": 0.024433070793747902,
+      "learning_rate": 1.7753877192913104e-07,
+      "loss": 0.0958,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9586692201487552,
+      "grad_norm": 0.07410872727632523,
+      "learning_rate": 1.683591368555737e-07,
+      "loss": 0.0631,
+      "step": 21525
+    },
+    {
+      "epoch": 0.9597826571059547,
+      "grad_norm": 0.058046724647283554,
+      "learning_rate": 1.5942215676620908e-07,
+      "loss": 0.0167,
+      "step": 21550
+    },
+    {
+      "epoch": 0.9608960940631541,
+      "grad_norm": 0.18262982368469238,
+      "learning_rate": 1.5072794101168132e-07,
+      "loss": 0.0668,
+      "step": 21575
+    },
+    {
+      "epoch": 0.9620095310203536,
+      "grad_norm": 0.008218363858759403,
+      "learning_rate": 1.4227659597223719e-07,
+      "loss": 0.0456,
+      "step": 21600
+    },
+    {
+      "epoch": 0.9631229679775531,
+      "grad_norm": 13.682443618774414,
+      "learning_rate": 1.340682250564096e-07,
+      "loss": 0.0474,
+      "step": 21625
+    },
+    {
+      "epoch": 0.9642364049347526,
+      "grad_norm": 0.005989130586385727,
+      "learning_rate": 1.2610292869977392e-07,
+      "loss": 0.0474,
+      "step": 21650
+    },
+    {
+      "epoch": 0.965349841891952,
+      "grad_norm": 0.015712061896920204,
+      "learning_rate": 1.1838080436369359e-07,
+      "loss": 0.0892,
+      "step": 21675
+    },
+    {
+      "epoch": 0.9664632788491515,
+      "grad_norm": 0.28373441100120544,
+      "learning_rate": 1.1090194653414543e-07,
+      "loss": 0.0158,
+      "step": 21700
+    },
+    {
+      "epoch": 0.9675767158063511,
+      "grad_norm": 0.08839291334152222,
+      "learning_rate": 1.0366644672056059e-07,
+      "loss": 0.0526,
+      "step": 21725
+    },
+    {
+      "epoch": 0.9686901527635505,
+      "grad_norm": 0.43165767192840576,
+      "learning_rate": 9.667439345469875e-08,
+      "loss": 0.0482,
+      "step": 21750
+    },
+    {
+      "epoch": 0.96980358972075,
+      "grad_norm": 0.2389592081308365,
+      "learning_rate": 8.992587228957128e-08,
+      "loss": 0.0577,
+      "step": 21775
+    },
+    {
+      "epoch": 0.9709170266779495,
+      "grad_norm": 0.033785369247198105,
+      "learning_rate": 8.342096579839087e-08,
+      "loss": 0.0506,
+      "step": 21800
+    },
+    {
+      "epoch": 0.972030463635149,
+      "grad_norm": 0.030277960002422333,
+      "learning_rate": 7.715975357356131e-08,
+      "loss": 0.0483,
+      "step": 21825
+    },
+    {
+      "epoch": 0.9731439005923485,
+      "grad_norm": 0.016112027689814568,
+      "learning_rate": 7.11423122257049e-08,
+      "loss": 0.0405,
+      "step": 21850
+    },
+    {
+      "epoch": 0.9742573375495479,
+      "grad_norm": 23.555667877197266,
+      "learning_rate": 6.536871538272538e-08,
+      "loss": 0.043,
+      "step": 21875
+    },
+    {
+      "epoch": 0.9753707745067475,
+      "grad_norm": 0.23819568753242493,
+      "learning_rate": 5.983903368890653e-08,
+      "loss": 0.0304,
+      "step": 21900
+    },
+    {
+      "epoch": 0.9764842114639469,
+      "grad_norm": 17.99365997314453,
+      "learning_rate": 5.45533348040439e-08,
+      "loss": 0.0104,
+      "step": 21925
+    },
+    {
+      "epoch": 0.9775976484211464,
+      "grad_norm": 0.010882832109928131,
+      "learning_rate": 4.9511683402627683e-08,
+      "loss": 0.0709,
+      "step": 21950
+    },
+    {
+      "epoch": 0.9787110853783458,
+      "grad_norm": 3.7435531616210938,
+      "learning_rate": 4.471414117303896e-08,
+      "loss": 0.07,
+      "step": 21975
+    },
+    {
+      "epoch": 0.9798245223355454,
+      "grad_norm": 0.04117279127240181,
+      "learning_rate": 4.0160766816796925e-08,
+      "loss": 0.0327,
+      "step": 22000
+    },
+    {
+      "epoch": 0.9809379592927449,
+      "grad_norm": 0.08280256390571594,
+      "learning_rate": 3.585161604785503e-08,
+      "loss": 0.0667,
+      "step": 22025
+    },
+    {
+      "epoch": 0.9820513962499443,
+      "grad_norm": 0.022770730778574944,
+      "learning_rate": 3.17867415918971e-08,
+      "loss": 0.085,
+      "step": 22050
+    },
+    {
+      "epoch": 0.9831648332071438,
+      "grad_norm": 0.05374102294445038,
+      "learning_rate": 2.7966193185706702e-08,
+      "loss": 0.06,
+      "step": 22075
+    },
+    {
+      "epoch": 0.9842782701643433,
+      "grad_norm": 10.886147499084473,
+      "learning_rate": 2.4390017576561008e-08,
+      "loss": 0.0439,
+      "step": 22100
+    },
+    {
+      "epoch": 0.9853917071215428,
+      "grad_norm": 0.22477033734321594,
+      "learning_rate": 2.1058258521642337e-08,
+      "loss": 0.0363,
+      "step": 22125
+    },
+    {
+      "epoch": 0.9865051440787422,
+      "grad_norm": 0.01279439590871334,
+      "learning_rate": 1.797095678752303e-08,
+      "loss": 0.0624,
+      "step": 22150
+    },
+    {
+      "epoch": 0.9876185810359418,
+      "grad_norm": 0.14910484850406647,
+      "learning_rate": 1.51281501496503e-08,
+      "loss": 0.0406,
+      "step": 22175
+    },
+    {
+      "epoch": 0.9887320179931413,
+      "grad_norm": 13.914883613586426,
+      "learning_rate": 1.2529873391895486e-08,
+      "loss": 0.0638,
+      "step": 22200
+    },
+    {
+      "epoch": 0.9898454549503407,
+      "grad_norm": 0.388899564743042,
+      "learning_rate": 1.0176158306118844e-08,
+      "loss": 0.091,
+      "step": 22225
+    },
+    {
+      "epoch": 0.9909588919075402,
+      "grad_norm": 19.062267303466797,
+      "learning_rate": 8.06703369178985e-09,
+      "loss": 0.081,
+      "step": 22250
+    },
+    {
+      "epoch": 0.9920723288647397,
+      "grad_norm": 0.013312091119587421,
+      "learning_rate": 6.202525355627487e-09,
+      "loss": 0.0281,
+      "step": 22275
+    },
+    {
+      "epoch": 0.9931857658219392,
+      "grad_norm": 0.02743523009121418,
+      "learning_rate": 4.582656111289385e-09,
+      "loss": 0.1062,
+      "step": 22300
+    },
+    {
+      "epoch": 0.9942992027791386,
+      "grad_norm": 0.01560298539698124,
+      "learning_rate": 3.2074457790876036e-09,
+      "loss": 0.1068,
+      "step": 22325
+    },
+    {
+      "epoch": 0.9954126397363381,
+      "grad_norm": 45.84857177734375,
+      "learning_rate": 2.0769111857510405e-09,
+      "loss": 0.0872,
+      "step": 22350
+    },
+    {
+      "epoch": 0.9965260766935377,
+      "grad_norm": 0.06956358253955841,
+      "learning_rate": 1.1910661642189348e-09,
+      "loss": 0.1165,
+      "step": 22375
+    },
+    {
+      "epoch": 0.9976395136507371,
+      "grad_norm": 0.021001625806093216,
+      "learning_rate": 5.499215534676694e-10,
+      "loss": 0.0312,
+      "step": 22400
+    },
+    {
+      "epoch": 0.9987529506079366,
+      "grad_norm": 0.00945682730525732,
+      "learning_rate": 1.5348519838198628e-10,
+      "loss": 0.0877,
+      "step": 22425
+    },
+    {
+      "epoch": 0.999866387565136,
+      "grad_norm": 25.856142044067383,
+      "learning_rate": 1.761949657286266e-12,
+      "loss": 0.0302,
+      "step": 22450
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.9805008477892265,
+      "eval_auc": 0.9956209792829493,
+      "eval_f1": 0.9869591765526867,
+      "eval_loss": 0.10316640883684158,
+      "eval_precision": 0.9820335893763833,
+      "eval_recall": 0.9919344233551045,
+      "eval_runtime": 4675.5639,
+      "eval_samples_per_second": 6.559,
+      "eval_steps_per_second": 0.273,
+      "step": 22453
+    }
+  ],
+  "logging_steps": 25,
+  "max_steps": 22453,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 5,
+        "early_stopping_threshold": 0.01
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 0
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.204190364178894e+19,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}